Letnik XXVII, številka 1-2, 2016 ^ Revija za teorijo in raziskave vzgoje in izobraževanja Šolsko polje Social and Emotional Aspects of Teaching and Learning ed. Maša Vidmar Šolsko polje Revija za teorijo in raziskave vzgoje in izobraževanja Letnik XXVII, številka 1-2, 2016 Šolsko polje je mednarodna revija za teorijo ter raziskave vzgoje in izobraževanja z mednarodnim uredniškim odborom. Objavlja znanstvene in strokovne članke s širšega področja vzgoje in izobraževanja ter edukacijskih raziskav (filozofija vzgoje, sociologija izobraževanja, uporabna epistemologija, razvojna psihologija, pedagogika, andragogika, pedagoška metodologija itd.), pregledne članke z omenjenih področij ter recenzije tako domačih kot tujih monografij s področja vzgoje in izobraževanja. Revija izhaja trikrat letno. Izdajajo Slovensko društvo raziskovalcev šolskega polja. Poglavitni namen revije je prispevati k razvoju edukacijskih ved in interdisciplinarnemu pristopu k teoretičnim in praktičnim vprašanjem vzgoje in izobraževanja. V tem okviru revija posebno pozornost namenja razvijanju slovenske znanstvene in strokovne terminologije ter konceptov na področju vzgoje in izobraževanja ter raziskovalnim paradigmam s področja edukacijskih raziskav v okviru družboslovno-humanističnih ved. Uredništvo: Valerija Vendramin, Zdenko Kodelja, Darko Štrajn, Alenka Gril, Igor Ž. Žagar, Eva Klemenčič in Mitja Sardoč (vsi: Pedagoški inštitut, Ljubljana) Glavni urednik: Marjan Šimenc (Pedagoški inštitut, Ljubljana) Odgovorna urednica: Mojca Štraus (Pedagoški inštitut, Ljubljana) Uredniški odbor: Michael W. Apple (University ofWisconsin, Madison, USA), Eva D. Bahovec (Filozofska fakulteta, Univerza v Ljubljani), Andreja Barle-Lakota (Urad za šolstvo, Ministrstvo za šolstvo in šport RS), Valentin Bucik (Filozofska fakulteta, Univerza v Ljubljani), Harry Brighouse (University ofWisconsin, Madison, USA), Randall Curren (University of Rochester, USA), Slavko Gaber (Pedagoška fakulteta, Univerza v Ljubljani), Milena Ivanuš-Grmek (Pedagoška fakulteta, Univerza v Mariboru), Russell Jacoby (University of California, Los Angeles), Janez Justin t (Pedagoški inštitut, Ljubljana), Stane Košir (Pedagoška fakulteta, Univerza v Ljubljani), Janez Kolenc t (Pedagoški inštitut, Ljubljana), Ljubica Marjanovič-Umek (Filozofska fakulteta, Univerza v Ljubljani), Rastko Močnik (Filozofska fakulteta, Univerza v Ljubljani), Zoran Pavlo-vič (Svetovalni center za otroke, mladostnike in starše, Ljubljana), Drago B. Rotar (Fakulteta za humanistične študije, Univerza na Primorskem), Harvey Siegel (University ofMiami, USA), Marjan Šetinc (Slovensko društvo raziskovalcev šolskega polja, Ljubljana), Pavel Zgaga (Pedagoška fakulteta, Univerza v Ljubljani), Maja Zupančič (Filozofska fakulteta, Univerza v Ljubljani), Robi Kroflič (Filozofska fakulteta, Univerza v Ljubljani), Marie-Helene Esteoule Exel (Universite Stendhal Grenoble III) Lektor (slovenskijezik), tehnični urednik, oblikovanje in prelom: Jonatan Vinkler Lektor (angleški jezik): Jason Brendon Batson Izdajatelja: Slovensko društvo raziskovalcev šolskega polja in Pedagoški inštitut © Slovensko društvo raziskovalcev šolskega polja in Pedagoški inštitut Tisk: Grafika 3000 d.o.o., Dob Naklada: 400 izvodov Revija Šolsko polje je vključena v naslednje indekse in baze podatkov: Contents Pages in Education; EBSCO; Education Research Abstracts; International Bibliography of the Social Sciences (IBSS); Linguistics and Language Behavior Abstracts (LLBA); Multicultural Education Abstracts; Pais International; ProQuest Social SciencesJournal, Research into Higher Education Abstracts; Social Services Abstracts; Sociological Abstracts; Worldwide Political Science Abstracts Šolsko polje izhaja s finančno podporo Pedagoškega inštituta inJavne agencije za raziskovalno dejavnost Republike Slovenije. Tiskana izdaja: ISSN 1581-6036 Izdaja na zgoščenki: ISSN 1581-6052 Spletna izdaja: ISSN 1581-6044 Letnik XXVII, številka 1-2, 2016 Revija za teorijo in raziskave vzgoje in izobraževanja Šolsko polje Social and Emotional Aspects of Teaching and Learning ed. Maša Vidmar SF Contents/Vsebina 1 EDITORIAL/UVODNIK 5 Maša Vidmar ■ Social and Emotional Aspects of Teaching and Learning 7 2 PAPERS/RAZPRAVE 11 Jennifer L. Hanson-Peterson, Kimberly A. Schonert-Reichl and Veronica Smith ■ Teachers' Beliefs about Emotions: Relations to Teacher Characteristics and Social and Emotional Learning Program Implementation 13 Maša Vidmar and Katja Kerman ■ The Development of Teacher's Relational Competence Scale: Structural Validity and Reliability 41 Ana Kozina and Ana Mlekuž ■ Intrinsic Motivation as a Key to School Success: Predictive Power of Self-perceived Autonomy, Competence andRelatedness on the Achievement in International Comparative Studies 63 Urška Aram, Nina Jurinec, Marina Horvat in Katja Košir ■ Samopodoba in socialna sprejetost identificiranih nadarjenih in visoko učno uspešnih osnovnošolcev 89 Klaudija Šterman Ivančič and Melita Puklek Levpušček ■ Motivational Goals and Academic Performance from the Perspective ofStudentsPerceivedQuality ofRelationship with Their Class Teachers at the Start of the Upper Secondary Education Level 113 Per F. Laursen and Anne Maj Nielsen ■ Teachers' Relational Competencies: the Contribution from Teacher Education 139 Jana Hafner and Maja Krajnc ■ Comparison of Interpersonal Communication and Interpersonal Relationship Between Early Years Educators and Children 163 3 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 3 REVIEW/RECENZIJA 185 Ana Mladenovic ■ Wendy R. Kohli in Nicholas C. Burbules (2013), Feminisms andEducationalResearch 187 4 ABSTRACTS/POVZETKI 193 5 AUTHORS/AVTORJI 207 4 1 EDITORIAL UVODNIK Social and Emotional Aspects of Teaching and Learning Masa Vidmar The social and emotional competences of students and its role for students' achievement and other developmental outcomes have received a lot of scientific and policy interest in the last decade (e.g. Durlak et al., 2011; OECD, 2015). Recently, the focus has broadened to include also teachers' social and emotional competences (SEC; Schonert-Reichl, Hanson-Peterson, & Hymel, 2015). In the current thematic issue Social and emotional aspects of teaching and learning we aim to widen this perspective to include also other social and emotional aspects of teaching and learning and present state-of-the-art research in the field. Social and emotional aspects of teaching and learning can be defined as the social and emotional competences and processes in the educational context within an individual (e.g. students' or teachers' emotions), between individuals (e.g. teacher-student relationship) or phenomena emerging as a result of these competences and processes at the classroom (e.g. class climate), school (e.g. school climate and culture) or community level. There are complex concurrent and longitudinal interrelations among these variables, as they contribute to successful learning and teaching as well as to academic achievement and other developmental outcomes. Some of the topics of social and emotional aspects of teaching and learning are well studied within the educational contexts (e.g. student motivation), while others are not well-studied (e.g. teachers' social and emotional competences). The aim of this issue is to elucidate some of the less-talked-about topics and feed into future research, practice, and policy. This thematic issue brings together a wide range of topics, reflecting the diversity and heterogeneity of the field. We are particularly proud to have all 7 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 educational levels represented: from early childhood education and care (see Hafner & Kranjc), through primary, lower, and upper secondary education with a focus on students (see Šterman Ivančič & Puklek Levpušček; Kozina & Mlekuž; Aram, Jurinec, Horvat, & Košir) and teachers (see Hanson-Peterson, Schonert-Reichl, and Smith; Vidmar & Kerman) all the way to tertiary education (see Laursen & Nielsen). Articles are diverse also in the sense of geography; the majority include Slovenian samples (see Šterman Ivančič & Puklek Levpušček; Kozina & Mlekuž; Aram et al.; Vidmar & Kerman), while others include international samples (see Hanson-Peterson et al; Laursen & Nielsen) or focus on the international comparison (see Hafner & Krajnc). The first article by Canadian researchers Hanson-Peterson and colleagues reports on the significant role of teachers' SEC (specifically teachers' emotion beliefs) in the implementation of SEC program as well the role of teachers' background characteristics for teachers' SEC. These findings demonstrate the need for teachers' SEC training in this respect. The following article by Vidmar and Kerman continues the topic of teachers' social and emotional competence by introducing the newly developed Teacher's Relational Competence Scale (TRCS), and examining its construct validity and reliability. It demonstrates that a teachers' respect for individuality and their responsibility for the teacher-student relationship (two dimensions of relational competence) can be reliably measured using the TRCS. In the next three articles, the focus shifts from teachers to students in primary and secondary education. Kozina and Mlekuž examine a series of international studies on student achievement; they find a significant effect of internal motivation (i.e. satisfied need for autonomy, competence, and relatedness) for student achievement; students' satisfied need for competence is the strongest and most consistent predictor of their achievement. In the article by Aram and colleagues the self-concept of gifted and high-achieving students in comparison to other students is examined. In general, they find no differences in academic, peer relations, or general self-concept between gifted and high-achieving students. However, gifted girls are identified as a possible high-risk subgroup of gifted students due to their lower peer relations self-concept. Šterman Ivančič and Puklek Levpušček focus their study on the sample from international study PISA. They focus on motivational goals and students' perceived quality of relationship with teachers (i.e. socio-emotional support and negative interactions). The results indicate that perceived support from teachers is especially important for students' motivational goals, while perceived negative interactions are detrimental for academic achievement. In the last two articles, the qualitative approach is used, 8 M. VIDMAR ■ SOCIAL AND EMOTIONAL ASPECTS OF TEACHING AND LEARNING bringing a wealth of interesting information. Laursen and Nielsen from Denmark present us with their findings on a programme aiming to develop relational competence in initial teacher education. The programme contributed to student-teachers taking a more reflective and experimental approach to teaching. In the last article, Hafner and Krajnc compare English and Slovenian interpersonal communication and interpersonal relationships in early childhood education settings. They conclude that there are not many differences with regards to non-verbal communication (expect for more physical contact to express affection in Slovenian settings), but find more educator-child verbal interactions in Slovenian settings. The future research should deepen our understanding about the interrelations of social and emotional aspects of teaching and learning with various student and teacher outcomes and the quality of the educational process. Implications for educational practice (e.g. how to consider social and emotional aspects of teaching and learning in everyday school interactions) and initial and continuous education of teachers in the field of social and emotional aspects of teaching and learning remain a challenge for future research and practice. It has been an honour to be guest editor of this thematic issue promulgating important issues often overlooked in education practice and policy. I hope this issue, with international authorship, contributes to raising awareness in regard to social and emotional aspects of students and teachers in the educational process and spurs discussion in the field. References Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., & Schellinger, K. B. (2011). Enhancing students' social and emotional development promotes success in school: Results of a meta-analysis. Child Development, 82, 474-501. OECD. (2015). Skills for social progress: the power of social an emotional skills. Paris: OECD Publishing. Schonert-Reichl, K. A., Hanson-Peterson, J. L., & Hymel, S. (2015). Social and emotional learning and pre-service teacher education. In J. A. Durlak, R. P. Weissberg, C. E. Domitrovich, and T. P. Gullotta (Eds.), Handbook of social and emotional learning: Research and practice (pp. 406-421). NY: Guilford. 9 2 PAPERS RAZPRAVE Teachers' Beliefs about Emotions: Relations to Teacher Characteristics and Social and Emotional Learning Program Implementation Jennifer L. Hanson-Peterson, Kimberly A. Schonert-Reichl and Veronica Smith Recent years have witnessed increased theoretical and empirical attention to the school-based promotion of children's social and emotional competence as educators, parents, policymakers, and other societal agencies contemplate solutions for contemporary problems such as declining academic motivation and achievement (Klem and Connell, 2004), increasing school bullying (Swearer et al. 2010), and rises in children's mental health problems (Institute of Medicine, 2009). Schools around the world are adopting social and emotional learning (SEL) programs aimed at preventing these issues and fostering social and emotional competencies (Ransford et al. 2009; Schonert-Reichl and Weissberg, 2014). Although a plethora of research suggests that SEL programs are largely effective, as evidenced in the meta-anal-ysis of SEL programs by Durlak et al. (2011), other evaluations of SEL programs have yielded non-significant findings (see Ransford et al. 2009). As espoused by many SEL researchers, future studies should move beyond the "black box" approach to program evaluation and investigate the role of teachers in delivering SEL programs, specifically teachers' beliefs about emotional socialization practices and the extent to which they deliver the program with fidelity (Beets et al. 2008; Durlak and DuPre, 2008; Jennings and Greenberg, 2009; Zinsser et al. 2014). Through their natural daily interactions, teachers play a critical role in the emotional life of the classroom, including student-teacher and student-student relationships. Teachers enter the classroom with their own levels of social-emotional competence (e.g. mindfulness, self-compassion), which has an impact on the quality of the learning environment, including the amount of emotional support offered to students (Jennings, 2014). Ad- 13 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 ditionally, whether conscious of it or not, teachers are constantly facilitating their students' social and emotional development through emotion socialization practices - modeling and communicating the extent to which students should reflect upon, control, and express their emotions in the classroom (Hargreaves, 2000). Several researchers have predicted and found a relationship between teachers' utilization of supportive emotion socialization practices (e.g. reacting to students' emotions in a supportive way) and their students' emotional competence (e.g. well-adjusted emotion regulation; Denham, Bassett and Zinsser, 2012; Horner and Wallace, 2013). Furthermore, evidence has suggested that students of emotionally sensitive and involved teachers are less likely to exhibit internalizing and externalizing problems (Murray and Greenberg, 2000; Zinsser et al. 2014), and show greater improvements in their social competence (Wilcox-Herzog and Ward, 2004; Zinsser et al. 2014). Despite the recognized influence of teachers on their students' social and emotional development, there remains a paucity of research examining a factor that may be associated with differences in teachers' emotion socialization practices and implementation of SEL programs: teachers' emotion beliefs (see Hyson and Lee, 1996). Teachers' emotion beliefs refer to beliefs that teachers hold about emotions in the classroom setting and their personal role in promoting the emotional development of their students (ibid.). Most of the research to date examining teachers' emotion beliefs has assessed the beliefs of early childhood educators (Ahn, 2005; Gos-ney, 2006; Huemer, 2010; Hyson and Lee, 1996; Jaramillo, 2006; Jumper 2005; for an exception, see Bellas, 2009). Furthermore, only one known study comprised of only early childhood educators has established a link between teachers' emotion beliefs and their implementation of an SEL program (Jaramillo, 2006). Therefore, to our knowledge, the current study is the first of its kind to examine elementary school teachers' emotion beliefs in relation to both their background characteristics and their implementation of an SEL program designed to promote children's emotional competence. Teachers' Background Characteristics and Emotion Beliefs It has been widely suggested that the beliefs people hold are shaped by their backgrounds and personal experiences (Pajares, 1992). A significant body of research has established a link between teachers' background characteristics and a variety of beliefs they hold. For instance, years of teaching experience has been found to be positively and significantly correlated with both teachers' self-efficacy beliefs regarding their ability to manage disruptive behavior and motivate learning in the classroom (Tschannen-Mo- 14 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... ran and Wolfolk Hoy, 2007). A handful of studies exist that have found no relation between years of teaching and emotion beliefs (Bellas, 2009; Hyson and Lee, 1996; Jumper, 2005). Nonetheless, the samples in these previous studies were either fully or partially comprised of early childhood teachers - a population with different educational training and attrition rates than elementary school teachers (Whitebook, 2014). Indeed, little is known about the relation between years of teaching experience and teachers' beliefs about emotions in the elementary school context. Prior research indicates that early childhood teachers' level of education is positively and significantly associated with the developmental appropriateness of their emotion beliefs (i.e. the congruency of these beliefs with their students' age-related social-emotional needs; Hyson and Lee, 1996; Jumper, 2005). The level of teacher preparation, however, can be quite different between elementary school teachers (who are required to have a bachelor's degree, at a minimum) and early childhood educators (with only about 50% holding a bachelor's degree or higher; Whitebook, 2014). Similarly, teachers' emotion socialization practices may depend on the grade level they teach, owing to the age-related differences in the social and emotional competence of children in elementary school compared to early childhood (e.g. advanced reasoning about emotions; Brackett and Rivers, 2008). Some evidence suggests as the grade level that teachers instruct increases, their willingness to engage the children in their classrooms in direct emotion socialization practices decreases. Ahn and Stifter (2006) found, for example, differences in the emotion socialization practices of toddler caregivers versus preschool teachers - with the former teacher group being more willing to physically comfort and verbalize emotions with their children compared to the latter. Taken together, understanding the beliefs elementary school teachers hold about emotions in the classroom can provide insight into whether accredited teacher training programs adequately prepare these teachers to engage in developmen-tally appropriate emotion socialization practices. Teachers' Emotion Beliefs and Emotion Socialization Practices Although teachers' beliefs about using particular emotion socialization practices and their execution of those emotional socialization practices are two independent processes, a body of empirical evidence indicates a link between these two factors (e.g. Ahn, 2005; Bellas, 2009). Some research has been conducted on teachers' emotion beliefs and their emotion socialization practices in classrooms not hosting a specific SEL program. In one study, it was found that the teachers' beliefs regarding the 15 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 importance of particular emotion socialization practices were relatively consistent with their actual emotion socialization practices via classroom observations (Ahn, 2005). In another study, teachers high in emotion support for students, compared to those who were moderately supportive as determined via quantitative assessments of their classroom interactions, were more likely to express in focus groups that they (i) held the beliefs that SEL was an integral part of interacting with their students and were equally as accountable as parents to foster children's emotional development, and (ii) purposefully used explicit emotion socialization practices with their students (e.g. drawing attention to and labelling emotions; Zinsser et al. 2014). Moreover, teachers with less developmentally appropriate emotion beliefs have been found to be more likely to react negatively (e.g. use punitive practices, minimize students' emotions) to their students' negative emotion expressions (Gosney, 2006). Gosney (2006), however, found that more developmentally appropriate emotion beliefs did not predict teachers' positive reactions to students' negative emotion expressions. Collectively, these findings highlight the importance of gaining more knowledge about teachers' emotion beliefs and emotion socialization practices. Research in this area is particularly important in classrooms hosting emotion-focused SEL programs, as the programs' structured activities may support teachers' sense of efficacy and motivation to deliberately carry out direct emotion socialization practices aimed at promoting their students' emotional competence. Teachers' Emotion Beliefs and SEL Program Implementation Burgeoning evidence indicates that the extent to which teachers implement preventive intervention programs with fidelity (quality and degree of implementation) is associated with the effectiveness of these programs (Durlak, 2015; Durlak and DuPre, 2008; Ransford et al. 2009). Although it is beyond the scope of the present study to examine program effectiveness, it is valuable to attempt to extend the understanding of underlying mechanisms that impact implementation fidelity. It has been suggested that future studies examine teacher-related factors that impact variations in the implementation of evidence-based preventive intervention programs as these variations affect the quality of the program and may undermine its success (Durlak and DuPre, 2008; Wanless and Domitro-vich, 2015). Further, Wanless and Domitrovich (2015) highlight that examining teacher-related factors that are present before the delivery of the program - such as beliefs, knowledge, and skills - expands the scant lit- 16 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... erature on "indicators of readiness" to implement the program (p. 1038). Research has found that teachers' implementation fidelity of SEL program lessons and practices is associated with a number of teacher beliefs: beliefs about whether the SEL program activities are aligned with their teaching approach (ibid.); beliefs about behavior management practices (Rimm-Kaufman and Sawyer, 2004); self-efficacy beliefs about teaching (Ransford et al. 2009; Reyes, et al. 2012); the level of comfort delivering the SEL curriculum (Brackett et al. 2012); dedication to developing their SEL skills (ibid.); perceptions of whether the school culture supports SEL instruction (ibid.); and perceptions of whether the school leader supports an SEL program (Brown et al. 2010). Nonetheless, these previous studies did not examine the teachers' beliefs that are specifically relevant to an underlying philosophy of emotion-focused SEL programs, that is, emotional competence can be enhanced through direct instruction and is not an innate or fixed characteristic (Gordon, 2000; Kress and Elias, 2006). Therefore, for teachers hosting an SEL program in their classrooms that aims to enhance students' emotional competence, the emotion beliefs of those teachers may be associated with differences in their emotion socialization practices as evidenced by the extent to which they implement SEL program activities. Only one study to date has examined teachers' emotion beliefs in relation to frequency of implementation of SEL program activities. Jar-amillo (2006) found that early childhood teachers' expressiveness beliefs (beliefs about teachers' candid expression of emotions around their students) were significantly and negatively correlated with the amount of SEL program activities they implemented. That is, teachers who reported being uncomfortable or unwilling to be emotionally expressive with their students were less likely to engage their students in emotion-focused activities in an SEL program, in contrast to teachers who reported being emotionally expressive in their interactions with students (ibid.). Clearly, additional research is needed to better understand the association between elementary school teachers' emotion beliefs and their implementation of SEL programs, particularly with regard to SEL programs that are emotion-focused in their intent. The Roots of Empathy: An Emotion-Focused SEL Program For the current study, teachers' emotion beliefs and their implementation of extension activities were examined in the context of one SEL program - the Roots of Empathy (ROE). ROE is a classroom-based SEL program for children from Kindergarten to 8th grade. The goal of the ROE pro- 17 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 gram is to increase students' emotional competence (i.e. emotional understanding, perspective-taking) and prosocial behaviors, and to decrease students' aggressive behaviors (Gordon, 2000). To date, there have been several outcome studies examining the efficacy of ROE (see Schonert-Reichl and Scott, 2009 for a review). Overall, research on the effectiveness of ROE has yielded consistent and highly promising findings regarding the impact of the program across age and gender. For instance, Schonert-Reichl et al. (2012) found that 4th to 7th grade children who participated in ROE, compared to those who did not, demonstrated advanced emotional and social understanding, as well as reduced aggressive behavior and increased prosocial behavior. Moreover, Santos et al. (2011), in their cluster randomized controlled field study and longitudinal follow-up of the ROE program, found that the positive effects of the program in decreasing aggression and increasing prosocial behavior were either maintained or improved, even after the program had ended. A trained and certified ROE instructor facilitates the ROE program over the course of nine months, and visits the classroom three times each month - a pre-family visit, a family visit, and a post-family visit. The cornerstone of the ROE program is the monthly ROE family visits, involving an infant and his/her parent(s) who visit the classroom to serve as a springboard for lessons on emotion knowledge, perspective-taking, and infant development. During these monthly visits, the ROE instructor encourages the students to observe the baby's emotional, social, and physical development, and reflect on the parent-infant bond and the ways in which the infant "relies on the parent to understand the world, to feel safe to explore and to learn how to regulate his/her emotions" (Roots of Empathy, 2013, para. 8). During the pre- and post-family visits, the ROE instructor draws on the students' observations of the infant via a series of se-quenced ROE lessons. These lessons are age-appropriate to the students and progress with the baby's natural development. The 27 ROE lessons consist of nine different themes: Meeting the Baby, Crying, Caring and Planning for the Baby, Emotions, Sleep, Safety, Communication, Who am I?, Goodbye and Good Wishes. All of the lessons are designed to help children understand and reflect on their own feelings and the feelings of others. Each ROE lesson aims to develop students' emotional competence by providing them with rich opportunities to identify, explain, and react to the emotions of the ROE infant, and discuss the emotions experienced by the infant, themselves, and others. For instance, during a structured ROE lesson, students are directed to recognize the nonverbal cues 18 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... and facial expressions of the infant (e.g. sadness, fear) and label the emotion, and are encouraged to engage in perspective-taking to identify possible reasons why the infant may be experiencing that emotion (e.g. having had a toy taken away, hearing a loud noise). Then, through a variety of additional activities (e.g. book discussions, art projects), the students are encouraged to reflect on their own and others' experiences with the emotion the infant was feeling (e.g. feeling sad or afraid). For the Caring and Planning for the Baby theme, for example, the instructor reads a story to the students about a young girl losing her first tooth. After the story, the instructor asks the students about the various emotions that can be experienced from losing a tooth (e.g. worry that others will laugh about the missing tooth). The group discussion gives students the opportunity not only to discuss their emotions, but to practice empathy through learning about and respecting their classmates' emotions as well. In the family visit, the students are encouraged to engage in perspective-taking by asking the infant's parent questions about the infant's experiences with teething (e.g. "How does it make you feel to see your baby in pain?" "What do you try to do to make your baby's pain go away?"). As posited by Schonert-Re-ichl et al. (2012), ROE draws on the functionalist approach to emotions, wherein emotion understanding and expressivity are seen as playing central roles in the establishment and maintenance of children's interpersonal relationships (Saarni, 2011). Learning to reflect upon, label, discuss, and express emotions helps students learn to regulate and exhibit their emotions in socially acceptable ways. Therefore, students are better equipped to demonstrate greater empathy, and accordingly more prosocial behavior and less aggression towards others (Schonert-Reichl et al. 2012). Although the ROE instructor facilitates the lessons and thus is the primary implementer of the program, in the ROE model the classroom teacher is encouraged to reinforce the valued concepts promoted by the ROE program by integrating extension activities into the existing academic curriculum (e.g. language arts lessons; Gordon, 2000). The instructor provides the teacher with several resources that can aid in the design of extension activities, including the ROE curriculum manual that outlines the goals and activities of the program, the lesson plans for each visit, and references to additional resources (ibid.). Despite their role as secondary implementers, no research to date has examined the ways in which teachers' beliefs about emotions may influence their implementation of ROE extension activities. Hence, examining elementary school teachers' emotion beliefs in association with their implementation of the ROE program extension activities can be highly informative. This information can 19 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 fill several gaps in the knowledge about teachers' emotion beliefs and the manner in which they deliver an emotion-focused SEL program. The Current Study The objectives of the current study were twofold: (a) to examine relations of teachers' background characteristics (i.e. elementary grade level taught, years of teaching experience) to their emotion beliefs, and (b) to examine relations of teachers' emotion beliefs to the implementation of extension activities in the context of one emotion-focused SEL program - the ROE program. Data for this study were drawn from two studies investigating the effectiveness of the ROE program - one of which was a randomized controlled trial (RCT) and the other a quasi-experimental study. To investigate the first research question, given the question is not concerned with teachers' implementation of a specific SEL program, pretest data for the intervention group and control group were combined in order to yield a larger sample size. To investigate the second research question, only the intervention group teacher data were analyzed since the control group did not implement the SEL program. Method For the present investigation, as noted above, data were derived from two studies of ROE - an RCT and a quasi-experimental design study. The former ROE study took place in a large urban public school district serving approximately 55,000 students located in a Western Canadian city. The latter ROE study was conducted in public elementary schools on the Isle of Man. The Isle of Man is located within the British Isles between Ireland and the islands of Great Britain, has a population of approximately 82,000 people, and has approximately 35 primary schools. Nearly half of the primary schools on the Isle of Man were part of the present study. Participants The participants included 58 elementary school teachers: 20 teachers from the Isle of Man and 38 teachers from Canada. The Isle of Man sample included primary grade teachers (Kindergarten to 3rd grade) recruited from 18 schools who were assigned to either host the ROE program (n = 10) or serve as controls by delivering standard academic instruction (n = 10). The Canadian sample included primary grade (Kindergarten to 3rd grade; n = 20) and intermediate grade teachers (4th to 7th grade; n =18) recruited from 16 schools who were randomly assigned to either host the ROE program (n = 19; 53% instructing primary grades and 47% instructing intermediate grades) or serve as controls by delivering standard academic in- 20 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... struction (n = 19; 53% instructing primary grades and 47% instructing intermediate grades). For both samples, informed consent was obtained from participants. Each consent form explained that the purpose of the study was to assess the effectiveness of a classroom-based program aimed at enhancing children's social and emotional understanding. In the current sample, analyses of background characteristics of the intervention and control teachers revealed no significant differences. The teacher participants were primarily female (86.2% program and 89.7% control), and were of Western European descent (82.8% program and 82.8% control). They predominantly instructed primary grades (69% program and 69% control), had 11 or more years of teaching experience (41.4% program and 48.3% control), and held an undergraduate-level degree (86.2% program and 89.7% control). Approximately half of the participants reported participating in SEL-related training and/or professional development (program 55.2% and 48.3% control). As the present analysis sought to examine only the subset of primary grade teachers, analyses of demographics reported by the Isle of Man and Canadian teachers instructing primary grades were also conducted and revealed no significant differences. Procedure For the ROE study on the Isle of Man (quasi-experimental design study), school leaders interested in participating in the study were invited to a meeting, at which time the study was explained and schools were assigned to the ROE program intervention group or the control group. After the groups were assigned, teachers were sent packages containing information about the study and the measures to be completed. It was explained that they could directly contact the research team at any time with questions. For the ROE study in Canada (RCT study), principals demonstrating an interest in implementing the program at their school were contacted and asked to invite their teachers to participate in the study. Participating classrooms were randomly assigned to the ROE group or the control group. After the groups were randomly assigned, the Principal Investigator of the research project delivered a short presentation about the study to each participating classroom and answered teachers' and students' questions. For both evaluations, baseline data collection occurred a few weeks prior to the commencement of the ROE program implementation in late autumn. Post-test data collection occurred a few weeks after the completion of the ROE program implementation in late spring, approximately 8 months after baseline data collection. Teachers completed a series of 21 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 self-report measures at baseline and post-test within a two-week time period. For their participation in the study, teachers received an honorarium (£25 honorarium for the Isle of Man teachers; $150.00 for the Canadian teachers). Program Implementation Beginning in late autumn and ending in late spring, the ROE program was implemented in the intervention-group classrooms over the course of the school year. All 27 structured lessons across the nine themes (i.e. Meeting the Baby, Crying, Caring and Planning for the Baby, Emotions, Safety, Sleep, Communication, Who Am I?, Goodbye, and Good Wishes) were facilitated by a trained and certified ROE instructor. For each of the nine themes, the ROE instructor facilitated three lessons: (1) the pre-fam-ily visit to introduce the students to the theme, (2) the ROE family visit in which the instructor directed the students to observe the family's interactions and the baby's development, and (3) the post-family visit to reflect on the ROE family visit and complete the theme. The intervention group teachers in the current study acted as secondary implementers and, therefore, they did not facilitate the structured ROE lessons. However, the teachers could support the ROE curriculum by integrating ROE extension activities into the academic curriculum. As stated earlier, the intervention group teachers did not receive ROE training, but were provided with resources to aid in the development of ROE extension activities. Measures Participating teachers completed three self-report measures that assessed (a) their background characteristics, (b) their implementation of ROE extension activities (this latter measure was completed by ROE program teachers only), and (c) their emotion beliefs. Assessment of Teachers' Background Characteristics At baseline, teachers completed a self-report measure that asked them to provide information on demographics (gender, race/ethnicity), level of education, grade level currently teaching, and years of teaching experience. For grade level, teachers were categorized into one of two levels: primary grades (K - 3rd) or intermediate grades (4th - 7 th). Assessment of Teachers' Emotion Beliefs At baseline, all participants completed the Teachers' Beliefs about Emotions (TBAE; Hyson and Lee, 1996) questionnaire - one of the only known measures to assess teachers' emotion beliefs. The TBAE is a 23- 22 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... item self-report measure of the beliefs that teachers hold about emotions in the classroom and the role the teacher plays in their students' emotional development. Teachers indicated the extent to which they agreed with each statement on a Likert-type scale ranging from 1 ("strongly disagree") to 5 ("strongly agree"). The rating scale was modified from Hyson and Lee's (1996) 6-point Likert scale in order to align with other teacher report measures used in the current study. The TBAE comprise six sub-scales: (1) Bonds - beliefs about the importance of teacher-student connections (4 items: e.g. "Children need to feel emotionally close to their teachers;" a = .60); (2) Expressiveness - beliefs about teachers' candid expression of emotions around students (4 items: e.g. "Teachers should 'let their feelings out' in the classroom;" a = .54); (3) Instruction/Modeling - beliefs about using direct instruction and demonstration to help illustrate to students appropriate emotion expression (4 items: e.g. "When a child is angry because another child won't share a toy, I often tell the child exactly what words she could use to express her feelings;" a = .66); (4) Talk/Label - beliefs about helping children identify and discuss their current emotion states (6 items: e.g. "When one of my children is upset about something, I usually try to put into words how he or she is feeling;" a = .45); (5) Protect - beliefs about shielding students from upsetting emotions (3 items: e.g. "Teachers should not read children stories that might make them sad or worried;" a = .55); and (6) Display/Control - beliefs about students' ability to regulate and exhibit emotions in a socially acceptable manner (3 items: "As a teacher, it's important for me to teach children socially acceptable ways of expressing their feelings;" a = .79). Cronbach's alphas for the TBAE in the current study were low to moderate, ranging from .45 (Talk/Label) to .79 (Display/Control). A decision was made to exclude subscales with alphas falling below .50 due to their low internal consistency; hence the Talk/Label subscale was discarded from further analyses. The five remaining subscales of the TBAE were retained for further analyses. Although an alpha level of .70 is customarily considered acceptable for research purposes, Ransford et al. (2009) suggest that .60 is acceptable when research is exploratory in nature, such as in the present study. Therefore, subscales with alphas close to or above .60 were retained as acceptable - two subscales with alphas slightly below .60 (i.e. .54 and .55) and three subscales at or above .60 (i.e. .60, .66, and .79). It should also be noted that the alphas found in the present study were higher than those found by Hyson and Lee (1996) in their research on the development of the TBAE. Hyson and Lee's alphas ranged from .41 (Protect sub-scale) to .62 (Bonds subscale). 23 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Assessment of Teachers' Implementation of ROE Extension Activities At post-test, only ROE program teachers completed a measure assessing the extent to which they delivered ROE extension activities across the general academic curriculum. The dimensions assessed included: (a) the number of subject areas in which the teacher implemented the ROE extension activities (Language Arts, Math, Science, Social Studies, Art, and Other), and (b) the overall frequency with which the teacher implemented the ROE extension activities across all subject areas. More specifically, teachers were provided with a list of curricular subject areas and were asked to indicate with a "yes" or "no" whether or not they had implemented ROE extension activities in each subject area. If the teacher participants reported "yes", they were prompted to indicate the frequency with which they delivered extension activities in that subject area. In the Isle of Man sample, teacher participants were provided a field to write in the frequency of implementation. In the Canadian sample, teacher participants were asked to indicate the frequency with which they implemented the extension activities on a 5-point Likert-type scale (Never, Once or twice, Monthly, Weekly, Daily). To merge the data sets, the qualitative data from the Isle of Man sample were re-coded to match the Canadian study's quantitative response options (e.g. "every day" was re-coded to the numeric value assigned to the Daily option). Results Preliminary Analyses Preliminary analyses were conducted to examine emotion beliefs in the Isle of Man and Canadian teachers to determine whether the two samples could be merged for further analyses (see Table 1). The results indicated that the means for emotion beliefs were comparable between the two groups (i.e. scores falling within one standard deviation range of each other) for four of the five subscales of the TBAE: Bonds, Expressiveness, Instruction/Modeling, and Protect beliefs. For the Display/Control beliefs, however, there was no overlap between the groups' scores. On average, the Canadian primary grade teachers reported stronger agreement that their students were developmentally ready to be taught how to express their feelings in socially acceptable ways than the Isle of Man teachers. The difference in the means for this particular dimension may be related to the cultural context in which these two teacher groups are instructing. Nevertheless, given the demographic and emotion belief similarities between these teacher groups overall, as well as the interest in increasing the statistical power of this study's analyses by having a larger sample size, the data for the two primary grade teacher groups were combined. 24 J. L. HANSON-PETERSON, K. A. SCHONERT-REICHL, V. SMITH ■ TEACHERS' BELIEFS ... Table 1. Means and Standard Deviations of Emotion Beliefs by Study and Grade Level Taught (N = 58) Isle ofMan, Grade K-3 Canadian, Grade K-3 Canadian, Grade 4-7 Emotion Beliefs Teachers (n = 20) Teachers (n = 20+) Teachers (n = 18) M SD M SD M SD Bonds 3-45 -59 3-91 -57 3-68 -56 Expressiveness 3-35 -59 3-54 -45 3-46 -53 Instruction /Modeling 3.40 .72 398 -69 3-95 72 Protect 2.28 -54 2-00 -35 1-76 -39 Display/ Control 3-97 -42 4-68 -46 4-70 -44 Note. + n = 19 for the Protect and Display/Control beliefs, due to missing data. Analytic Strategy To examine the first research question regarding the extent to which grade level taught and years of teaching experience are associated with teachers' emotion beliefs, the combined pre-test data from teachers in both the control and intervention groups were analyzed. Specifically, a series of 2 (grade level taught: primary grades, intermediate grades) x 3 (years of teaching experience: 0-5 years, 6-10 years, 11+ years) analyses of variance (ANOVAs) were conducted - one for each subscale on the TBAE. Huberty and Morris (1989) posit that when multiple outcome variables are of interest, some statisticians suggest conducting a multivariate analysis of variance (MANOVA) prior to performing multiple ANOVAs to help decrease the probability of committing a Type I error. However, Huberty and Morris argued that the results of a MANOVA, versus those of multiple ANOVAs, answer different research questions and thus suggested using both or only the latter depending on the purpose of the research. If the researcher is not "seeking any linear composite of the outcome variables" and "an underlying construct is of no concern" (ibid. p. 303), then performing multiple ANOVAs alone is viewed as appropriate. Therefore, it was deemed acceptable for two reasons to conduct a series of ANOVAs for the present study instead of one MANOVA. Firstly, Hyson and Lee (1996) perceived each emotion belief dimension to be conceptually independent from the others. Secondly, as previously noted, the present study aimed to be descriptive in nature due to the dearth of extant research on teachers' emotion beliefs. The steps for performing ANOVAs described by Pallant (2007) were followed, in which a dependent variable (i.e. composite score on a TBAE 25 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 subscale) and the fixed factors (i.e. teacher background characteristic variables) were entered into the model to examine main and interaction effects. Moreover, when a statistically significant difference was found with a fixed factor with more than two levels, post-hoc Tukey's HSD tests were conducted to identify which group means significantly differed from each other. Finally, to examine effect sizes, eta squared was calculated using Brown's (2008) formula: SSeffect/SST. The results were interpreted according Cohen's (1988) effect size index: small effect size = .01; medium effect size = .06; and large effect size = .14. To examine the second research question regarding the relation of teachers' emotion beliefs related to both the amount of subject areas in which they implement ROE extension activities and the frequency of their implementation across all subject areas, the data of the intervention group teachers were examined. Specifically, correlational analyses were conducted. In these analyses, the teacher participant scores on the TBAE subscales were assessed in relation to the number of subject areas in which the teacher implemented the ROE extension activities and the overall frequency with which the teacher implemented the ROE extension activities across all subject areas. The effect sizes were examined for the significant correlations by squaring the correlation coefficients (Hoyt, Leierer and Millington, 2006). The effect sizes of the correlational analyses were also interpreted according to the aforementioned effect size index (Cohen, 1988). Descriptive Analyses Table 2. Means, Standard Deviations, Minimum and Maximum Values of Teachers' Emotion Beliefs (N = 58) Emotion beliefs M SD Min Max Bonds 3.68 .60 2.25 5.00 Expressiveness 3-45 -53 2.25 5.00 Instruction/Modeling 377 -75 2.33 5.00 Protect 2.02 .48 1.00 3.33 Display/Control 4-44 .56 3.00 5.00 Note. Ns range from 57 to 58 due to missing emotion beliefs data. The response scale for each emotion belief item ranged from 1 "Strongly disagree" to 5 "Strongly agree." Each teacher's composite score for each sub-scale was the average of their responses to the items comprising the given subscale. Table 2 presents the profiles of the teacher participants' emotion beliefs, regardless of intervention or control group. This includes the means, 26 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... standard deviations, and ranges of the emotion beliefs of all teacher participants. As can be seen, on average the teachers reported moderately agreeing with the Expressiveness beliefs and agreeing with the Display/ Control beliefs. They also reported disagreeing with the Protect beliefs. Moreover, teachers reported moderately agreeing with both the Bonds beliefs and Instruction/Modeling beliefs; although, there was more variability in teachers' responses for these two belief areas compared to the others. Relations Between Teachers' Background Characteristics and Emotion Beliefs ANOVAs were performed to examine the relationship between teachers' background characteristics and emotion beliefs. More specifically, differences in each of the emotion beliefs (Bonds, Expressiveness, Instruction/Modeling, Protect, Display/Control) were assessed in relation to their background characteristics using a series of two-way 2 (grade level taught) x 3 (years of teaching experience) ANOVAs - one for each emotion belief. Table 3 provides a summary of the analyses for the relationship between teachers' grade level taught and emotion beliefs. Table 4 provides a summary of the analyses for the relationship between teachers' years of experience and emotion beliefs. The results of these analyses indicated that there were no interaction effects. Several significant main effects were found and are discussed below. No main effects were found for Bonds or Expressiveness beliefs in relation to either background characteristic. As can be seen in Table 3, no main effect emerged for Instruction/ Modeling beliefs and grade level taught. However, as illustrated in Table 4, Instruction/Modeling beliefs (i.e. beliefs in using direct instruction and demonstration to help illustrate to students appropriate emotion expression) were significantly higher for experienced teachers than novice teachers. Post-hoc Tukey's HSD test revealed teachers with either 6-10 years (M = 4.14, SD = .58) or 11+ years (M = 3.89, SD = .75) of teaching experience had higher beliefs on this dimension than novice teachers with 0-5 years of experience (M = 3.31, SD = .69). Regarding the effect size, 1% of the between-subject variance in Instruction/Modeling beliefs was explained by years of teaching experience, which is a small effect. Table 3 indicates that Protect beliefs (i.e. beliefs in shielding students from strong emotions) were significantly higher for primary grade teachers (M = 2.15, SD = .48) than intermediate grade teachers (M = 1.76, SD = .39). The effect size was 1%, which is a small effect. Main effects did not emerge for Protect beliefs and years of teaching experience, as seen in Table 4. 27 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Table 3 also reveals that Display/Control beliefs (i.e. beliefs in students' abilities to regulate and exhibit emotions in a socially acceptable manner) were significantly higher for intermediate grade teachers (M = 4.73, SD = .44) than primary grade teachers (M = 4.29, SD = .57). The effect size was 0%, which is a non-significant effect. Main effects did not emerge for Display/Control beliefs and years of teaching experience, as seen in Table 4. Table 3. Results ofANOVAs for Emotion Beliefs by Grade Level Taught (N= 58) Emotion Beliefs Group Primary Intermediate SS df MS F 1. Bonds 3.69 (.62) 3.68(.56) .01 .00 .00 .95 .00 2. Expressiveness 3-42(-53) 3.47(.53) .04 .04 .14 .72 .00 3. Inst./Modeling 37o(76) 3.87(72) .33 .33 .68 .42 .00 4. Protect 2.15O48) i.76(.39) 1.47 1.47 7.22* .01 .01 5. Display/Control 4.29(.57) 4.73C44) 2.23 2.23 8.04* .01 .00 Note. Ns range from 57 to 58 due to missing emotion belief data. Standard deviations appear in parentheses. For grade level taught, primary grades = grades K-3, intermediate grades = grades 4-7. Inst./Modeling = Instruction/Modeling. p < .°5. Table 4. Results ofANOVAs for Emotion Beliefs by Years ofExperi-ence (N = 58) Group Emotion Beliefs 0-5 years 6-10 years 11+years SS df MS F P n2 1. Bonds 3.58C65) 3.76(.52) 37°(.6i) .23 2 .11 .30 .74 .00 2. Expressiveness 3.43(.47) 3.48(.62) 3.43C5O .03 2 .02 .05 .95 .00 3. Inst./Modeling 3.3i(.69) 4.14C58) 3.89(75) 4.92 2 2.46 5.00* .01 .01 4. Protect i.99(.28) 2.0i(.5i) i.93(.48) .06 2 .°3 .15 .86 .00 5. Display/Control 4.67C54) 4.48O65) 4.39C5O .58 2 .29 1.04 .36 .00 Note. Ns range from 57 to 58 due to missing emotion belief data. Standard deviations appear in parentheses. p < .°5. Correlational Analyses To examine teachers' emotion beliefs in relation to the amount of academic subject areas in which they implemented ROE program activities and the frequency of their implementation of ROE extension activities across all subject areas, correlational analyses were conducted. As can be P n 28 J. L. HANSON-PETERSON, K. A. SCHONERT-REICHL, V. SMITH ■ TEACHERS' BELIEFS seen in Table 5, particular emotion beliefs' dimensions were significantly and positively related to the implementation of ROE extension activities. Teachers' Expressiveness beliefs (i.e. beliefs in teachers' candid expression of emotions around their students) were significantly and positively correlated with the overall frequency with which they implemented the ROE extension activities across the academic subject areas. The variance-ac-counted-for effect size was 23%, a moderate effect. Teachers' Protect beliefs significantly and negatively correlated with both the number of subject areas in and frequency with which ROE extension activities were implemented. The effect sizes were 27% and 20%, respectively, which are fairly moderate effects. Finally, teachers' Display/ Control beliefs were significantly and positively correlated with both the number of subject areas in and frequency with which ROE extension activities were implemented. Respectively, the effect sizes were 41%, a moderate to strong effect, and 21%, a moderate effect. Table5. Correlations and Descriptive Statistics for Emotion Beliefs and Implementation of ROE Extension Activities (N = 29) Variables i 2 3 4 5 6 7 Emotion beliefs: (1) Bonds - -35** .°9 -.°9 .°9 .°3 .2° (2) Expressiveness -- -.07 .04 .37** .3° .48* (3) Instruction/Modeling .°3 J9 -.°3 -.25 (4) Protect -.44** -.52** -.45* (5) Display/Control .64** .46* Extension activity implementation: (6) Number ofsubject areas .91** (7) Frequency -- M 3.68 3.44 3.72 2.00 4.49 2.65 5.32 SD -57 .62 .73 .45 .57 2.3° 5.47 Note. Ns range from 25 to 29 due to missing data. For the number of subject areas in which extension activities were implemented, there were six subject areas in total: Language Arts, Math, Science, Social Studies, Art, and Other. For Frequency of implementation across all subject areas, the scores for all six subject areas (0 = Never, 1 = Once or twice, 2 = Monthly, 3 = Weekly, 4 = Daily) were totalled. p < .05. p < .01. Discussion The current literature on the emotion beliefs of important adults in the lives of children, who influence their social and emotional development, focuses almost solely on parents and on early childhood teachers instruct- 29 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 ing in preschool settings (see Bellas, 2009; Hyson and Lee, 1996; Jaramil-lo, 2006; Jumper, 2005). Additionally, most research studies concerned with the implementation of evidence-based preventive intervention programs "focus on identifying concurrent factors that explain variation in fidelity during the implementation phase as opposed to examining factors that are in place before an intervention is selected or that emerge early on when an intervention starts" (Wanless and Domitrovich, 2015, p. 1038). The current study was designed to address these gaps in the research. This study is unique as it is one of the first to investigate the emotion beliefs of elementary school teachers, and is the only known study to include in its sample intermediate grade teachers who instruct students in middle childhood (Grades 4-7). Additionally, this study is unique as it is one of the first to investigate teacher-related factors before program implementation, particularly emotion beliefs, with the findings indicating that these factors are related to program implementation. These findings add to the emerging literature on 'indicators of readiness' to implement SEL programs (ibid.) and further open the black box of SEL program implementation to understand the role of the teacher in its implementation (Jennings and Greenberg, 2009). Understanding the factors that impact teachers' readiness to implement SEL programs is beneficial as it can guide the knowledge, training, and support offered by SEL program developers, coaches, school leaders, and even pre-service teaching programs; this can help enhance the social-emotional competence of teachers, ensure their needs are met, and that they have the capacity to implement the evidence-based program with fidelity and thus increase the likelihood that positive student outcomes will be achieved (Wanless and Domitrovich, 2015; Domitrovich et al. 2015; Jennings, 2014). Teachers' Background Characteristics and Emotion Beliefs The results of the current study indicate that teachers' background characteristics are significantly related to particular emotion beliefs dimensions, indicating the need to address these issues in pre-service and/or in-service teacher training. The finding that experienced teachers had higher Instruction/Modeling beliefs compared to novice teachers may indicate that novice teachers felt uncertain or less prepared to take responsibility for showing their students how to express their emotions appropriately; whereas experienced teachers may have acquired strategies through first-hand experience that enhanced their sense of efficacy to explicitly guide their students in this way (see Tschannen-Moran and Woolfolk Hoy, 2007). Primary grade teachers had higher Protect beliefs and lower Display/ Control beliefs compared to intermediate grade teachers. These beliefs for 30 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... both groups may be developmentally appropriate - that primary grade children are not developmentally ready to be exposed to stories or circumstances that could upset them and are unable to control the way they express their emotions, whereas intermediate-aged children tend to cope on their own by using problem-solving strategies when upset (Saarni, 2011). The lower Display/Control beliefs also indicate that primary grade teachers are less likely to believe it is their role to help children in their classrooms learn to control their emotions in socially appropriate ways compared to intermediate grade teachers. Regardless of the possibility that primary grade children are not developmentally ready to regulate their emotions on their own, they would still benefit from receiving support from their teachers to develop their emotion regulation skills. Together these findings suggest the importance of providing teachers, particularly novice and primary grade teachers, with knowledge about social-emotional development and the important role that teachers play in socializing this development in their students. Additionally, the findings suggest that if these teachers are expected to implement an emotion-focused SEL program, it may be beneficial to provide them with ongoing SEL coaching to build their self-efficacy and motivation to engage their students in the emotion socialization practices called for by the program (Tschannen-Moran and Woolfolk Hoy, 2007). It should be reiterated that the effect sizes were minimal. Nonetheless, Trusty, Thompson and Petrocelli (2004) have cautioned researchers not to interpret minimal effect sizes as an indication that the significant relationships among the variables are not important. They noted that whether the findings are comparable to the findings of other similar studies may be of greater importance (ibid.). At this stage, however, minimal research has been conducted using an elementary school sample to explore similar links. Thus, the implications of the minimal effect sizes for the present study are indeterminate. Teachers' Emotion Beliefs and Implementation of ROE Extension Activities The findings indicating that differences in teachers' implementation dosage of an emotion-focused SEL program are associated with their emotion beliefs support the notion that SEL program implementation does not occur in a vacuum. Rather, a variety of factors present before the implementation of a program begins, such as emotion beliefs, can influence the implementer's readiness to implement the program (Wanless and Domi-trovich, 2015), their perceptions of the importance of the program, and/or 31 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 their motivation and sense of efficacy to implement the program activities (see Durlak and DuPre, 2008). A central finding of the present study was that teachers with lower Expressiveness beliefs implemented ROE extension activities less frequently, compared to those with higher beliefs on this dimension. Similar to Jaramillo's (2006) study, these findings suggest that teachers with higher Expressiveness beliefs may be more comfortable, willing, and/or motivated to engage in activities that can involve communicating their own emotions to their students. Another key finding of the current study was that teachers with higher Protect beliefs or lower Display/Control beliefs implemented ROE program activities in fewer subject areas and with less frequency. These findings may indicate that teachers' perceptions of their students' developmental readiness to cope with and learn to regulate strong emotions are associated with teachers' motivation and willingness to engage their students in activities that can provoke strong emotions in their students. That is, despite the fact that the ROE program is tailored to the developmental level of the participating students, these teachers may not want to take responsibility or do not feel well-equipped to provide emotional support and guidance to students who become upset during the emotion-laden activities. Such reasoning builds on the research literature on implementer-related factors that influence the readiness to implement and thus the implementation fidelity of preventive intervention programs (e.g. Durlak and DuPre, 2008; Wanless and Domitrovich, 2015). Limitations and Implications A few limitations of the current study must be acknowledged to consider their impact on this research and provide direction for future research in this area. Although the findings of the current study add to the knowledge base on the links among elementary school teachers' background characteristics, emotion beliefs, and SEL program implementation, the lack of prior research in this area meant there was little empirical direction in which to forecast our results. Furthermore, the study's sample size was relatively small, hence limiting the statistical power to discern effects. For instance, this may explain the lack of interaction effects for the ANOVAs and the few sizeable, almost-significant results (±.20 and above) found for the correlational analyses (e.g. a strong positive correlation between Bonds beliefs and implementation frequency; a strong positive correlation between Expressiveness beliefs and implementation in total subject areas). Moreover, the sample was rather homogenous regarding gender and ethnicity, with the teachers being predominantly female and of Western Eu- 32 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... ropean descent. On the other hand, teachers of Western countries typically fit these demographics. As previously noted, the data for the Grades K-3 Canadian and Isle of Man teachers were merged to form one group. Although the current study ensured that there were no statistically significant differences concerning the demographics between these teacher groups, cultural and contextual differences were not controlled for. Owing to the recent emphasis the British Columbia Ministry of Education has placed on social responsibility (e.g. ethical and democratic behavior, peaceful problem-solving) in its schools (British Columbia Ministry of Education, 2001), the Canadian teachers, compared to the Isle of Man teachers, could have been more familiar and comfortable with fostering the student outcomes that the ROE program aims to promote. Merging the data may have tempered some of the findings related to the primary grade teachers. It is suggested that future research consider cultural and contextual differences in relation to emotion beliefs. An additional limitation is the internal consistencies of the subscales of the TBAE (Hyson and Lee, 1996) questionnaire, which were weak to moderate. It is suggested that future research perform a content validation of the questionnaire to assess whether all items should be retained. For instance, the framing of the items is not consistent throughout the questionnaire, as some items ask teacher participants to reflect on their own emotion socialization practices (e.g. "When a child is angry because another child won't share a toy, I often tell the child exactly what words she could use to express her feelings" [italics added]), whereas other items are concerned with their perceptions of social norms (e.g. "Teachers should avoid showing children how to express their feelings" [italics added]). Another limitation is the reliance on self-report data. Specifically, we relied solely on teachers' reports of their implementation of SEL program activities. As we did not obtain reports from other informants or observational data, we cannot be certain about the degree to which teachers' reports are accurate representations of what they did in practice. As noted by Durlak and DuPre (2008), "[t]here are some indications that observational data are more likely to be linked to outcomes than self-report data ... but few studies have directly compared these two strategies" (p. 331). Therefore, our findings on SEL program implementation fidelity should be interpreted with some caution. We suggest future research on SEL program implementation fidelity include multiple informants and/or observations in addition to self-reports. As noted, the teacher participants in the ROE intervention were secondary implementers of the program. Although this did not interfere 33 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 with our efforts to assess teachers' emotion beliefs in relation to their implementation ofROE activities, it is suggested that future research include in its sample teacher participants who are primary implementers. This can be valuable for comparing primary versus secondary implementers and to consider the benefit of receiving training directly related to the SEL program being implemented. The findings of the current study revealed that particular emotion beliefs held by teachers were associated with their implementation dosage of the ROE program. Due to the fact that the research has indicated that SEL program implementation fidelity is linked to student SEL outcomes (Durlak and DuPre, 2008), it would be valuable for future researchers to examine whether teachers' emotion beliefs are positively correlated with students' emotional development in a classroom hosting an emotion-focused SEL program. In conclusion, this study has the potential to inform SEL program design, teacher training, and future SEL research. The findings can inform the design of SEL programs as they provide insight into what teachers might think about the suitability of the programs in relation to (a) their current belief systems, competencies, and methods of operating emotionally in their classrooms, and (b) their students' current level of development and whether they are ready for emotion-laden activities, even if the program is said to be tailored to the particular age group of the participating students. These findings may also inform the training with which teachers are provided, whether in pre-service teacher training or in-service professional development, to potentially guide their emotion beliefs. This might include providing teachers with information about the emotional development of children and about teachers' role in socializing their students' emotional development, and encouraging engagement in meaningful experiences (e.g. a student teaching practicum that involves observing experienced teachers implement SEL programs, on-going support from a trained SEL coach) that can enhance their sense of self-efficacy and motivation to implement SEL program activities (see Larsen and Samdal, 2012). The findings of this study also support and extend an area of research in the SEL field that is receiving increased attention: the social-emotional competence (SEC) of teachers (Brown et al. 2010; Jennings and Green-berg, 2009). In fact, teachers' emotion beliefs may provide insight into particular dimensions of their social-emotional competence (e.g. emotion understanding, emotion regulation skills). Therefore, this research may support the importance of promoting the SEC of teachers, such as through coursework or workshops that help teachers express and man- 34 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... age their emotions properly. This study may also act as a platform for future research investigating whether teachers' emotion beliefs can be guided and the most effective methods for guiding these beliefs. References Ahn, H. J. (2005) Child care teachers' beliefs and practices regarding socialization of emotion in young children. Journal of Early Childhood Teacher Education 26 (3), pp. 283-295. Ahn, H. J., Stifter, C. (2006) Child care teachers' response to children's emotional expression. Early Education and Development 17 (2), pp. 253-270. Beets, M. W., Flay, B. R., Vuchinich, S., Acock, A. C., Li, K. K., Allred, C. (2008) School climate and teachers' beliefs and attitudes associated with implementation of the positive action program: A diffusion of innovations model. Prevention Science9 (4), pp. 264-275. Bellas, V. M. (2009) Emotion in the classroom: A theory-based exploration of teachers' emotion socialization beliefs and behaviors. PhD thesis, Clark University. [Online] ProQuest Dissertations and Theses Da-tabase.http://www.proquest.com (12.6.2012). Brackett, M. A., Reyes, M. R., Rivers, S. E., Elbertson, N. A., Salovey, P. (2012) Assessing teachers' beliefs about social and emotional learning. Journal of Psychoeducational Assessment30 (3), pp. 219 -236. Brackett, M. A., Rivers S. E. (2008) What is emotional literacy? In Brackett, M. A., Kremenitzer, J. P. with Maurer, M., Carpenter, M., Rivers, S.E., Katulak, N. (eds.). Emotional literacy in the classroom: Upper elementary. Portchester, NY: National Professional Resources, Inc. British Columbia Ministry of Education (October, 2001) BC performance standards. Social responsibility: A framework. Victoria, British Columbia: Ministry of Education, Student Assessment and Program Evaluation Branch. http://www.bced.gov.bc.ca/perf_stands/sintro. pdf (10.6.2012). Brown, J. D. (2008) Statistics corner: Questions and answers about language testing statistics: Effect size and eta squared. Shiken: JALT Testing Evaluation SIG Newsletter 12 (2), pp. 38-43. http://jalt.org/ test/PDF/Brown28.pdf (23.4.2016). Brown, J. L., Jones, S. M., LaRusso, M. D., Aber, J. L. (2010) Improving classroom quality: Teacher influences and experimental impacts of the 4Rs program. Journal of Educational Psychology 102 (1), pp. 153167. Cohen, J. (1988) Statistical power analysis for the behavioral sciences (2nd ed.). Hillside, NJ: Lawrence Erlbaum Associates, Publishers. 35 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Denham, S. A., Bassett, H. H., Zinsser, K. (2012) Early childhood teachers as socializers of young children's emotional competence. Early Childhood Education Journal 40 (3), pp. 137-143. Domitrovich, C. E., Pas, E. T., Bradshaw, C. P., Becker, K. D., Keperling, J. P., Embry, D. D., Ialongo, N. (2015) Individual and school organizational factors that influence implementation of the PAX Good Behavior Game intervention. Prevention Science 16, pp. 1064-1074. Durlak, J. A. (2015) What everyone should know about implementation. In Durlak, J. A., Domitrovich, C. E., Weissberg, R. P., Gullotta, T. P. (eds.). Handbook of social and emotional learning: Research and practice. New York: Guilford, pp. 395-405. Durlak, J. A., DuPre, E. P. (2008) Implementation matters: A review of research of the influence of Implementation on program outcomes and the factors affecting implementation. American Journal of Community Psychology 41 (1), pp. 327-350. Durlak, J. A., Weissberg, R. P., Dymnicki, A. B., Taylor, R. D., Schellinger, K. B. (2011) The impact of enhancing students' social and emotional learning: A meta-analysis of school-based universal interventions. Child Development 82 (1), pp. 405-432. Gordon, M. (2000) The Roots of Empathy: Curriculum manual. Toronto, Canada: Roots of Empathy. Gosney, S. L. (2006) Child care providers' beliefs about and reactions to children's negative emotions. Master's thesis, Oklahoma State University. [Online] ProQuest Dissertations and Theses Database. http://www. proquest.com (10.6.2012). Hargreaves, A. (2000) Mixed emotions: Teachers' perceptions of their interactions with students. Teaching and Teacher Education 16 (8), pp. 811-826. Horner, C. G., Wallace, T. L. (2013) Measuring emotion socialization in schools. Journal of School Health 83 (10), pp. 697-703. Hoyt, W. T., Leierer, S., Millington, M. J. (2006) Analysis and interpretation of findings using multiple regression techniques. Rehabilitation Counseling Bulletin 49 (4), pp. 223-233. Huberty, C. J., Morris, J. D. (1989) Multivariate analysis versus multiple univariate analyses. Psychological Bulletin 105 (2), pp. 302-308. Huemer, K. M. (2010) Infant/toddler teachers' emotion beliefs and children's emotional Competence. Master's thesis, Sacramento State University. [Online] Sacramento State Scholarworks Database. http://csus-dspace.calstate.edu/handle/10211.9/1016 (23.4.2016). 36 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... Hyson, M. C., Lee, K. (1996) Assessing early childhood teachers' beliefs about emotions: Content, contexts, and implications for practice. Early Education and Development 7 (1), pp. 59-78. Institute of Medicine (2009) Preventing mental, emotional, and behavioral disorders among young people: Progress and possibilities. The National Academies Press, Washington, DC. Jaramillo, A. M. (2006) Early childhood educators' experiences implementing a social competence promotion program for preschool-aged children: The "Safe Spaces"program. Master's thesis, University of British Columbia. [Online] UBC Theses and Dissertations Database. https://open. library.ubc.ca/cIRcle/collections/ubctheses/83i/items/i.0 054469 (23.4.2016). Jennings, P. A. (2014) Early childhood teachers' well-being, mindfulness, and self-compassion in relation to classroom quality and attitudes towards challenging students. Mindfulness. http://link.springer.com/ article/i0.i007%2Fsi267i-0i4-03i2-4 (25.6.2016). Jennings, P. A., Greenberg, M. T. (2009) The prosocial classroom: Teacher social and emotional competence in relation to student and classroom outcomes. Review of Educational Research 79, pp. 49^525. Jumper, R. L. (2005) Teacher reports ofchildren's emotion regulation: Is there concordance with trained observers? Master's thesis, Auburn University. [Online] Auburn University Theses and Dissertation Database. https://etd.auburn.edu/bitstream/handle/i04i5/752/JUMPER_ RACHEL_2i.pdf?sequence=i (23.4.20^). Klem, A. M., Connell, J. P. (2004) Relationships matter: Linking teacher support to student engagement and achievement. Journal of School Health 74 (7), pp. 262-273. Kress, J. S., Elias, M. J. (2006) School-based social and emotional learning programs. In Renninger, K. A., Sigel, I. E. (Series eds.). Handbook of Child Psychology: Vol. 4. Child psychology in practice. Hoboken, NJ: John Wiley, pp. 592-6i8. Larsen, T., Samdal, O. (20i2) The importance of teachers' feelings of self efficacy in developing their pupils' social and emotional learning: A Norwegian study of teachers' reactions to the Second Step program. School Psychology International 33 (6), pp. 63^645. Murray, C., Greenberg, M. T. (2000) Children's relationship with teachers and bonds with school: An investigation of patterns and correlates in middle childhood. Journal of School Psychology 38 (5), pp. 423-445. Pajares, M. F. (i992) Teachers' beliefs and educational research: Cleaning up a messy construct. Review of Educational Research 62 (3), pp. 307332. 37 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Pallant, J. (2007) SPSS survival manual (3rd ed.). New York, NY: McGraw-Hill. Ransford, C. R., Greenberg, M. T., Domitrovich, C. E., Small, M., Jacob-son, L. (2009) The role of teachers' psychological experiences and perceptions of curriculum supports on the implementation of a social and emotional learning curriculum. School Psychology Review 38 (4), pp. 510-532. Reyes, M. R., Brackett, M. A., Rivers, S. E., Elbertson, N. A., Salovey, P. (2012) The interaction effects of program training, dosage, and implementation quality on targeted student outcomes for the RULER approach to social and emotional learning. School Psychology Review 41 (1), pp. 82-99. Rimm-Kaufman, S. E., Sawyer, B. E. (2004) Primary-grade teachers' self-efficacy beliefs, attitudes toward teaching, and discipline and teaching practice priorities in relation to the Responsive Classroom approach. Elementary School Journal 104 (4), pp. 321-341. Roots of Empathy. (2013) What we do: Program pillars. http://www.root-sofempathy.org/en/what-we-do/program-pillars.html (11.10.2013) Saarni, C. (2011) Emotional development in childhood. In Tremblay, R. E., Barr, R. G., Peters, R. DeV., Boivin, M. (eds.). Encyclopedia on early childhood development: Emotions. Montreal, Quebec: Centre of Excellence for Early Childhood Development and Strategic Knowledge Cluster on Early Child Development, pp. 1-7. Santos, R. G., Chartier, M. J., Whalen, J. C., Chateau, D., Boyd, L. (2011) Effectiveness of school-based violence prevention for children and youth: Cluster randomized controlled field trial of the Roots of Empathy program with replication and three-year follow up. Healthcare Quarterly 14 (2), pp. 80-91. Schonert-Reichl, K. A., Scott, F. (2009) Effectiveness of the "Roots of Empathy" program in promoting children's emotional and social competence: A summary of research findings. In Gordon, M. (ed.). Roots of empathy: Changing the world child by child. Toronto, Ontario: Thomas Allen Publishers, pp. 239-262. Schonert-Reichl, K. A., Smith, V., Zaidman-Zait, A., Hertzman, C. (2012) Promoting children's prosocial behaviors in school: Impact of the "Roots of Empathy" program on the social and emotional competence of school-aged children. School Mental Health 4 (1), pp. 1-21. Schonert-Reichl, K. A., Weissberg, R. P. (2014) Social and emotional learning: Children. In Gullotta, T. P., Bloom, M. (eds.). Encyclopedia of Primary Prevention and Health Promotion Part II. (2nd ed.), New York: Springer. 38 j. l. hanson-peterson, k. a. schonert-reichl, v. smith ■ teachers' beliefs ... Swearer, S. M., Espelage, D. L., Vaillancourt, T., Hymel, S. (2010) What can be done about school bullying? Linking research to educational practice. Educational Researcher 39 (1), pp. 38-47. Tschannen-Moran, M., Woolfolk Hoy, A. (2007) The differential antecedents of self-efficacy beliefs of novice and experienced teachers. Teaching and Teacher Education 23 (6), pp. 944-956. Trusty, J., Thompson, B., Petrocelli, J. V. (2004) Practical guide for reporting effect size in the Journal of Counseling and Development. Journal of Counseling and Development 82 (1), pp. 107-110. Wanless, S. B., Domitrovich, C. E. (2015) Readiness to implement school-based social-emotional learning interventions: Using research on factors related to implementation to maximize quality. Prevention Science 16, pp. 1037-1043. Whitebook, M. (2014) Building a skilled teacher workforce. Berkeley, CA: Center for the Study of Child Care Employment, University of California, Berkeley. http://www.irle.berkeley.edu/cscce/wp-content/ uploads/20i4/0 9/Building-a-Skilled-Teacher-Workforce_Septem-ber-2014_9-25.pdf (10.2.2016). Wilcox-Herzog, A., Ward, S. (2004) Measuring teachers' perceived interactions with children: A tool for assessing beliefs and intentions. Early Childhood Research and Practice 6 (2). http://ecrp.uiuc.edu/ v6n2/herzog.html (23.4.2016). Zinsser, K. M., Shewark, E. A., Denham, S. A., Curby, T. W. (2014) A mixed-method examination of preschool teacher beliefs about social-emotional learning and relations to observed emotional support. Infant and Child Development 23 (5), 471-493. 39 The Development of Teacher's Relational Competence Scale: Structural Validity and Reliability Maša Vidmar and Katja Kerman Social and emotional competences have often been an umbrella term for a wide range of competences, from emotional intelligence, interpersonal skills to cognitive regulation (Jones, Bouffard, & Weissbourd, 2013). Collaborative for Academic, Social, and Emotional Learning (CASEL, 2013) outlines five dimensions of social and emotional learning in students that have also been applied to teachers (Schonert-Reichl, Hanson-Peterson, & Hymel, 2015): self-awareness, self-regulation, social awareness, relationship skills and responsible decision-making. In the recent years, it has been suggested that teachers' social and emotional competence (SEC) are vital not only for the development of social and emotional competences in students (Schonert-Re-ichl, Roeser et al., 2015), but also for students' learning and development in general (Jennings & Greenberg, 2009; Jensen, Bengaard Skibsted, & Veds-gaard Christensen, 2015; Jones et al., 2013). Currently, in theoretical discourses, empirical research, teacher education programs, and policy debates little attention is paid to the cultivation and promotion of socio-emotional competences of teachers. Scientific evidence is scarce in this respect, thus much of the theoretical and empirical work is ahead. This study addresses the gap by developing a measure of teacher's SEC, specifically their relational competence, to advance the science and research in the field. What is Relational Competence? The topic of relational competence (also referred to as interpersonal competence) has been mostly examined in the context of relationships with peers and/or romantic partners (e.g. Adamczyk & Pilarska, 2012; Engels, Finkenauer, Meeus, & Dekovic, 2001; Niederberger, 2013; Ngu & Florsheim, 2011), al- 41 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 though some studies have also been done with teachers in education context (e.g. Jensen et al., 2015; Pantic & Wubbels, 2012). Teachers' relational competence can be placed in the general framework of teachers' SEC. Several definitions exist, but establishing/sustaining quality (positive, supportive, encouraging) relationships with students lies in the core. In our work, the relational competence refers to a concept proposed by Juul and Jensen (2010) and is defined as teacher's ability to see a student as a unique being and to consequently adapt their own actions (behaviour) without abandoning the leadership role and their authenticity, as well taking full responsibility for teacher-student relationship. Looking at this definition we can extract specific components of relational competence, such as seeing the student, leadership (of the educational process), teachers' authenticity and teachers' responsibility for the relationship. Authors (Juul & Jensen, 2010) provide an explanation of each of these components; however, they do not provide very explicit definitions or descriptions. In their work they refer to the ability and willingness to take full responsibility for the relationships as pedagogic ethics (stance), whereas other components are referred to as pedagogic skill (craft) (ibid.). According to Juul and Jensen (2010) the teachers' ability to see a student refers to the fact that teachers sees beyond the most obvious apparent behaviour or words of a student (e.g. see worry or fear behind hyperactivity, see non-verbal resistance in student's body even if a student verbally says 'yes') - an adult collects all this information and shapes their own view of a student and is able to give a student full recognition and acknowledgement. As stated by Juul and Jensen (2011), the basis for high-quality relationships is that students/children are understood and treated as individuals - as autonomous people who play an active role in building and maintaining relationships. This means that the teacher does things with the student and not to the student. This also means that a student, with their reactions within this relationship, provides information about who the student is and which parts of relationship make them feel good (or less good). Students' reactions are therefore not interpreted as an expression of what a student is, but who they are in this particular relationship. The key is to observe and identify students' reactions and signals. Leadership implies teachers' ability to plan and realize educational processes without damaging student's sense of personal integrity (i.e. students feel secure and relaxed). It means that a teacher is able to lead (guide) towards the (educational) goal and have this goal all the time in mind, but at the same time being able to see students as individuals and acknowledge them. 42 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... Authenticity refers to teachers' ability and willingness to be personal (i.e. to be present and to share own thoughts, values, boundaries) in the relationship and to develop a subject-subject relationship (i.e. two individuals engaged in educational process), rather than a subject-object relationship (i.e. teachers as the one doing teaching, transferring knowledge to the student). It also refers to the match between professional and personal values (i.e. is the teacher able to act in accordance with their own values and beliefs - about teaching, learning, education etc). The quality of relationship depends on how authentic adults (teachers) are in communication and how included children (students) feel. Responsibility for the quality of the student-teacher relationship includes the ability to establish and maintain the relationship as well as to take an important position in students' life; it is solely on the side of the teacher. The adult has to consider both, his inner reality and the understanding of the child. The concept of teachers' responsibility for the student-teacher relationship refers to the fact that student-teachers relationships are asymmetrical (Pianta, Hamre, & Stuhlam, 2003) and that teachers are responsible for creating contact and the quality (reciprocity, dynamics) of the relationship. Thus when positive, supportive, and accepting relationship with a student or a group of students does not develop, the teacher asks themselves what are they doing that this positive relationship is not being built (and adapts their behaviour accordingly). The teacher holds the responsibility for creating good interactions and a good learning environment, and for engaging in development-supporting relations (Jensen et al., 2015). Teachers are models of how to communicate. Thus, teachers need to know how to form, maintain, improve and strengthen the quality of the relationships: how to work consciously and systematically with the relation as a space for development and learning. All of these components are related with each other. However, we view the two components - namely teacher's ability to see a student and teacher's leadership in educational process in accordance with the 'seen' - so interrelated that we propose to merge them in a single component called respect for individuality. We bring these two components together because teacher's (exclusive) focus on seeing a student as an individual may imply overshadowing of the process of teaching and learning that is to take place in the classroom; when in fact the opposite is true - taking students psychological needs into account does not mean that learning is no longer important, rather that it is a prerequisite for learning. Thus 'seeing' the student is inextricably linked with leadership role in teaching and learning. It is proposed that relational competence is composed of three components (dimensions): respect for individuality, authenticity, and responsibility for the relationship. 43 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 It is important to note that relational competence is not only about communication techniques, but also about the dialogue which is based on the sincere wish and the competence of the adults to react openly and with sensitivity; it is 'an ability to meet students with openness and respect, to show empathy and be able to take responsibility for one's own part of the relation' (Jensen et al., 2015). In a recent work, Jensen and colleagues (2015) proposed the model of relational competence (also using Juul and Jensen's definition), that is composed of five sub-elements: (1) context (relational competence includes the ability to reflect on the influence of the context for interaction and learning); (2) appreciation (respect for other people's worlds of experience - this can be understood as fundamental attitude and also as more specific relational skills, such as listening, understanding, tolerating, confirming); (3) change of perspective (taking other people's viewpoint); (4) empathy (ability to recognize and understand others' feelings); and (5) attention and presence of mind (being present in relation to yourself). The link between this sub-elements and the definition of relational competence remains unclear. Why is Relational Competence Important? The idea of developing teachers' relational competence is based on the cen-trality of relationships in human development, which has been demonstrated in many theories and studies (e.g. attachment theory, Bowlby, 1969; zone ofproximal development, Vygotsky, 1978). This formed a foundation for understanding the teacher-student relationship and led to hypothesis that teacher-student relationship has an important impact on students. Indeed, several longitudinal studies provide evidence that a teacher's report of a supportive relationship with a student has positive effects on elementary students' behavioural and academic adjustment (e.g. Curby, Rimm-Kaufman, & Ponitz, 2009; Hamre & Pianta, 2001; Hughes, Cavell, & Jackson, 1999; Ladd, Birch, & Buhs, 1999; Meehan, Hughes, & Cavell, 2003; O'Connor & McCartney, 2007; Valiente, Lemery-Chalfant, Swan-son, & Reiser, 2008). Similarly, PISA 2012 (OECD, 2013) demonstrated that in all countries and economies, among students with equal performance and similar socio-economic status, those who attend schools with better teacher-student relations reported a stronger sense of belonging to school. Sense of belonging is linked to academic achievement (Anderman 2002; Pittman & Richmond, 2007). Moreover, in a seven-year study of 400 elementary schools, Bryk and Schneider (2004) found that the quality of social relationships among the school community (principals, teach- 44 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... ers, students, and parents) is central to student's functioning, and strongly predicts positive student outcomes. Hattie (2012) listed the teacher-student relationships as highly influential for student achievement in his review. Moreover, teacher-student relations have an exceptionally powerful influence over teachers' job satisfaction (OECD, 2014) and are also related to teachers' sense of efficacy (Yoon, 2002). These studies indicate the importance of relationship skills for teachers - teachers need to know how to develop quality relationships, how to improve and strengthen the relationships, i.e. they need to possess relational, interpersonal competences. Two questions emerge immediately: how can we measure teachers' relational competence and how can relational competence be developed (taught) in teachers? The present paper aims to develop an instrument to measure teachers' relational competence, whereas developing teachers relational competences is in the focus of another article in this issue (Laurs-en & Nielsen, this issue). Measures of Relational Competence Wubbels and colleagues (Wubbels, Brekelmans, den Brook, & Van Tartwijk, 2006; Wubbels et al., 2012) developed a student-report and self-report measure of interpersonal competence (Questionnaire for Teacher Interactions, QTI) based on their theoretical Model for Interpersonal Teacher Behaviour. In their model, teachers' behaviour is described along two independent dimensions: control (dominance-submission) and affiliation (cooperation-opposition). These dimensions define eight types of teacher interpersonal relations: Leadership (e.g. This teacher is sure about what they want in the classroom), Helpful/friendly (e.g. This teacher is friendly), Understanding (e.g. This teacher is willing to explain again if we don't understand), Student responsibility/ freedom (e.g. We can decide some things in this teacher's class), Uncertain (e.g. This teacher does not seem sure), Dissatisfied (e.g. This teacher is bad-tempered), Admonishing (e.g. This teacher gets angry quickly) and Strict (e.g. This teacher is strict) (examples from Kokkinos, Charalambous, & Davazoglou, 2009). The authors (Wubbels et al., 2012) also list the following five teacher competences as comprising interpersonal competence: (1) providing guidance (e.g. is able to make pupils active learners), (2) setting norms and standard (e.g. acts as a role model), (3) correcting undesirable pupil behaviour (e.g. checks whether pupils respond to their corrections), (4) paying attention to pupils (e.g. shows personal interest in pupils), and (5) giving pupils responsibility and freedom (e.g. gives the pupils an appropriate level of responsibility). The model proposed by Wubbels and colleagues (2006, 45 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 2012) resembles the work on teacher styles that are based on Baumrind's parenting dimensions - control and responsiveness (see Wentzel, 2002). Although the focus in Wubbels' (and colleagues, 2006, 2012) and Juul and Jensen's work is similar - teachers' relationship skills - and the terms used imply substantial overlap (relational versus interpersonal competence), a closer look shows that the Juul and Jensen's conceptualization focuses much more on the relationship between teacher and student per se. In this context, correcting students' undesirable behaviour (one of the competences in Wubbels et al., 2012 model) is irrelevant, because students' behaviour is viewed and understood as a signal of who and how the student is within the relationship. We decided to develop a measure of relational competence based on the work ofJuul and Jensen (2010). When constructing the scale we focused on the main three components of relational competence - respect for individuality (i.e. seeing student and leadership), authenticity and responsibility for the relationship (see Method for details on scale development). The Present Study The need for a measure of relational competence is practice-based on one hand (e.g. to demonstrate the need for teachers' professional development on the topic), but also research-based on the other hand (e.g. to evaluate effects of initial or in-service training for teachers, see e.g. Laursen & Nielsen, this issue; to examine links with student and teacher outcomes). Thus, in the present study, we aimed to develop and verify a new measure of relational competence (Teachers' Relational Competence Scale, TRCS) that is grounded in the conceptual work of Juul and Jensen (2010). The overall aim was to develop a self-report questionnaire for teachers and to investigate the reliability and structural validity of the new instrument. The three-factor structure assessing respect for individuality, authenticity and responsibility has been presupposed. The specific aims of the study were to: (1) identify the items that are reflected by the underlying factors in the expected manner; (2) examine whether the presupposed three factors indeed emerge and have satisfactory reliabilities, item loadings and model fit indices; (3) propose suggestions for further development of the instrument (e.g. alternative factor structure, new items). Method Participants Teachers, who participated in the present study, also participated in main TIMSS 2015 study (Trends in Mathematics and Science Study). The TIMSS study is an international study of student achievement that is con- 46 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... ducted every four years at the 4th and 8th grades; students, teachers and school heads participate in the study. Out of all TIMSS 2015 teachers (n = 257 for the 4th grade and n = 882 for the 8th grade) 562 teachers responded to our invitation; 127 were 4th grade teachers (49% response rate) and 478 were 8th grade teachers (54% response rate). Teachers came from 136 Slovenian schools; most participants were females (85%). Instruments Teachers' Relational Competence Scale (TRCS - pilot II1, Vidmar, Rutar Leban, & Niederberger, 2015) is a newly developed instrument for measuring teachers' relational competence as defined in the work of Juul and Jensen (2010). The development of the TRCS is described below, followed by a description of the instrument. An expert team of three psychologists (two researchers and one psy-chotherapeutic counsellor) studied the original work of Juul and Jensen (Juul & Jensen, 2010, 2011; Jensen & Jensen, 2011). We followed explanations and descriptions of the relational competence dimensions - respect for individuality (i.e. seeing student and leadership), authenticity and responsibility for the relationship. For each component we constructed items that would reflect its content as much as possible. The items content was reviewed also by a Danish expert for relational competence. This resulted in the scale, comprised of 33 self-report items (TRCS-pilot II, Vidmar et al., 2015, see Table 1; only two items remained from the pilot I version). The items were assessed on a 5-point Likert scale (from Very rarely or never, to Always or very often). The items presumably described teachers' relational competence along the three dimensions - individuality (9 items, e.g. I take into consideration that each student's thoughts, feelings and understanding of a given situation may differ from mine), authenticity (12 items; e.g. I am authentic in my relationships with students) and responsibility (12 items; e.g. When I can't build a good relationship with a student, I ask them for help). The respect for individuality refers to a teacher's abil- 1 The first version of the TRCS (TRSC-pilot I, Vidmar & Niederberger, 2014) was developed based on the work of Niederberger (2013) that measured relational competence within parent-child and romantic relationships. The TRSC-pilot I, comprising of 26 items was tested in preliminary study on the sample ofover 100 Slovenian teachers (n = 121) ofthe 4th and 8th grade, who participated in the TIMSS 2015 field study. Teachers completed the TRSC on-line. Using exploratory factor analyses (EFA) 1-, 2-, 3-, and 4-factor models were tested. Fit indices were low (CFI, RMSEA, SRMR) and there were several non-significant loadings and several cross-loadings. After removing the non-significant items, fit indices remained low. The items did not load onto the factors in the expected manner; the items that had reverse coding loaded on one factor (in the 2-, 3-, and 4- factor solution), whereas the 'positive' items loaded together on another factor (in the 3- and 4-factor solution, the remaining factors were weak with only one or two items loading). The results indicated that comprehensive revision ofthe TRSC was needed. 47 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 ity to recognize and acknowledge students as individuals (with their own (psychological) needs, goals, values) and teacher' ability to take this into account when leading the teaching and learning. The authenticity refers to teacher's ability to be personal in the relationship with students and to be able to act in accordance with their own values and beliefs (about teaching, learning, education) in their professional life. The responsibility refers to a teacher's ability to take exclusive responsibility for the quality of the relationship with students (i.e. for what is happening in the relationship between the teacher and the student). The presupposed factor structure (Table 1) as well as final factor structure (Tables 2 and 3) and scale reliability are presented in the results. Procedure Teachers who participated in the TIMSS 2015 were invited via e-mail to complete the on-line TRCS - pilot II (using on-line survey tool ika (https://www.ika.si/)). The participation was voluntary. Statistical Analyses Using statistical package IBM SPSS Statistics 23, we computed descriptive statistics, correlations and alpha coefficients. Using the "random" function in SPSS, we split the sample into two equal groups. We conducted exploratory factor analysis (EFA) on the first sample half and exploratory structural equation modelling (ESEM) on the other sample half, using Mplus Version 6.12. Full information maximum likelihood (FIML) algorithm was used to assess the parameters in the model. The ESEM approach is similar to confirmatory factor analyses (CFA), because it allows the pre-specification of target and non-target loadings in a confirmatory manner; in the ESEM all factor loadings are estimated with the cross-loadings targeted to be close to zero (but not fixed at 0 as is the case in the CFA) and main (target) loadings are estimated freely (Morin, Arens, & Marsh, 2016). Item loadings were interpreted according to Thabachnick and Fidell (2006) who suggest cut-off values going from 0.32 (poor), 0.45 (fair), 0.55 (good), 0.63 (very good) or 0.71 (excellent); following this rule of thumb all items with loadings smaller than 0.30 were excluded from further analysis. Model fit was interpreted following the recommendations by Hu and Bentler (1998): the comparative fit index, CFI > 0,95, the root mean square error of approximation, RMSEA < 0,06 and the Standardized Root Mean Square Residual SRMR < 0,08. 48 m. vidmar, k. kerman ■ the development of teacher's relational ... Results In the first section, descriptives for all items are presented (see Table 1). This is followed by presentation of EFA for the first random half of the sample and ESEM results for the second half of the sample. For the items in the ESEM model, inter-item correlations are also presented as well as mean differences according to grade taught and teachers' sex. Table 1. Item descriptive statistics Item/ factor Mean Std. dev. Skew. Kurt. i/R When a student doesn't agree with a certain decision I've made, I consider his/her opinion. 4,25 0,75 -0,78 0,42 2/I I take into consideration that each student's thoughts, feelings and understanding of a given situation may differ from mine. 4,38 0,63 -0,60 -0,i6 3/A As a teacher I am able to act in accordance with my values and beliefs. 4'35 0,63 -0,50 -0,44 4 */A Respecting the teacher authority is a prerequisite for effective teaching. 4,4i 0,72 -1,29 2,28 5/R When I find myself in disagreement with a certain student, I actively seek for new opportunities to (re)establish a harmonic relationship. 4'57 0,59 -1,17 i,i5 6/A I make sure, I don't expect (such) behaviours from students, that I don't express myself. 4,23 0,83 -1,10 i,45 7/R When a student behaves or expresses in an inappropriate or unsuitable way, I try to understand what lies under his/her behaviour or words. 4,04 0,72 -0,40 0,i8 8/I I am aware of values, feelings, thoughts and goals of each student. 3,61 0,74 -0,54 0,60 9/A In my demands/expectations I refer to my personal boundaries (e.g. "I do not allow this behaviour."). 4,07 0,72 -0,64 0,93 10/I I am open to student ideas and suggestions and I consider them when teaching. 4,29 0,60 -0,24 -0,6i ii/I I take into consideration that each student experiences a given situation from a different perspective. 4,i7 0,68 -0,44 0,i5 12/R As a teacher, I take full responsibility for the quality of the student-teacher relationship. 4,08 0,83 -0,75 0,48 13/R When I realise I made a mistake, I apologise to the student. 4,81 0,48 -3,08 i2,ii i4/I When I talk to a student, I encourage him/her to express his/her thoughts. 4,66 0,50 -i,00 -0,32 15/I Feelings, emotions and thoughts of my students are important to me. 4,4i 0,62 -0,87 i,54 49 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Item/ factor Mean Std. dev. Skew. Kurt. i6/A When I am with students, I focus on the present moment; I don't think about past situations nor think about the future ones. 4,II 0,76 -0,75 0,68 17/R I can effectively collaborate with every student or class. 4,06 0,68 -0,38 0,I7 18/R I am not insulted by students' inappropriate/ offensive behaviour or statements; I think of them as expression of imbalances between the student and myself/environment. 3.85 0,8I -0,52 0,46 19/A B uilding a personal teacher-student relationship is a prerequisite for effective teaching. 4,18 0,82 -0,84 0,50 20/A I am authentic in my relationships with students. 4,66 0,52 -I,I7 0,77 21/A I share my personal experiences with students when their content is appropriate and they deepen our relationship. 3,88 0,93 -0,5I -0,I7 22 */I As a teacher I try to treat all students in the same way. 4,74 0,5I -2,I9 5,76 23/A I develop a personal relationship with each student. 3,98 0,87 -0,86 0,9I 24/R When I can't build a good relationship with a student, I ask him/her for help. 3,39 0,93 -0,22 -0,I5 25/A I am aware my behaviour sets an example for the students. 4,77 0,46 -I,90 3,53 26/R After a disagreement with a student, I make sure we talk about it. 4,60 0,56 -I,02 0,04 27/I A sense of mutuality and equality in my relation with students is important to me. 4,56 0,64 -I,69 4,55 28/I I am aware that each student has his/her own way of thinking and functioning, so I try to adjust my behaviour accordingly. 4,33 0,62 -0,36 -0,66 29 */A In my demands/expectations I refer to the commonly accepted rules of behaviour (e.g. "This is no behaviour."). 3,90 0,80 -0,42 -0,07 30/R If I find myself in repeated conflicts with a certain student or a group of students, I consider my behaviour and usually modify it accordingly. 3,87 0,77 -0,35 0,20 31/R I see relationship with each student as an opportunity for my personal and professional growth. 4,18 0,75 -0,59 -0,I6 32 */R Difficult classes/students are the cause of teacher burnout. 3,88 0,85 -0,39 -0,28 33 */A As a teacher I don't show my weaknesses to the students. 3,5I 0,88 -0,20 0,07 * Reverse coding. N = 562. I = respect for individuality, A = authenticity, R = responsibility for the relationship. 50 m. vidmar, k. kerman ■ the development of teacher's relational .. As indicated in Table 1, only item 13 had excessive skew and kurto-sis that violates the recommendations on assuring multivariate normality (for all other items, univariate skewness is below 2.0 and kurtosis is below 7.0; Curran, West, and Finch, 1996). Exploratory Factor Analyses (EFA) A series of EFAs was conducted on the first random half of the sample (N = 281) to examine the factor structure and item loadings of the newly developed Teachers' Relational Competence Scale (TRCS - pilot II, Vidmar et al., 2015). Firstly, we conducted 1-, 2-, 3- and 4-factor EFA including all 33 items2. Based on fit indices (RMSEA, CFI, ^2 and SRMR), all models showed a poor fit (values ranged from .05 to .07 for RMSEA; .71 - .87 for CFI; 1205.173 (df = 495), p = .000 - 713.889 (df = 402), p = .000 for ^2 and .07 - .05 for SRMR). Across all factor solutions, two items loaded on none of the factors (items 3, 32; e.g. As a teacher I am able to act in accordance with my values and beliefs) and reversely coded items loaded positively rather than negatively on the factors (items 4, 22, 29, 33; e.g. As a teacher I don't show my weaknesses to the students). Generally, reversely coded items also correlated positively rather than negatively with other items. These six items were dropped from further analyses. Looking at the item content of these items reveals that four of the dropped items presumably tap authenticity, one individuality and one responsibility. Secondly, after dropping inadequate items (specifically items 3, 4, 22, 29, 32 and 33) 1-, 2-, 3- and 4-factor analyses were conducted again. This did not result in improved fit indices, particularly the CFI remained low (CFI < .81). In the next step we continued with exploratory factor analysis (EFA), including only items, that showed appropriate factor loadings (i.e. loaded together with other items designed to load the same factor; loading > .30), specifically items 2, 7, 10, 11, 12, 17, 18, 24, 28, 30, 31, and tested the 2- and 3-factor models. The models show appropriate fit (^2(25) = 38.947, p < .05; CFI = .98; SRMR = .03; RMSEA = .05 and ^2 (34) = 66.68; p < .001; CFI = .95; SRMR = .04; RMSEA = .06) for the 3- and 2-factor model, respectively. Even though fit indices generally meet the recommended values (Hu and Bentler, 1998) for both models, the 3-factor solution shows one dominant factor with two weaker factors (see appendix, table Ai for item loadings). In light of model parsimony, we opted for the 2-factor solution. In Table 2, we can see that all item loadings are above .35, with more items loading into factor 2 than factor 1. Factor 1 is comprised of four items (items 2, 10, 11 and 28), with item 28 cross-loading into both factors, whereas factor 2 is comprised of seven items 7, 12, 17, 18, 24, 30, 31). 2 Factor loadings for the 1-, 2-, 3, and 4-factor solution are available from the first author. 51 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Based on item content, we named the first factor Individuality and the second factor was named Responsibility. Table 2. Item loadings for EFA Factor 1 Factor 2 Item 2: I take into consideration that each student's thoughts, feelings and understanding of a given situation may differ from mine. 0.62 -0.09 Item 7: When a student behaves or expresses in an inappropriate or unsuitable way, I try to understand what lies under his/her behaviour or words. 0.24 0.44 Item 10: I am open to student ideas and suggestions and I consider them when teaching. 0.41 0.24 Item 11: I take into consideration that each student experiences a given situation from a different perspective. 0.78 0.01 Item 12: As a teacher, I take full responsibility for the quality of the student-teacher relationship. 0.20 0.38 Item 17: I can effectively collaborate with every student or class. 0.03 044 Item 18: I am not insulted by students' inappropriate/offensive behaviour or statements; I think of them as expression of imbalances between the student and myself/environment. 0.03 0.48 Item 24: When I can't build a good relationship with a student, I ask him/ her for help. -0.01 0.61 Item 28: I am aware that each student has his/her own way of thinking and functioning, so I try to adjust my behaviour accordingly. °.35 °.35 Item 30: If I find myself in repeated conflicts with a certain student or a group of students, I consider my behaviour and usually modify it accordingly. <0.01 059 Item 31: I see relationship with each student as an opportunity for my personal and professional growth. -0.05 0.75 Note. The highest loading is marked in bold. Cross-validation with Exploratory Structural Equation Modeling (ESEM) In the next set of analyses, we attempted to cross-validate the 2-factor structure obtained in the first sample half with EFA, using ESEM on the second half of the sample (N = 281). We specified the model based on EFA 2-factor model, assigning items 2, 10, 11 and 28 to factor 1 and items 7, 12, 17, 18, 24, 30 and 31 to factor 2. In the EFA, item 28 loaded on both factors; we decided to keep item 28 in presupposed target factor 1 based on: (1) item content, (2) higher correlation with other factor items (mean correlation with other items of the factor was .33 and .29 for factor 1 and factor 2, respectively) and (3) the higher loadings on the factor 1 compared to factor 2 in the 3-factor EFA solution (see appendix, table Ai). ESEM fit indices 52 m. vidmar, k. kerman ■ the development of teacher's relational .. were within the recommended range (RMSEA = .04, CFI = .98; ^2 (34) = 50.259, p < .05 and SRMR = .03). Factor loadings are presented in Table 3. Table 3. Item loadings to target factors (ESEM) Individuality p-value Responsibility p-value Item 2 0.60 0.00 0.03 0.79 Item 10 °-55 0.00 0.17 0.08 Item 11 0.74 0.00 0.08 0.46 Item 28 0.14 0.15 0.54 0.00 Item 7 0.48 0.00 0.29 0.00 Item 12 0.17 0.12 °.43 0.00 Item 17 0.09 0.54 0.40 0.00 Item 18 -0.00 °.99 044 0.00 Item 24 0.19 0.13 °.43 0.00 Item 30 0.11 0.33 °.53 0.00 Item 31 -0.16 0.12 0.87 0.00 Note. The loadings on target factors are marked in bold. Table 3 shows satisfactory loadings for all items (above .30, except for items 7 and 28). Item 28 loads more strongly on Responsibility factor, with loading on Individuality being statistically non-significant. We repeated the above described ESEM, using the complete data set, due to the cross-loading. In the full data ESEM, item 28 loads statistically significant onto to the target Individuality factor; however, it still loads more strongly onto Responsibility factor (.40) than onto Individuality factor (.27). In light of these results, further revision of item 28 (I am aware that each student has his/her own way of thinking and functioning, so I try to adjust my behaviour accordingly) is needed. Based on the final EFA and ESEM results (presented in Table 2 and 3) and item content, we named the first factor Individuality and the second factor was named Responsibility. Items 2, 10, 11 and 28 comprise the Individuality factor, with item correlations ranging from .28 to .45 (see Table 4) and Cronbach's alpha coefficient of .70. Responsibility factor is comprised of items 7, 12, 17, 18, 24, 30 and 31, with inter-item correlations ranging from .18 to .49 (see Table 4), and Cronbach's alpha coefficient of .76. Although some correlations are low, all correlations are statistically significant. The latent factors correlate moderately (r = .73, p < .001). 53 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Table 4. Inter-item correlation matrix for the final set of items 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 1. Item 2 2. Item 10 .28** 3. Item 11 .45** .42** 4. Item 28 .29** .36** .44** 5. Item 7 .30** .33** .37** .32** 6. Item 12 .16** .36** .33** .28** .33** 7. Item 17 .07 .21** .23** .27** .22** .36** 8. Item 18 .22** .15* .23** .28** .33** .23** •32** 9. Item 24 .13* .30** .25** .33** .42** .37** .26** .24** 10. Item 30 .18** .21** .29** .30** .32** .22** .18** .32** .37** 11. Item 31 .15* .35** .29** .41** .38** .29** .31** .36** .39** .49** ** Correlation is significant at the 0.01 level (2-tailed). * Correlation is significant at the 0.05 level (2-tailed). Sex and Grade Differences As this study was conducted in an exploratory manner, we also wanted to test differences in the two factors of relational competence based on teacher's gender and the grade they teach (4th and 8th) using MannWhitney U test3. The results showed that there are statistically significant differences (or come close to statistical significance) for both dimensions, based on sex (U = 4540,000 and p = .055; U = 4137,500 and p = .007; for Individuality and Responsibility, respectively) and based on grade taught (U = 5411,000 and p = .002; U = 5194,000 and p = .001 for Individuality and Responsibility, respectively). More specifically, for both Individuality and Responsibility, the higher average was found for females compared to males (^ = 17,23; a = 1,82 for Individuality and ^ = 32,39, a = 3,84 for Responsibility) and for 4th grade teachers compared to 8th grade teachers (^ = 17,72; a = 1,73 for Individuality and ¡z = 33,40 a = 3,31 for Responsibility). Discussion In the present study, we investigated the reliability and structural validity of the newly developed TRCS questionnaire for measuring teachers' relational competence, using exploratory factor analysis (EFA) and exploratory structural equation modelling (ESEM). EFAs were conducted on the 3 Mann-Whitney U test was used due to non-normal distribution of latent factors, com- puted with Shapiro-Wilk test (W = .95, p < .001 and W = .987, p < .001 for Individuality and Responsibility, respectively). 54 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... first half of the teacher sample and then cross-validated with ESEM on the other half of the teacher sample. Exploratory Factor Analyses (EFA) A series of EFAs led us to the construction of 11-items model with 2 factors. The model showed adequate fit and appropriate item loadings to respective factors (only item 28 cross-loaded into two factors). Items loadings ranged from fair to excellent. Item 28 was kept in further analysis on factor 1, based on its content, higher correlation with other items on factor 1 (compared to factor 2) and higher loading on factor 1 (compared to factor 2) in 3-factor EFA solution. The two obtained factors were named Individuality and Responsibility. Individuality consists of 4 items and taps items that reflect a teacher's ability to respect and consider each student on an equal-footing - as an individual with their own experiences, perspectives and ways of functioning. This also means that teachers are able to see beyond the most apparent behaviour or words of a student, take this into consideration in the interactions and yet remain in their leadership (guidance) role in the context of learning and teaching. Responsibility consists of 7 items and taps items that reflect a teacher's ability and willingness to take responsibility for what is happening in the teacher-student relationship - to take responsibility for the relationship in general as well as in conflicting/challenging situation (e.g. bad relationship with a student, inappropriate students' behaviour). It includes the ability to establish and maintain relationship. The two obtained factors are consistent with two components of relational competence as defined by Juul and Jensen (2010); however, the third components emphasized in their definition - teachers' authenticity - did not emerge as a cohesive factor. Thus, even though the three-factor structure (individuality, responsibility, authenticity) was presupposed, our study did not support this. Items tapping authenticity loaded on different factors and did often not correlate with each other significantly (.004 < r < .52, .00 < p < .98). Looking closely at items presumably tapping authenticity, we can see that four items have already been dropped from analyses after the first set of EFAs, due to low loadings or loadings in the wrong direction (e.g. As a teacher I don't show my weaknesses to the students; As a teacher I am able to act in accordance with my values and beliefs). These items were conceptually based, but empirically do not seem to measure authenticity. Moreover, looking at content of other items presumably tapping authenticity, they seem to also cover concepts like mindfulness (e.g. When I am with students, I focus on the present moment; I don't think about past situations 55 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 nor think about the future ones) and self-disclosure (I share my personal experiences with the students when their content is appropriate and they deepen our relationship). This heterogeneity of item content could result in the fact that authenticity did not emerge as an independent factor or did not emerge with more items on one of the other two factors. This indicates that in the future studies, the construct of authenticity should be revisited and re-examined with several newly developed items. Given that this is pioneering work in the field, and that items for the scale were newly constructed based on the conceptual work of Juul and Jensen (2010), there are not many referential studies to which we could compare our findings. Similar to Wubbels and colleagues (2006, 2012), our study supports the notion that teacher's relational competence (or more general social and emotional competences) can be reliably measured using self-report. The dimensions examined in this study differ from Wubbels' work; Juul and Jensen's (2010) concept includes components such as respect for individuality, authenticity and responsibility for the relationship, while Wubbels concept includes components such as providing guidance, setting norms and standards, correcting undesirable pupil behaviour, paying attention to pupils and giving pupils responsibility and freedom (Wubbels et al., 2012). The first and the last two of Wubbels' components could be seen as taping respect for individuality, while authenticity and responsibility for the relationship are not captured in his conceptualization. As indicated in the results, the 3-factor EFA solution with 11 items also showed adequate fit (2-factor was chosen because it is more parsimonious). Comparing the 2- and 3-factor solutions that emerged in the EFA shows that two items from the second factor emerged as a third factor in 3-factor solution. In the 3-factor solution, factor Responsibility from the 2-factor solution was decomposed in such a way, that two distinct types of responsibility occurred: on one hand responsibility for restoring relationship in time of conflict or lack of good relationship (e.g. item 30: If I find myself in repeated conflicts with a certain student or a group of students, I consider my behaviour and usually modify it accordingly) and on the other hand responsibility for the relationship in general (item 12: As a teacher, I take full responsibility for the quality of the student-teacher relationship). Future studies should bear in mind this possible distinction of the two types of responsibility when investigating the dimensions of relational competence. 56 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... Cross-validation using exploratory structural equation modelling The final 2-factor model was then cross-validated on the second data half, using exploratory structural equation modelling (ESEM). Fit indices were within the recommended range, confirming that the 2-factor structure with presumed item loadings fits the data well and most item loadings were found to be statistically significant (when repeating the ESEM on the complete dataset, all loadings proved to be statistically significant). Item 28 cross-loaded into both factors; but loaded more strongly and statistically significant onto the non-target Responsibility factor. Looking at the item content, cross-loading is not surprising. It seems that the first part of the items indeed taps respect for individuality (I am aware that each student has his/her own way of thinking and functioning..^), however the second part also describes behaviour that is indicative of a teacher's responsibility for the relationship (.so I try to adjust my behaviour accordingly). The item was kept in the final version, because the behaviour describes relational competence of teachers; the item was designated to measure factor 1 (Individuality). We suggest a revision of item 28, specifically the second part (e.g....so I try to take it into account when teaching or otherwise interacting with students). Overall, with the ESEM results, we were able to find support for the 2-factor structure of the newly developed TRCS questionnaire (as indicated by the EFA results). Sex and Grade Differences Finally, we tested whether dimension scores differ based on the teacher's sex and grade taught. Results showed statistically significant differences on both dimensions, for both categories (results for sex differences on Individuality are marginally significant). These findings indicate that female teachers (compared to male counterparts) and 4th grade teachers (compared to 8th grade teachers) report that they are better at respecting students' individuality and taking responsibility for the relationship with students. Additionally, these results, at least to some extent, demonstrate the discriminant validity of the instrument. Limitations and Strengths One of the strengths of this study is a large initial sample size, which enabled us to cross-validate the factor structure of the TRCS questionnaire. Additionally, advanced statistical tools, such as FIML and ESEM were used. The limitation is that no other aspects of validity, except structural (i.e. external, Messick, 1995) were investigated within this study; however further studies of validity will be possible after TIMSS 2015 data becomes available and will be merged with our data on relational competence. Fi- 57 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 nally, we have doubts about the generalizability of our findings. As our initial sample was representative (of Slovenian school teachers), only 50% of teachers completed the TRCS. The low response rate may also indicate a response bias, possibly related to the teachers' relational competence. Conclusion In accordance with the aims of our study we have: (1) identified items that are adequately reflected by the underlying factors using EFAs - 11 items were kept and 22 items of the initial 33 were dropped due to low or non-target loadings; (2) not found support for the presupposed three-factor structure of TRCS (individuality, responsibility, authenticity), because authenticity did not emerge as a cohesive factor; and (3) suggested to develop new authenticity items. Nevertheless, our study shows that teachers' relational competence (its two dimensions - individuality and responsibility) can now be reliably measured, using the newly developed TRCS. The scale was tested on a large sample size using advanced statistical procedures. We found that a 2-factor model, consisting of 11 items fits the data best; the first factor represents Individuality (items 2, 10, 11 and 28; a = .70) and the second factor represents Responsibility (items 7, 12, 17, 18, 24, 30, 31; a = .76). These results were confirmed with ESEM, but a modification of item 28 (that cross-loaded on both factors) was suggested. The third presumed factor, authenticity, proved to be the weakest (it did not emerge), with items usually loading onto other factors and items correlating poorly. To appropriately address the authenticity aspect of the relational competence, the construct needs to be revised and new items need to be developed and examined in future studies whether they load on a separate factor or load together with any of the other two factors. Future studies on validity of the instrument are needed. References Adamczyk, K., Pilarska, A. (2012) Attachment style, relationship status, gender and relational competences among young adults. Polish Psychological Bulletin 43 (2), pp. 59-69. Anderman, L. H. (2003) Academic and social perception as predictors of change in middle school students' sense of school belonging. Journal of Experimental Education 72 (1), pp. 5-22. Bryk, A. S., Schneider, B. (2004) Trust in schools: A core resource for improvement. Contemporary Sociology 33 (3), pp. 368-369. Bowlby, J. (1969) Attachment and loss: Vol 1. Attachment. New York: Basic Books. 58 M. viDMAR, K. KERMAN ■ THE DEvELopMENT of TEAcHER's RELATioNAL ... CASEL, Collaborative for academic, social, and emotional learning (2013) 2013 CASEL guide: Effective social and emotional learningpro-grams—preschool and elementary school edition. Chicago: Author. Curran, P. J., West, S. G., Finch, J. F. (1996) The robustness of test statistics to nonnormality and specification error in confirmatory factor analysis. Psychological Methods 1, pp. 16-29. Curby, T. W., Rimm-Kaufman, S. E., Ponitz, C. C. (2009) Teacher-child interactions and children's achievement trajectories across kindergarten and first grade. JournalofEducationalPsychology 101, pp. 912925. Engels, R. C. M. E, Finkenauer, C., Meeus, W., Dekovic, M. (2001) Parental attachment and adolescents' emotional adjustment: The associations with social skills and relational competence. Journal of Counseling Psychology 48 (4), pp. 428-439. Hamre, B., Pianta, R. (2001) Early teacher-child relationships and the trajectory of children's school outcomes through eighth grade. Child Development72, pp. 625-638. Hattie, J. (2012) Visible learningfor teachers. Maximizing impact on learning. Routledge. Hu, L. Bentler, P. M. (1998) Fit indices in covariance structure modeling: Sensitivity to underparameterized model misspecification. Psychological Methods 3 (4), pp. 424-453. Hughes, J. N., Cavell, T. A., Jackson, T. (1999) Influence of teacher-student relationships on aggressive children's development: A prospective study. Journal of Clinical Child Psychology 28, pp. 173-184. Jennings, P. A., Greenberg, M. T. (2009) The prosocial classroom: Teacher social and emotional competence in relation to student and classroom outcomes. Review ofEducational Research 79, pp. 491-525. Jensen, E. Jensen, H. (2011) Dialog s starši [Dialogue with parents]. Ljubljana: Inštitut za sodobno družino Manami. Jensen, E., Bengaard Skibsted, E., Vedsgaard Christensen, M. (2015) Educating teachers focusing on the development of reflective and relational competences. Educational Research for Policy and Practice, 14 (3), pp. 201-212. Jones, S. M., Bouffard, S. M., Weissbourd, R. (2013) Educators' social and emotional skills vital to learning. Phi Delta Kappan, 94, pp. 62-65. Juul, J., Jensen, H. (2010) Od poslušnosti do odgovornosti. [From obedience to responsibility] Radovljica: Didakta. Juul, J., Jensen, H. (2011) Vzgojno delovanje v vrtcu in šoli [Up-bringing acting in kindergarten and school]. Šolsko svetovalno delo3-4, pp. 6174. 59 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Kokkinos, C. M., Charalambous, K., Davazoglou, A. (2009) Interpersonal teacher behaviour in primary school classrooms: A cross-cultural validation of a greek translation of the questionnaire on teacher interaction. Learning Environments Research 12 (2), pp. 101-114. Ladd, G. W., Birch, S. H., Buhs, E. S. (1999) Children's social and scholastic lives in kindergarten: Related spheres of influence? Child Development 70, pp. 1373-1400. Laursen, P. Nielsen, A. M. (this issue) Teachers' relational competencies: the contribution from teacher education. Šolsko polje 27 (1-2), pp. 139161. Meehan, B. T., Hughes, J. N., Cavell, T. A. (2003) Teacher-student relationships as compensatory resources for aggressive children. Child Development 74, pp. 1145-1157. Messick, S. (1995) Validity of Psychological Assessment: Validation of Inferences From Persons' Responses and Performances as Scientific Inquiry Into Score Meaning. American Psychologist50 (9), pp. 741-749. Morin, A. J., Arens, A. K., Marsh, H. W. (2016) A bifactor exploratory structural equation modeling framework for the identification of distinct sources of construct-relevant psychometric multidimension-ality. Structural Equation Modeling: A Multidisciplinary Journal 1, pp. 116-139. Ngu, L., Florsheim, P. (2011) The development of relational competence among young High-Risk fathers across the transition to parenthood. Family Process, 50 (2), pp. 184-202. Niederberger, U. (2013) Relational Competence Theory: Can Respect, Authenticity, and Re-sponsibility for the Relationship Predict Relationship Quality?. Bachelor thesis, Kristianstad University, Sweden. http://hkr.diva-portal.org/smash/record.jsf?pid=diva2%3A63540 6d-swid= OATDFullTextWindow (1.3.2013). O'Connor, E., McCartney, K. (2007) Examining teacher-child relationships and achievement as part of an ecological model of development. American Educational Research Journal 44, pp. 240-269. OECD (2013) PISA 2012 results: ready to learn: students' engagement, drive and self-beliefs (volume iII). Paris: OECD Publishing. OECD (2014) A Teachers' guide to talis 2013: teaching and learning international survey. Paris: OECD Publishing. Pantic, N., Wubbels, T. (2012) Teachers' moral values and their interpersonal relationships with students and cultural competence. Teaching and Teacher Education 28 (3), pp. 451-460. Pianta, R. C., Hamre, B. K., Stuhlman, M. (2003) Relationships etween teachers and children. In Reynolds, W., Miller, G. (eds.). Comprehen- 60 M. VIDMAR, K. KERMAN ■ THE DEVELOPMENT OF TEACHERS RELATIONAL .. sive handbook ofpsychology (vol. 7) educational psychology . New York: John Wiley Sons, pp. 199-234. Pittman, L. D., Richmond, A. (2007) Academic and psychological functioning in late adolescence: The importance of school belonging. Journal of Experimental Education 75 (4), pp. 270-290. Schonert-Reichl, K. A., Hanson-Peterson, J. L., Hymel, S. (2015) Social and emotional learning and pre-service teacher education. In Dur-lak, J. A., Weissberg, R. P., Domitrovich, C. E., Gullotta, T. P. (eds.). Handbook of social and emotional learning: Research and practice. New York: Guilford, pp. 406-421. Schonert-Reichl, K. A., Roeser, R. W., Oberle, E., Whitehead, J., Kitil, M. J., Lawlor, M. S. (2015, April) Singular and combined effects of mindfulness-based education programs on teacher and student well-being and classroom context. In R. W. Roeser (Chair), Systematic Approaches to mindfulness-based interventions for teachers and students: Early childhood and elementary school settings. Symposium presented at the annual meeting of the American Educational Research Association, Chicago, IL. Tabachnick, B. G., Fidell, L. S. (2006) Using multivariate statistics (5th edition). Needham Heights, MA, USA: Allyn Bacon, Inc. Valiente C., Lemery-Chalfant K., Swanson J., Reiser M. (2008) Prediction of children's academic competence from their effortful control, relationships, and classroom participation. Journal of Educational Psychology 100, pp. 67-77. Vidmar, M., Niederberger, U. (2014) Lestvice odnosne kompetentnosti uciteljev-pilotna I [Teacher relational competence scale - pilot I], unpublished. Vidmar, M., Rutar Leban, T., Niederberger, U. (2015) Lestvice odnosne kompetentnosti uciteljev-pilotna II [Teacher relational competence scale - pilot II], unpublished. Vygotski, L. S. (1978) Mind and society: The development of higher psychological processes. Cambridge: Harvard University Press. Wentzel, K. R. (2002) Are effective teachers like good parents? Teaching styles and student adjustment in early adolescence. Child Development 73 (1), pp. 287-301. Wubbels, T., Brekelmans, M., den Brok, P., Van Tartwijk, J. (2006) An interpersonal perspective on classroom management in secondary classrooms in the Netherlands. In Evertson, C. M., Weinstein, C. S. (eds.). Handbook of classroom management: Research, practice, and contemporary issues. New York: Lawrence Erlbaum Associates, pp. 1161-1191. ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Wubbels, T., Brekelmans, M., den Brok, P., Levy, J., Mainhard, T. Van Tartwijk, J. (2012) Let's make things better. In Wubbels, T., den Brok, P., van Tartwijk, J., Levy, J. (eds.). Interpersonal Relationships in Education. An Overview of Contemporary Research. Rotherdam, the Netherlands: Sense Publishers, pp. 225-249. Yoon, J. S. (2002) Teacher characteristics as predictors of teacher-student relationships: Stress, negative affect, and self-efficacy. Social Behavior and Personality 30 (5), pp. 485-493. Appendix Item Loadings for the Final 3-factor EFA model Table Ai. Item loadings for the final 3-factor model EFA model Fi F2 F3 Item 2 0.666 -0.016 -0.164 Item 7 0.256 0 344 0.118 Item 10 °.398 0.120 0.187 Item 11 0.752 0.034 -0.013 Item 12 0.164 -0.004 0.666 Item 17 -0.002 0.234 °.35i Item 18 0.051 °.457 0.025 Item 24 -0.003 °.438 0.252 Item 28 °.353 0.345 0.006 Item 30 0.012 0.682 -0.116 Item 31 -0.064 0.797 -0.012 Note. The highest loading is marked in bold. 62 Intrinsic Motivation as a Key to School Success: Predictive Power of Self-perceived Autonomy, Competence and Relatedness on the Achievement in International Comparative Studies Ana Kozina and Ana Mlekuz When addressing the possible predictors of academic achievement in various domains, one firstly comes across the concept of motivation. For instance, one of the frequently asked questions is: How to motivate students to learn more, to achieve more and to have higher conceptual knowledge. One of the most empirically supported theories of contemporary psychology of motivation is Self-Determination Theory (SDT) (Ryan & Deci, 2002). SDT was chosen as a framework of present paper due to its in-depth model of intrinsic and extrinsic motivation with practical implications also for the field of education. SDT Theory SDT focuses on the quality of motivation and not so much on the quantity of motivation. SDT theory differentiates amotivation, controlled (extrinsic) motivation and autonomous (intrinsic) motivation. Amotivation is defined as a total lack of motivation. Intrinsic motivation is the inherent propensity to seek out novelty and challenge, to extend and exercise ones capacities, to explore and to learn (Reeve, 20i5). On the other side, extrinsic motivation arises from environmental incentives (rewards, consequences, punishments) that are separate from activity itself (Reeve, 2015). All three types of motivation can be placed on the continuum of perceived locus of control or self-determination. The type of motivation is closely linked to the perceptions that individuals have on the origins of their behaviour (whether they are inside or outside of their control). On one end of the continuum is amotivation (a total lack of intentionality and motivation). In the continuum, amotivation is followed by four types of extrinsic motivation that can be distinguished be- 63 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 tween one another depending on the degree of autonomy: external regulation (not at all autonomous), introjected regulation (somewhat autonomous), identified regulation (mostly autonomous) and integrated regulation (fully autonomous). On the other end of the continuum is intrinsic motivation as the highest level of self-determination (Ryan & Deci, 2000). One of the many advantages and practical implications of SDT is that it explains how amotivation can be changed to extrinsic motivation (in the process of internalization) first using external regulation (the task is done in order to obtain rewards or avoid negative consequences), then introjected regulation (the task is done in order to improve self-esteem and avoid shame, guilt and anxiety) to identified regulation (the task is done because students feel it is important and related to their personal goals - they consciously apply value to it) and finally to integrated regulation (the task is done because it represents an integral part of values and needs of student). The level of self-determination, perceived autonomy, increases as we move on through the continuum. The building stone of the change form external to internal motivations is fulfilment of three basic psychological needs: autonomy, competence and relatedness. SDT in Educational Setting Two decades of research in the educational setting provided empirical support for this conclusion: intrinsically (autonomously) motivated students thrive in educational setting (Reeve, 2002; Miserandino, 1996; Flink et al., 1992). Students who have high levels of intrinsic motivation have higher academic achievement and more conceptual knowledge (Guay & Valle-rand, 1997; Hardre and Reeve, 2003). Research showed that low achieving students typically have lower levels of intrinsic motivation and identified regulation and higher levels of amotivation (Legault, Green-Demers, & Pelletier, 2006). The mechanisms linking intrinsic motivation and academic achievement are: engagement and effort. When students are intrinsically motivated they experience engagement. The higher a person's intrinsic motivation the greater will be his or her engagement in task (e.g. school tasks in school setting), effort to pursuit their goals (Sheldon & Elliot, 1999) and focused attention in class (Hardre & Reeve, 2003; Skinner & Belmont, 1993) which is all related to higher academic achievement. Students that are intrinsically motivated perceive their school-related tasks as decided by themselves (self-determined) and based on their personal values and interests (Alivernini & Lucidi, 2011). On the other side, the problem with extrinsic motivations is that when these environmental incentives are withdrawn the behavior stops as well - for instance if student is externally motivated to be in school (e.g. grades, parental pressure) 64 a. kozina, a. mlekuz ■ intrinsic motivation .. and if these external rewards or punishers are gone (or student no longer finds them relevant), the student will become less engaged which leads to lower academic achievement and perhaps even early school leaving. All this support the notion that it is crucial to support students' intrinsic motivation in order to see students succeed in school and in later life. According to the SDT framework, the type of motivation depends on the fulfilment of three basic psychological needs: need for autonomy, need for competence and need for relatedness. People are naturally intrinsically motivated to learn and in a supporting environment all three basic needs are met and learning is intrinsically motivated and of a higher quality. The need for autonomy refers to being the perceived origin or source of one's own behaviour (Deci & Ryan, 2002). The need for competence refers to feeling effective in one's ongoing interactions with the social environment and experiencing opportunities to exercise and express ones capacities (Deci & Ryan, 2002). The need for relatedness refers to feeling connected to others, to caring for and being cared for by others, to having a sense of belonging both with other individuals and with one's community (Ryan, 1995). Students become more intrinsically motivated when their basic psychological needs of autonomy, competence and relatedness are fulfilled. The need for competence and autonomy are the most important ones in the development of intrinsic motivation whereas the need to relatedness is crucial when transforming external regulation to autonomous regulation and supporting the process of internalization (Deci & Ryan, 2002). Legault and colleagues (Legault et al., 2006) found that the lack of support for the three needs contributed to amotivation (a total lack of motivation or the lowest level of self-determination). Students that are amotivated do not want to study and they feel they cannot change their academic outcomes and the most likely consequence of those feelings is that these students would leave their schooling as soon as they can. Experimental work shows when students are tested or given rewards for activities that for them are intrinsically motivated, their intrinsic motivation decreases due to lowering their sense of autonomy. In contrast, providing students with choice (thus supporting autonomy) and positive feedback (thus supporting competence) typically increases intrinsic motivation. The satisfaction of all three needs results in strong intrinsic goals (e.g. personal growth, affiliation, community) that is linked to greater psychological well-being and better academic and non-academic outcomes (Ryan & Deci, 2009). An educational setting can promote all three psychological needs. For instance Vallerand and colleagues (Vallerand, et al., 1997) introduced the model in which low levels of autonomy supportive behaviours from 65 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 critical social agents (teachers, parents, school administration) undermine students' perceptions of their own autonomy and competence which in turn decreases self-determined motivation. They (ibid.) studied lower educational achievement in the light of early school leaving. They revealed the contextual and motivational predictors of early school leaving by assessing students for their perception of their autonomy and the autonomy support and investigating which students would be more likely to still be in school a year after. They found that the students that felt more autonomous and had more autonomy support felt more competent and were more likely to stay in school a year after. In the classrooms where teachers are more autonomy supportive (e.g. letting students choose from various alternatives, listening to them and asking them for their point of view), students tend to become more intrinsically motivated, perceive themselves as more competent, and feel better about themselves, whereas in a classroom where teachers were more controlling (e.g. giving rigid directions or orders, supervising and monitoring too closely or not giving students the opportunity to propose choices and opinions that differ from those expressed by adults), students tended to lose intrinsic motivation, perceived competence and self-esteem (Ryan & Grolnick, 1986; Tsai et al., 2008), have lower educational achievement and are more prone to early school leaving (Vallerand, et al., 1997). Intrinsic motivation is not only related to a higher quantity of knowledge but also to higher quality knowledge. In an experiment (Benware & Deci, 1984) students were given three hours to read a text. The first group was told they are going to be tested afterwards (low intrinsic motivation is expected) and the other that they will be given a chance to use their knowledge in practice by teaching others (higher intrinsic motivation is expected). The two groups did not differ significantly in the information memorized but did differ in conceptual knowledge. The findings were replicated in numerous studies around the world (Grolnick & Ryan, 1987; Kage & Namiki; 1990; Fortier, Vallerand & Guay, 1995). Aim of the Study Based on the rich empirical support and sound theoretical conception we aim at investigating the predictive power of intrinsic motivation for academic achievement in international comparative studies. In order to identify the level of intrinsic motivation of students we will focus on the indicators (items) measuring fulfilment of the three psychological needs that lead to intrinsic motivation: competence, autonomy and relatedness. International comparative studies together with reliable and valid measurement of achievement (in specific domains) measure also student's back- 66 a. kozina, a. mlekuz ■ intrinsic motivation .. ground information using background questionnaires. In order to analyse the predictive power of all three psychological needs for achievement, we firstly have to identify items (in the background questionnaires) indicating the fulfilment of all three psychological needs. In order to increase generalizability of the findings, we used four different international comparative studies: PIRLS (Progress in International Reading Literacy Study), ICCS (International Civic and Citizenship Survey), TIMSS (Trends in International Mathematics and Science Study), TIMSS Advanced (Trends in International Mathematics and Science Study Advanced). All studies measure achievement in a specific domain (PIRLS - reading comprehension; TIMSS and TIMSS Advanced - math and science achievement; ICCS - civic and citizenship literacy) and in a specific age group (PIRLS - 4th grade, TIMSS - 8th grade; TIMSS Advanced -13th grade); ICCS - 9th grade). The most recent data for Slovenia from the selected studies are used. Specific objectives of this paper are: - to identify indicators of competence, autonomy and relatedness in the pool of all items from background questionnaires of selected studies; these indicators have up to now not yet been identified and will be used in the present paper for the first time; - to identify the predictive power (regression analyses) of perceived autonomy support for achievement in selected international comparative studies (controlling for SES); - to identify the predictive power (regression analyses) of perceived competence for achievement in selected international comparative studies (controlling for SES); - to identify the predictive power of perceived relatedness (regression analyses) for achievement in selected international comparative studies (controlling for SES). Due to differences in the item pools of the studies used, the findings will be preliminary and used for further more in depth analyses of international comparative studies achievement predictors. The content of the items is nevertheless similar (also due to the studies being conducted by the same international organization: IEA International Evaluation Association). 67 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Method Participants In order to increase the generalizability of our findings, the aim was to include the wider possible age and content range. Therefore four different international comparative studies (data bases) for Slovenia were used. Below each study with participating sample is presented in detail. 4th Grade - Reading. Progress in International Reading Literacy Study 2011 (PIRLS 2011) is an international assessment of reading comprehension at the fourth grade which is conducted every five years (Martin & Mullis, 2013b). For our analysis data for Slovenia from 2011 study cycle were used. PIRLS uses a two-stage random sample design, with a sample of schools drawn as a first stage and one or more intact classes of students selected from each of the sampled schools as a second stage (Joncas & Foy, 2013). The target population of PIRLS are all students in their fourth year of formal schooling and therefore their mean age at the time of testing in each country is 9.5 years. Each sampled student answered a cognitive test and a background questionnaire. 8 th Grade - Mathematics. Trends in International Mathematics and Science Study 2011 (TIMSS 2011) is an international assessment of mathematics and science comprehension at the fourth and eighth grade of formal schooling, which is conducted every four years. As PIRLS, TIMSS also uses a stratified two-stage random sample design with firstly sampling schools and then sampling one or more whole classes at a sampled school (Joncas & Foy, 2013). The target population of TIMSS are students in their fourth and eighth year of formal schooling averagely aged at least 9.5 (fourth grade) or 13.5 (eighth grade) years at the time of testing. Since we used the PIRLS data for fourth grade and as already stated at the beginning, the aim of the article is to cover as wide an age range and knowledge range as possible, TIMSS 2011 Slovenian eighth grade sample for Mathematics was used for the purposes of this analysis. Since mathematics and science achievements in TIMSS 2011 are highly correlated (r = 0.85) we only used mathematics sample. Each sampled student answered a cognitive test and a background questionnaire. 9th Grade - Civics and Citizenship Knowledge. International Civic and Citizenship Survey 2009 (ICCS 2009) focused on civics and citizenship content and knowledge. Again to cover the wid- 68 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... est age range possible ICCS Slovenian additional grade1 database is used (students enrolled in Grade 9) in the analyses. The population of schools with Grade 9 students tested was identical to the population of schools with Grade 8 students tested (Schulz, Ainley & Fraillon, 2011) and the average age of students enrolled in Grade 9 should be 14.5 or higher. Slovenia used a two-stage cluster sampling, where schools were sampled within the country using probability proportional to size measured by the number of students enrolled in a school. Within each sampled school, an intact class from the target grade was sampled randomly and all students in that class were surveyed (Schulz et al., 2011). With these sampling procedures the representativeness of the selected test population was ensured. Each student answered a cognitive test and a background questionnaire. 13 th Grade - Mathematics. Trends in International Mathematics and Science Study Advanced 2008 (TIMSS Advanced 2008) is an international assessment of achievements in advanced Mathematics and Physics in the final year of upper secondary school (in Slovenia these are students in their 13 th year of schooling). The survey measured achievement in two student populations, one in Mathematics and one in Physics. Again the Slovenian mathematics sample of students included in TIMSS Advanced 2008 was used for the purpose of this analysis. Mathematics sample was used since it was larger than the physics sample and therefore it is more representative for the age group. TIMSS Advanced uses the same two-stage sampling design as PIRLS 2011 and TIMSS 2011 (Arora, Foy, Martin & Mullis, 2009). Instruments and Included Variables For selected studies the data gathered with achievement test and background questionnaires were included. Included variables are listed below separately for all age groups. All ages - All Surveys Achievement Scores (Plausible Values). To test achievement, all surveys use matrix-sampling approach where achievement items are divided into groups, blocks or sets and each achievement booklet is then made up of these sets of items according to a systematic arrangement (Mullis et al., 2009a; Mullis et al., 2009c; Schulz, Ainley, Fraillon, 2011; Arora, Foy, Martin & Mullis, 2009). The PIRLS cognitive assessment is composed of 10 blocks of items. Each booklet consists of two blocks, with each block containing a reading passage and 12-17 associated items. One half of the blocks assess the 1 The additional grade sample was used for estimating trends from CIVED 1999. 69 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 literary purpose and the other half assesses informational purpose. Eight blocks are then assigned to 12 booklets according to a specific plan that enables linking among booklets and balances position effects. The remaining two blocks (one literary and the other informational) are presented in a magazine format in the PIRLS reader2 (Mullis et al., 2013). TIMSS cognitive test is composed of packages of the entire pool of mathematics and science items. Each item appears in two booklets, providing a mechanism for linking together the student responses from various booklets. In TIMSS 2011 28 assessment blocks are distributed across 14 student achievement booklets. Each booklet consists of four blocks of items, two blocks of mathematics and two blocks of science items. (Mullis et al., 2009c). ICCS cognitive test consisted of 80 items which were allocated to seven clusters that were assembled into a fully balanced rotated test design comprising of seven paper-based booklets (Schulz, Ainley, Fraillon, 2011). The ICCS assessment framework includes four content (civic society and systems; civic principles; civic participation; civic identities) and two cognitive (knowing; reasoning and analysing) domains. The cognitive test or the assessment in TIMSS Advanced is composed of 14 item blocks - a total of 72 advanced mathematics items were included in the assessment. These items were distributed across 8 booklets. The design was chosen to ensure that each student responded to a sufficient number of items to provide a reliable measure (Arora, Foy, Martin & Mullis, 2009). To derive student achievement scores for analysis and reporting Rasch one-parameter item response theory (IRT) model was used in all surveys. Since each student answers only some questions the surveys use multiple imputations (plausible values3) to obtain proficiency scores. In order to enhance the reliability student responses are combined with information about student's background using the "conditioning" scaling approach (Foy, Brossman & Galia, 2013; Schulz et al., 2011). Socioeconomic Background (SES) Variables or scales that each respective survey uses as a socioeconomic background indicator in their respective international reports were used in the present article as control variables of socioeconomic background in the regression analyses. 2 PIRLS reader is a magazine-type format of reading booklet with the questions in a separate booklet. 3 There are five plausible values for each student in each ofthe databases and all five plausible values for each survey were included in the computing of the achievement score for a respective student. 70 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... PIRLS and TIMSS use scale for home resources for learning (AS-BGHRL in PIRLS and BSBGHRL in TIMSS). The scale combines answers to the questions on parents' education, parents' occupation, number of children's books and books at home and the availability of internet connection and their own room (Mullis et al., 2012b; Mullis et al., 2012a). In ICCS three different measures for socioeconomic background of students were used (parental occupational status, parental educational attainment and home literacy resources). Since the parental occupational status (HISEI) explained the highest percentage ofvariance in civic knowledge we used this scale as a measure for socioeconomic background in our analyses (Schulz et al., 2010). In TIMSS Advanced report number of books at home (MS2G-BOOK) was used as a measure of socioeconomic background (Mullis et al, 2009b). 4th Grade (PIRLS 2011). Autonomy. The autonomy of the student is measured with one set of questions ASBG07. In the set of questions ASBG07 students had to evaluate how often (every day or almost every day, once or twice a week, once a month or twice a month or never or almost never) the following things happen at home: My parents ask me what I am learning in school; I talk about my schoolwork with my parents; My parents make sure that I set aside time for my homework; My parents check if I do my homework. Cronbach alpha's for this set of questions is 0.61 which is relatively low due to small number of items included. This will be considered in further interpretation of results. In order to define the underlying structure of the question set measuring autonomy support the correlation matrix of the question set was subjected to factor analyses (method: principal axis factoring). The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.661; Bartlett's Test of Sphericity (^2) = 1868.123; p < .001). The Kai-ser-Guttman criteria (Eigenvalue over 1) revealed one factor explaining 28.75 % of the question set total variance. We named the factor autono-my_PIRLS. Factor loading ranged from 0.403 to 0.616 for selected sub questions. Higher values reflect higher degrees of autonomy. In the regression models autonomy_PIRLS was used. Competence. For competence Students Confident in Reading (ASBGSCR) scale was used. The scale or factor already existed in the PIRLS database and it was created based on students' degree of agreement (agree a lot, agree a little, 71 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 disagree a little, disagree a lot) with seven statements: I usually do well in reading (+)4; Reading is easy for me (+); Reading is harder for me than for many of my classmates (-); If a book is interesting, I don't care how hard it is to read (+); I have trouble reading stories with difficult words (-); My teacher tells me I am a good reader (+); Reading is harder for me than any other subject (+). Cronbach's alpha for these questions is 0.77. The higher values on this scale reflect perceptions of higher reading competence (Martin & Mullis, 2013a). In the regression model scale ASBGSCR (Student Confident in Reading) was used as measure for competence. Relatedness. We used a set of questions ASBG08 to measure relatedness. Students had to evaluate their agreement (agree a lot, agree a little, disagree a little, disagree a lot) on three statements: I like being in school (+); I feel safe when I am at school (+); I feel like I belong at this school (+). Cronbach alpha's for this set of questions is 0.66 which is relatively low due to small number of items included. This will be considered in further interpretation of results. In order to define the underlying structure of the question set measuring relatedness the correlation matrix of the question set was again subjected to factor analyses (method: principal axis factoring). The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.655; Bartlett's Test of Sphericity (^2) = 1917.110; p < .001). The Kaiser-Gutt-man criteria (Eigenvalue over 1) revealed one factor explaining 40.07 % of the question set total variance. We named the factor relatedness_TIMSS. Factor loading ranged from 0.564 to 0.690 for selected sub questions. The higher values on this factor represent higher students' perception of relatedness. Factor relatedness_TIMSS was used in the regression model as measure for relatedness. 8th grade (TIMSS 2011). Autonomy. As a measure of autonomy BSBG11 set of questions from the TIMSS background questionnaire was used. These questions were formulated in the same way as the question ASBG07 from PIRLS 2011, therefore see the previous section of the article (PIRLS 2011; Autonomy). Cronbach alpha's for this set of questions is 0.68 which is relatively low due to small number of items included. This will be considered in further interpretation of results. To define the underlying structure of the question set measuring autonomy the correlation matrix of the question set was subjected to factor 4 In the bracket the method of item valuation is presented. The same approach ofpresenting the item valuations was used throughout the article. 72 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... analyses (method: principal axis factoring). The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.715; Bartlett's Test of Sphericity (^2) = 2683.139;p < .001). The Kaiser-Guttman criteria (Eigenvalue over 1) revealed one factor explaining 35.56 % of the question set total variance. Factor loading ranged from 0.515 to 0.667 for the selected set of questions. We named the factor autonomy_TIMSS. The higher values on this factor reflect higher students' perception of autonomy. In the regression model factor autonomy_TIMSS was used. Competence. The scale Student Confident in Mathematics (BSBGSCM) was used as a measure of competence. The scale is based on students' degree of agreement (agree a lot, agree a little, disagree a little, disagree a lot) to the nine following statements: I usually do well in mathematics (+); Mathematics is more difficult for me than for many of my classmates (-); Mathematics is one of my strengths (+); I learn things quickly in mathematics (+); Mathematics makes me confused and nervous (-); I am good at working out difficult mathematics problems (+); My teacher thinks I can do well in mathematics lessons with difficult materials (+); My teacher tells me I am good at mathematics (+); Mathematics is harder for me than any other subject (-). Cronbach's alpha for this set of questions is 0.89. The higher values on this scale reflect perceptions higher mathematics competence (Martin & Mullis, 2013a). In the regression model scale BSBGSCM (Student Confident in Mathematics) was used as competence predictor. Relatedness. For a measure of relatedness BSBG12 set of questions was used. Again the question was formulated the same as the question ASBG08 in PIRLS 2011 (see PIRLS 2011, Relatedness). Cronbach's alpha for this set of questions is 0.71. The correlation matrix of the question set was subjected to factor analyses (method: principal axis factoring) in order to define the underlying structure of the question set measuring autonomy. The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.667; Bartlett's Test of Sphericity (^2) = 2339.618; p < .001). The Kaiser-Guttman criteria (Eigenvalue over 1) revealed one factor explaining 44.40 % of the question set total variance. Factor loading ranged from 0.597 to 0.735 for the selected set of questions. We named the factor relatedness_ TIMSS. The higher values on this factor reflect higher students' perception of relatedness. In the regression model factor relatedness_TIMSS was used. 73 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 9th Grade (ICCS 2009). Autonomy. To estimate autonomy scale Student's perception of openness in classroom discussion (OPDISC) was used. To derive this scale six statements to which students could answer with never, rarely, sometimes or often were used. These statements were: teachers encourage students to make up their own minds (+); teachers encourage students to express their opinions (+); students bring up current political events for discussion in class (+); students express opinions in class even when their opinions are different from most of the other students (+); teachers encourage students to discuss the issues with people having different opinions (+); teachers present several sides of the issues when explaining them in class (+). The scale's reliability (Cronbach's alpha) is 0.77. The higher values on the scale reflect perceptions of higher levels of classroom discussion and therefore a higher level of autonomy (Schulz, Ainley, & Fraillon, 2011). In the regression model scale OPDISC was used as a measure of autonomy. Competence. As a measure of competence we used Students' sense of internal political efficacy (INPOLEF). This scale is composed of six statements: I know more about politics than most people my age (+); when political issues or problems are being discussed, I usually have something to say (+); I am able to understand most political issues easily (+); I have political opinions worth listening to (+); As an adult I will be able to take part in politics (+); I have a good understanding of the political issues facing this country (+). Response options ranged from strongly agree to strongly disagree. The scale's reliability (Cronbach's alpha) is 0.86. The higher level on this scale reflects a higher sense of internal political efficacy and consequently competence (Schulz et al., 2011). As a measure of competence scale INPOLEF was used in the regression model. Relatedness. To measure relatedness scale Student's perception of student-teacher relation at school (STUTREL) was used. Scale is composed of five variables -students' answers to questions: Most of my teachers treat me fairly, Students get along well with most of the teachers, Most teachers are interested in students' well-being, Most of my teachers really listen to what I ha ve to say and If I need extra help, I will receive it from my teachers. Response options ranged from strongly agree to strongly disagree. The scale's reliability (Cronbach's alpha) is 0.76. The higher values on this scale reflect perceptions of strong relations between students and teachers and therefore relatedness (Schulz, 74 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... et al., 2011). In the regression model scale STUTREL was used as a measure for relatedness. 13th Grade (TIMSS Advanced 2008). Autonomy. As a measure of autonomy three sub questions were used: MS2MACWP, MS2MDL05 and MS2MDL06. The students had to evaluate how often (every or almost every lesson, about half the lessons, some lessons and never) they do the following activities in their mathematic lessons: We work problems on our own (+); We decide on our own procedures for solving complex problems (+); We communicate our arguments (+). Cronbach alpha's for this set of questions is 0.64 which is relatively low due to small number of items included. This will be considered in further interpretation of results. The correlation matrix of the question set was subjected to factor analyses (method: principal axis factoring) in order to define the underlying structure of the question set measuring autonomy. The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.567; Bartlett's Test of Sphericity (^2) = 1082.188; p < .001). The Kaiser-Gutt-man criteria (Eigenvalue over 1) revealed one factor explaining 48.82 % of the question set total variance. Factor loading ranged from 0.366 to 0.942 for the selected. We named the factor autonomy_TIMSSAdv. The higher values on this factor reflect higher students' perception of autonomy. In the regression model factor autonomy_TIMSSAdv was used. Competence. In the analyses three sub questions, namely MS2MWSWM, MS2M-WSLT and MS2MWSEP, were used as a measure of competence. The students had to evaluate how important (very important, important, unimportant or very unimportant) were for them the following reasons for studying advanced mathematics: I usually do well in mathematics (+); Studying or doing mathematics homework does not take me a lot of time (+); I expect that I will easily pass the tests (+). Cronbach alpha's for this set of questions is 0.6 which is relatively low due to small number of items included. This will be considered in further interpretation of results. In order to define the underlying structure of the question set measuring competence the correlation matrix of the question set was subjected to factor analyses (method: principal axis factoring). The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.637; Bartlett's Test of Sphericity (^2) = 619.652;p < .001). The Kaiser-Guttman criteria (Eigenvalue over 1) revealed one factor explaining 32.99 % of the question set total variance. Factor loading ranged from 0.549 to 0.604 for the se- 75 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 lected. We named the factor competence_TIMSSAdv. The higher values on this factor reflect higher students' perception of competence. In the regression model factor competence_TIMSSAdv was used as a measure for competence. Relatedness. Relatedness was measured with nationally added School climate scale (Kozina, Rožman, Vršnik Perše & Rutar Leban, 2012) that measure four dimensions of school climate: relations-school, relations-teacher and relations-students, organizational structure. We used first three dimensions in the analyses. The factors were composed of sub questions from Slovenian national question N5 (19 of 22 sub questions), which tried to evaluate the degree of agreement (completely agree, agree, uncertain, disagree, completely disagree) to statements on students' relations at school in general (e. g. I like being at school; I am proud at my school, etc.), student-teacher relations (e. g. We get on well with our teachers; Inappropriate behaviour is always duly punished at our school, etc.) and relations among students (e. g. Students at our school have good relations etc.). The scale is reliable (0.66 < a > 0.80). We used total scores on these three factors to create a new multidimensional relatedness scale. In order to define the underlying structure of the three factors measuring relatedness, the correlation matrix of the factors was subjected to factor analyses (method: principal axis factoring). The preliminary test showed the data are suitable for this kind of analyses (KMO = 0.609; Bartlett's Test of Sphericity (^2) = 1354.404; p < .001). The Kaiser-Guttman criteria (Eigenvalue over 1) revealed one factor explaining 50.46 % of the question set total variance. Factor loading ranged from 0.483 to 0.935 for the selected. We named the factor relatedness_ TIMSSAdv. The higher values on this scale reflect higher students' perception of relatedness. In the regression model factor relatedness_TIMSS-Adv was used as a measure for relatedness. Procedure and Statistical Analyses. For the analyses of the data we used two statistical programmes. First, we used IBM SPSS Statistics 23 for the calculation of internal reliability and for defining the underlying structure of the factors with the method Principal Axis Factoring, and secondly, we used IDB Analyser 3.2.19 for regression analyses, since this programme, unlike IBM SPSS Statistics, takes into account all five plausible values for student achievement in all mentioned surveys. Since the present analysis is a preliminary one, we used simple regression analyses instead of HLM. In the regression mod- 76 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... el achievement scores (all five plausible values) for each of the surveys are used as predicting variables. Results The results present the regression analyses where achievement score is predicted by competence, autonomy and relatedness. Moreover, we also included social and economic status in the regression models as a control variable. Table 1. Descriptive statistics Gender (%) Average N Age (s. e.) Female (s. e.) Male (s. e.) achievement score (s. e.) PIRLS 2011 4466 9,9 (0,0l) 48 (0,8) 52 (0,8) 530 (2,0) TIMSS 2011 4415 13,9 (0,0l) 49 (0,9) 51 (0,9) 505 (2,2) ICCS 2009 3042 i4,8 (0,0i) 49 (0,8) 5i (0,8) 540 (2,6) TIMSS Advanced 2156 18,8 (0,01) 60 (i,8) 40 (i,8) 457 (4,4) 2008 The average age of students included in PIRSL 2011 was almost 10 (9.9) and their average achievement score was 530 score points. Moreover, 4466 students participated in the PIRLS survey, among which 48% were girls and 49% were boys. In TIMSS 2011 4415 students were included in Slovenia, of which 49% were girls and 51% were boys. Their average achievement score on the cognitive test was 505 score points and their average age was 13.9 years. Likewise, in Slovenia 3042 students were included in ICCS (49% female and 51% male). Their average age was almost 15 (14.8) and their average score on ICCS cognitive test was 540 score points. In TIMSS Advanced 2156 students participated. There were 60% girls and 40% boys in the sample and their mean age was almost 19 years (18.8). Their average mathematics achievement was 457 score points. The data shows competence, autonomy and socioeconomic background as significant predictors of achievement in PIRLS 2011, whereas relatedness is not a statistically significant predictor of achievement. If perceived reading competence increases by one unit (the average value), the students' reading achievement score increases by 12 score points (if autonomy, relatedness and SES are held constant). If autonomy raises by one unit while controlling for other predictors, then the students' reading achievement increases by almost 9 score points. However, if SES raises by one unit while controlling for other predictors, then the students' achievement increases by 17 score point. The regression model for PIRLS accounts for 31% of the variance of the students' reading achievement score. 77 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Table 2. Predictive power of self-determination theory concepts for achievement. b (SE) P (SE) R2* (SE) 4th Grade - PIRLS 2011 constant 235-°° (9-99) Competence (ASBGSCR) 11.52 (0.68)* a35 (0.02)* Autonomy (autonomy_PIRLS) 8.62 (1.61)* 0.10 (0.02)* Relatedness (relatedness_PIRLS) -0.34 (1.64) 0.00 (0.02) SES (ASBGHRL) 17.08 (0.91)* a35 (0.02)* 0.31 (0.02) 8th Grade - ■ TIMSS 2011 Constant 167.03 (10.52) Competence (BSBGSCM) 19.78 (0.80)* 0.50 (0.02)* Autonomy (autonomy_TIMSS) 12.02 (1.32)* 0.14 (0.02)* Relatedness (relatedness_TIMSS) 2.70 (1.61) 0.03 (0.02) SES (BSBGHRL) 12.98 (0.83)* 0.25 (0.02)* 0.41 (0.01) 9th Grade - ICCS 2009 constant 299.20 (14.46) Competence (INPOLEF) 1.92 (0.18)* 0.24 (0.02)* Autonomy (OPDISC) 1.43 (0.24)* 0.15 (0.02)* Relatedness (STUTREL) a38 (0.30 0.04 (0.03) SES (HISEI) 1.27 (0.12)* 0.23 (0.02)* 0.16 (0.01) 13th Grade - TIMSS Advanced 2008 constant 460.09 (3.81) Competence (competence_TIMSSAdv) 27.16 (2.84)* 0.24 (0.03)* Autonomy (autonomy_TIMSSAdv) 9.98 (3.°2)* 0.11 (0.03)* Relatedness (relatedness_TIMSSAdv) 16.58 (3.17)* 0.18 (0.03)* SES (MS2GBOOK) -5.86 (6.14) -0.03 (0.03) 0.13 (0.02) Notes. Owing to the nested sampling design, data were weighted with Total Student Weight (PIRLS, TIMSS, TIMSS Advanced) or Final Student Weight (ICCS). To increase the possibility of generalisation, we used adjusted R2. Multiple regression was conducted (forced entry method) on IDB Analyzer (IEA DPC, v.3.2). The assumption ofmul-ticollinearity was tested on all data sets and was not violated (the VIFs were lower than 10 (Myers, 1990) and varied between 1.12 and 1.25). Statistically significant (p > 0,05) coefficients are marked with *. The regression analysis for TIMSS 2011 also shows that related-ness is not a statistically significant predictor of achievement. Howev- 78 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... er, SES, competence and autonomy are statistically significant predictors of achievement. Every unit increase in perceived competence is associated with a 20 score point increase in the students' mathematics achievement (if other predictors are held constant) and every unit increase in perceived autonomy is associated with a 12 score point increase in students' mathematics achievement (if other predictors are held constant). Moreover, every unit increase in SES is associated with a 13 score point increase in students' mathematics achievement (if other predictors are held constant). The regression model for TIMSS achievement accounts for 41% of the variance of the students' mathematics achievement. The data again shows SES, competence and autonomy as significant predictors of achievement in ICCS 2009 survey as well, whereas related-ness again is not a statistically significant predictor of achievement. The results show that if competence increases by one unit, the students' civic and citizenship achievement score increases by 2 score points (if all other predictors are constant). If autonomy increases by one unit, then the students' civic and citizenship score increases by a little bit more than 1 score point while controlling for other predictors. Moreover, if SES increases by one unit, then the students' civic and citizenship knowledge increases by 1 score point too (if other predictors are held constant). The model accounts for 16% of the variance of the students' civic and citizenship achievement. The regression analysis for TIMSS Advanced however shows that all three SDT predictors (competence, autonomy and relatedness) of mathematics achievement are statistically significant, whereas SES is not a statistically significant predictor. The results show that if competence increases by one unit, the students' advanced mathematics score increases by 27 score points (if all other predictors are held constants). If autonomy increases by one unit, then the students' advanced mathematics score increases by almost 10 score points (again if all other predictors are constant). Moreover, if autonomy and competence are held constant and re-latedness increases by one unit, then students' advanced mathematics score increases by almost 17 score points. The model accounts for 13% of the variance in the students' advanced mathematics achievement. Except for TIMSS Advanced, where the strongest predictor for advanced mathematics score is autonomy, and PIRLS, where the strongest predictor for reading score is SES, competence is the strongest predictor among all other predictors for other surveys (TIMSS and ICCS). Moreover, relatedness is a statistically significant predictor only in TIMSS Advanced regression model. In all other models relatedness is not a statistically significant predictor of achievement. 79 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Discussion The study aimed at identifying the possible indicators of psychological needs fulfilment in the background questionnaires of selected international comparative studies in order to analyse the predictive power of these indicators for educational achievement across content domains. Rough comparisons were made across different studies, different age groups and content domains. The study is the first of this type (to use the international comparative studies in order to find possible indicators of psychological needs fulfilment among students and their relationship with academic achievement) in Slovenia. Even though the data available and analyses conducted are rough and preliminary, the findings show consistency across studies and across age groups. In all samples included, self-perceived confidence and self-perceived autonomy are significant predictors of students' academic achievement. The more competent and the more autonomy supported students feel, the higher on average is their achievement (when controlling for SES). The percentages of explained achievement variances with the included indicators of psychological needs (SES included) are largest in 4th and 8th grade students, followed by 13 and 9th grade students. Self-perceived competence is a significant predictor of student's achievement in various content domains: reading, maths and civic education and across age groups from 4th to 13th grade. The competence is the strongest predictor out of the ones included (compared to autonomy, re-latedness and SES) in 8th, 9th and 13th grade indicating that an increase in perceived competence would result in significant increase in achievement. In 4th grade, SES is the strongest predictor and is followed by perceived self- competence. When students feel their efforts and abilities are being recognized, they use the learning situation for building their academic success. We see consistency across samples even though the achievement domain varies and competence measures vary. The items measuring self-perceived competence varied across studies and focused on specific domains of the study (e.g. self-perceived reading competence in PIRLS) and varied in the number of items measuring competence (from 4 in TIMSS Advanced to 9 in TIMSS). The content of the items is nevertheless similar (also as already mentioned due to the studies being conducted by the same international organization). The results are in line with research literature indicating self-perceived competence (self-efficacy) as one of the most prominent predictors of academic success. Self-efficacy is directly related to behaviour in academic tasks (e.g. the effort, persistence in the tasks despite obstacles 80 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... and challenges) (Maddux, 2009; Ragozzino et al., 2003, Motti-Stefanidi & Masten, 2013). High perceived self-efficacy for self-regulated learning contributes to better learning outcomes and also increases the likelihood of remaining in school (Caprara, Fida, Vecchione, Del Bove, Veccio, & Barbaranelli, 2008). Competence being the strongest predictor is in line also with initial research findings of Ryan and Deci (2002) that autonomy itself is not enough for gains in academic achievement - the competence takes (according to the authors of the framework) the leading role. Optimal functioning and high engagement (with active participation in learning) in school setting is both a result of high autonomy support and self-perceived competence. Practical implications when promoting competence at the school and classroom level involve mastery goal structure (Wang & Holcombre, 2010). Mastery goal structure promotes positive and effort based praise while avoiding pressuring students for correct answers or high grades (Wang & Holcombe, 2010) as opposed to performance goal structure. Mastery type school climate provides students with more opportunities to feel successful. Mastery goal structure also provides more opportunities for students to work together (support for relatedness) and not compete against each other. In order to foster student's sense of competence, school community collaboration can be of use as well. For instance different forms of project community based work can provide an example of students' knowledge being directly used and reinforce their perception of self-efficacy. These types of activities (mentoring and tutoring programs, contextual learning and job shadowing) have research support as well (Epstein et al., 2009). And also as shown in the experiment by Benware and Deci (1984) when students learn in order to use their knowledge, their knowledge is more conceptual. One example of this type of collaboration would be for instance project work on the agricultural planning of planting local green areas in which representatives of a local community would cooperate with biology teachers and students of a local school. The second most important predictor of academic achievement, out of included indicators of psychological needs fulfilment, is self-perceived autonomy. In 8th, 9th and 13 grade the predictive power of self-perceived autonomy is similar (8th grade) or even higher (9th and 13 grade) when compared to SES as a predictor of academic achievement. Autonomy is a building stone of intrinsic motivation. Students with a greater sense of autonomy in school have better school outcomes such as classroom engagement, persistence, enjoyment and achievement (Wang & Holcombe, 2010). 81 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Autonomy was measured differently in selected studies: in PIRLS and TIMSS the only items related to autonomy were indicated to perceived autonomy support at home (a lack of control over school work). In ICCS and in TIMSS Advanced studies the autonomy measure represented the perceived autonomy in classroom (autonomy support from teachers). The autonomy support is a significant predictor of student's achievement when the autonomy is supported at home or in the classroom. The perceived parental autonomy support was significantly related also to achievement in PISA study (math, science and reading literacy) (Rutar Leban, Vršnik Perše, Kozina, Pavlovic, 2009). Practical implications suggest various activities inside the classroom as well as in the form of school-community collaboration. For instance, given the choice and supporting autonomy in organizing and conducting project work fosters their sense of autonomy. As already mentioned in the introduction autonomy support from teachers (e.g. letting students choose from various alternatives, listening to them and asking them for their point of view) increases students intrinsic motivation (Ryan & Grolnick, 1986; Tsai et al., 2008) and have higher educational achievement (Vallerand, et al., 1997). Previously mentioned school community collaboration fosters autonomy as well. With students planning the whole project their autonomy would be supported. Relatedness was a significant predictor in TIMSS Advanced study (13th grade students) but not in other age groups. This could be due to developmental characteristics of the sample. Although the need to connect and belong is likely to be pervasive throughout one's life, research has suggested that during the period of adolescence the need to connect with others through mutually supportive relationships is at its peak (Wang & Hol-combe, 2010). Nevertheless since the 8th and 9th grade students are in the period of adolescence as well the results would need some extra elaboration in future research. Relatedness refers in its core to the strengths of ones connections to others. And a sense of connectedness to teachers and peers in school is associated with multiple indicators of academic motivation and engagement (Wang & Holcombe, 2010) indicating that the association between relatedness and achievement is indirect (through engagement) which results in nonsignificant prediction in most of the samples included in the analyses. Engagement is according to Wang and Holcombe (2010) composed of three interrelated dimensions: behavioural, emotional and cognitive engagement. When all three are addressed and supported, academic outcomes are increased. Relatedness supports emotional engagement. And emotional engagement is, according to re- 82 a. kozina, a. mlekuz ■ intrinsic motivation ... search review (Wang & Holcombe, 2010), not directly related to academic achievement. Even though our results have not wholly supported the importance of the relatedness in academic achievement we believe that it has to be addressed also as a support for autonomy and competence. In future research one focus could be in the elaboration in more detail of the relationship between all three predictors. Social support fosters feeling of social connectedness, which is required in order for children to internalize social standards (for instance value of education) and to develop respect for social institutions (also school) (Ellias & Hayes, 2008). With collaboration in and outside of school students build their own social network, social capital that is as an important well-being indicator as is material capital (Morenoff & Sampson, 2008). Even if a child or adolescent possesses the required skill for school success, motivation to use them is related to perception of social support for school related activities (from parents and community). Together with the research relevance of the findings (congruency with the SDT theoretical assumptions), the findings have high practical value as well. Individuals seek experiences that fulfil their need for competence, autonomy and relatedness through interaction with the environment (Deci & Ryan, 2002). As said the theory proposes that the degree to which students perceive that the school context meets these psychological needs determines the level of student's engagement in school. We can use school related activities as a source of activities that can foster student's self-perceived autonomy, competence and relatedness. Intrinsically motivated learning can be greatly influenced by social environments (Ryan & Deci, 2009). For instance, the key feature is autonomy supporting teacher's behaviour. When teachers support students' autonomy they achieve more, learn conceptually and stay in school longer (Reeve, 2002). An important emphasis has to be put also on teachers' motivation and the quality of their teaching. Various studies of elementary and high school students (e.g. Hardre & Reeve, 2003; Jang, Reeve and Deci, 2010) show that teachers' autonomy support is related to teachers' own autonomous motivation and later work engagement. As pointed out by Ryan and La Guardia (1999, in: Ryan & Deci, 2009) the importance of autonomy and competence support needs to be recognized in the school setting also in the light of addressing students at risk (e.g. low achieving students, students prone to early school leaving) even more since the first response of teachers and parents in situations of low achievement and anticipated early school leaving is adding more controls (e.g. scheduled time for studying, constant control over school 83 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 work...) and additional pressures to the students, which in a way closes the door for intervention and even lowers their motivation for learning and staying in school. Support for autonomy inside the classroom, inside school and also using community relevant activities, which would affect the relatedness, autonomy and competence as well - increase students' motivation for learning and continuing their education. As seen in our data the parental support (lack of controlling behaviour) for autonomy plays a vital part as well indicating our role in educating parents as parents meaning their role of autonomy support and of the developmental characteristics of their adolescents. Conclusion The findings support the importance of intrinsic motivation for academic achievement in various content domains as well as in various age groups. Even though the measures are rough and not directly comparable the findings show consistency. Since the results show somehow a different pattern in i3th grade, these results would benefit from further investigation of the role that all three psychological needs play in academic achievement in different developmental periods (with the same and comparable measure). The measures used are at this point rough - the international comparative studies are not designed to specifically measure perceived competence, autonomy and relatedness, therefore the findings should be understood as preliminary and as a starting point for future more in depth analyses with more consistent measures of all predictors as well as the use of more advanced statistical techniques (e.g. hierarchical regression models). References Alivernini, F., Lucidi, F. (20ii) Relationship between social context, self-efficacy, motivation, academic achievement, and intention to drop out of high school: A longitudinal study. The Journal of Educational Research 104, pp. 24^252. Arora, A., Foy, P., Martin, M. O. Mullis, I. V. S. (2009) TIMSS Advanced 2008 technical report. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Benware, C., Deci, E. L. ^984) Quality of learning with an active versus passive active set. American Educational Research Journal 21, pp. 755765. Caprara G. V., Fida, R., Vecchione, M., Del Bove, G., Veccio, G. M., Bar-baranelli, C. (2008) Longitudinal analyses of the role of perceived self-efficacy for self-regulated learning in academic continuance and achievement. Journal of Educational Psychology 100 (3), pp. 525-534. 84 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... Deci, E. L., Ryan, R. M. (2002) Handbook of self-determination research. Rochester: The University of Rochester Press. Ellias, M. J., Haynes, N. M. (2008) Social competence, social support, and academic achievement in minority, low income, urban, elementary school children. School Psychology Quarterly 23 (4), pp. 474-495. Epstein, J. L, Sanders, M. G., Sheldon, S. B., Simon, B. S., Clark Salinas, K., Rodrigey Jansorn, N., Van Voorhis, F. L., Martin, C. S., Thomas, B. G., Greenfeld, M. D., Hutchins, D. J., Williams, K. J. (2009) School, family and community partnership. London: Corwin Press. Flink, C., Boggiano, A. K., Main, D. S., Barrett, M. Katz, P. A. (1992) Children's achievement-related behaviors: The role of extrinsic and intrinsic motivational orientations. In Boggiano, A. K., Pittman, A. K. (eds.). Achievement and motivation: A social- develop- mental perspective. New York: Cambridge University Press, pp. 189-214. Fortier, M. S., Vallerand, R. J., Guay, F. (1995) Academic motivation and school performance: Toward a structural model. Contemporary Educational Psychology 20, pp. 257-274. Foy, P., Brossman, B., Galia, J. (2013) Scaling the TIMSS and PIRLS 2011 Achievement Data. In Martin, M. O., Mullis, I. V. S. (eds.). TIMSS and PIRLS methods and procedures. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College and International Association for the Evaluation of Educational Achievement, pp. 1-28. Grolnick, W. S., Ryan, R. M. (1987) Autonomy in children's learning: An experimental and individual difference investigation. Journal of Personality and Social Psychology 5, pp. 890-898 Guay, F., Vallerrand, V. J. (1997) Social context, students motivation and academic achievement: Toward a process model. Social Psychology of Education 1, pp. 211-233. Hardre, P. L., Reeve, J. (2003) A motivational model of rural students' intentions to persist in, versus drop out of, high school. Journal ofEdu-cationalPsychology 95 (2), pp. 347-356. Jang, H., Reeve, J., Deci, E. L. (2010) Engaging students in learning activities: It's not autonomy support or structure, but autonomy support and structure. Journal of Educational Psychology 102, pp. 588-600. Joncas, M., Foy, P. (2013) Sampling Design in TIMSS and PIRLS. In Martin, M. O., Mullis, I. V. S. (eds.). Methods and Procedures in TIMSS and PIRLS 2011. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College and International Association for the Evaluation of Educational Achievement, pp. 1-21. 85 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Kage, M., Namiki, H. (1990).The effects of evaluation structure on children's intrinsic motivation and learning. Japanese Journal of Educational Psychology38, pp. 36-45. Kozina, A., Rožman, M., Vršnik Perše, T. Rutar Leban, T. (2012) Napov-edna vrednost različnih ocen šolske klime za dosežke v raziskavah TIMSS [The predictive power of different school cliate evaluations in the TIMSS studies]. Didactica Slovenica 27 (1/2), pp. 127-144. Legault, L., Green-Demers, I., Pelletier, L. G. (2006) Why do high school students lack motivation in the classroom? Toward an understanding of academic amotivation and social support. Journal of Educational Psychology98, pp. 567-582. Maddux, J. E. (2009) Self-Efficacy: The power of believing you can. In Lopez, S. J., Snyder, C. R. (eds.). The Oxford handbook of positive psychology (2nd ed.). New York: Oxford university press, pp. 335-345. Martin, M. O., Mullis, I. V. S. (2013a) TIMSS and PIRLS methods and procedures. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College and International Association for the Evaluation of Educational Achievement. Martin, M. O., Mullis, I. V. S. (2013b) TIMSS and PIRLS 2011: Relationships Among Reading, Mathematics, and Science Achievement at the Fourth Grade - Implications for Early Learning Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College and International Association for the Evaluation of Educational Achievement. Miserandino, M. (1996) Children who do well in school: Individual differences in perceived competence and autonomy in above average children. Journal of Educational Psychology 88, pp. 203-214. Morenoff, J. D., Sampson, R. J. (2008) Constructing community indicators of child well-being. Brown, B. V. (ed.). Key indicators ofchildand youth well-being. New York, London: Psychology Press, pp. 307-331. Motti-Stefanidi, F., Masten, A. S. (2013) School success and school engagement of immigrant children and adolescents. European Psychologist 18 (2), pp. 126-135. Mullis, I. V. S., Drucker, K. T., Preuschoff, C., Arora, A., Stanco, G. M. (2013) Assessment Framework and instrument development. In Martin, M. O., Mullis, I. V. S. (eds.). TIMSS and PIRLS methods and procedures. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College and International Association for the Evaluation of Educational Achievement, pp. 1-22. 86 A. KOZINA, A. MLEKUZ ■ INTRINSIC MOTIVATION ... Mullis, I. V. S., Martin, M. O., Foy, P., Arora, A. (2012a) TIMSS 2011 International Results in Mathematics. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Mullis, I. V. S., Martin, M. O., Foy, P., Drucker, K. T. (2012b) PIRLS 2011 International Results in Reading. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Mullis, I. V. S., Martin, M. O., Kennedy, A. M., Trong, K. L., Sainsbury, M. (2009a) PIRLS 2011 Assessment Framework. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Mullis, I. V. S., Martin, M. O., Robitaille, D. F., Foy, P. (2009b) TIMSS Advanced 2008 International Report: Findings from IEA's Study of Achievement in Advanced Mathematics and Physics in the Final Year of Secondary School. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Mullis, I. V. S., Martin, M. O., Ruddock, G. J., O'Sullivan, C. Y., Preu-schoff, C. (2009c) TIMSS 2011 Assessment Framework. Chestnut Hill, MA: TIMSS PIRLS International Study Center, Boston College. Ragozzino, K., Resnik, H., Utne-O'Brien, M., Weissberg, R. P. (2003) Promoting academic achievement through social and emotional learning. Educational Horizons 81, pp. 169-171. Reeve, J. (2002) Self-determination theory applied to educational settings. In Deci, E. L., Ryan, R. M. (eds.). Handbook of self-determination research. Rochester, NY: University of Rochester Press, pp. 183-203. Reeve, J. (2015) Understanding motivation and emotion. Hoboken: Wiley. Rutar Leban, T., Vršnik Perše, T., Kozina, A., Pavlovic, Z. (2009) Elementi vzgojnega sloga staršev in dosežki mladostnikov [The elements of parental styles and students' academic achivement]. Šolsko polje 20 (3/4), pp. 87-103. Ryan, R. M., Deci, E. L. (2000) Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being. American Psychologist55, pp. 68-78. Ryan, R. M., Deci, E. L. (2002) An overview of self-determination theory: An organismic-dialectical perspective. In Deci, E. L., Ryan, R. M. (eds.). Handbook of self-determination research. Rochester, NY: University of Rochester Press, pp. 3-37. Ryan, R. M. (1995) Psychological need and the facilitation of integrative processes. Journal of Personality 63, pp. 397-427. Ryan, R. M., Deci, E. L. (2009) Promoting self-determined school engagement - motivation, learning and wellbeing. In Wentzel, K. R., Wig- 87 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 field, A. (eds.). Handbook of motivation at school. New York: Routh-ledge, pp. 171-196. Ryan, R. M., Grolnick, W. S. 1986) Origins and pawns in the classroom: Self-report and projective assessments of individual differences in children's perceptions. Journal of Personality and Social Psychology S0, pp. 550-558. Ryan, R. M., La Guardia, J. G. (1999) Achievement motivation within a pressured society: Intrinsic and extrinsic motivations to learn and the politics of school reform. In Urdan, T. (ed.). Advances in motivation and achievement. Greenwich, CT: JAI Press, pp. 45-85. Schulz, W., Ainley, J., Fraillon, J. (2011) ICCS 2009 Technical Report. Amsterdam: International Association for the Evaluation of Educational Achievement. Sheldon, K. M., Elliot, A. J. (1999) Goal striving, need-satisfaction, and longitudinal wellbeing: The self-concordance model. Journal of Personality and Social Psychology 76, pp. 482-497. Skinner, E. A., Belmont, M. J. (1993) Motivation in the classroom: Reciprocal effects of teacher behavior and student engagement across the school year. Journal of Educational Psychology 85 (4), pp. 571-581. Tsai, Y., Kunter, M., Ludtke, O., Trautwein, U., Ryan, R. M. (2008) What makes lessons interesting? The role of situational and individual factors in three school subjects. Journal of Educational Psychology 100, pp. 460-472. Vallerand, R. J., Fortier, M. S., Guay, F. (1997) Self-determination and persistence in a real life setting: Toward a motivational model of high school dropout. Journal of Personality and Social Psychology 72 (5), pp. 1162-1176. Wang, M. T. Holocombe, R. (2010) Adolescents' perceptions of school environment, engagement, and academic achievement in middle school. American Educational Research Journal 47 (3), pp. 633-662. 88 Samopodoba in socialna sprejetost identificiranih nadarjenih in visoko učno uspešnih osnovnošolcev Urška Aram, Nina Jurinec, Marina Horvat in Katja Košir Veliko je raziskav (npr. Lopez in Sotillo, 2009; Luftig in Nichols, 1990), ki preučujejo socialno-emocionalne razlike med nadarjenimi in ostalimi učenci v razredu; manj pa je raziskav, ki bi se ukvarjale z razliko na omenjenih področjih med nadarjenimi in visoko učno uspešnimi učenci. Zaradi različnih opredelitev nadarjenosti v različnih regionalnih in nacionalnih kontekstih (glej npr. Carman, 2013) raziskave ne dajejo konsistentnih rezultatov. V slovenskem okolju raziskav, ki bi se ukvarjale s podobnim vprašanjem, nismo zaznali, zato nas je v pričujoči raziskavi najprej zanimalo, 1) ali na vzorcu slovenskih učencev obstajajo razlike v socialni sprejetosti in samopo-dobi med učno uspešnimi (nadarjenimi in visoko učno uspešnimi) ter ostalimi učenci, 2) in nato še bolj podrobno, ali obstajajo razlike v socialni spreje-tosti in samopodobi med identificiranimi nadarjenimi učenci in tistimi, ki so jim po ocenah v šoli najbolj podobni - visoko učno uspešnimi učenci, ki niso bili identificirani kot nadarjeni, ter 3) ali med zgoraj navedenimi skupinami učencev, razdeljenimi glede na nadarjenost in učno uspešnost, obstajajo razlike med spoloma v samopodobi in socialni sprejetosti. V nadaljevanju so podrobneje predstavljeni izsledki raziskav, ki so preučevale omenjene pojave pri nadarjenih in visoko učno uspešnih učencih oziroma pri nadarjenih učencih in učencih, ki niso bili prepoznani kot nadarjeni. Nadarjeni učenci Starejše opredelitve operacionalizirajo nadarjenost zelo ozko in nadarjene učence največkrat enačijo z visoko inteligentnimi učenci (Carman, 2013). Novejše definicije pa poudarjajo, da so nadarjeni učenci zelo heterogena 89 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 skupina (Robinson, 2002), saj se lahko nadarjenost oziroma talent kaže na različnih področjih (Olszewski-Kubilius, Subotnik in Worrell, 2015). Ena bolj uveljavljenih definicij je Renzullijeva (1977) opredelitev nadarjenosti, ki nadarjenost opisuje kot soobstoj visokih sposobnosti, ustvarjalnosti in motivacije. Renzullijev trikrožni model predstavlja v Sloveniji osnovo za izvajanje postopka odkrivanja nadarjenih učencev v osnovni šoli (Juriševič, 2009). Zaradi neenotnih opredelitev nadarjenosti so merila za nadarjenost odvisna od nacionalnih in/ali šolskih kontekstov, kar pomeni, da se lahko razlikujejo od države do države in celo od šole do šole. Pomanjkanje enotne definicije nadarjenosti oziroma enotnih meril za identifikacijo nadarjenih učencev tako ne vodi samo v različne operacionalizacije (Carman, 2013), ampak tudi v razlike v deležu identificiranih nadarjenih učencev v različnih regionalnih in nacionalnih kontekstih, posledično pa onemogoča neposredno primerjavo nadarjenih. Ker je pri nas definicija nadarjenosti precej široko zastavljena, predstavlja Slovenija eno od držav z največjim deležem nadarjenih učencev (Juriševič, 2012). Prepoznavanje nadarjenih učencev v Sloveniji V Sloveniji poteka odkrivanje nadarjenih učencev v treh korakih (Koncept: Odkrivanje in delo z nadarjenimi učenci v devetletni osnovni šoli, 1999): 1) evidentiranje učencev; 2) identificiranje nadarjenih učencev, kjer mora posameznik, po slovenskem modelu, izpolnjevati vsaj eno od naslednjih treh meril: - učiteljeva ocena učenčeve uspešnosti: učitelj s pomočjo ocenjevalne lestvice oceni delovanje učenca na različnih področjih; - kognitivne sposobnosti: nadpovprečni rezultati na testu inteligentnosti; - ustvarjalnost: nadpovprečni rezultati na Torrancovih testih ustvarjalnega mišljenja; 3) seznanitev in mnenje staršev. Postopek identifikacije se običajno izvaja v četrtem razredu, učenca pa se lahko evidentira tudi v višjih razredih. Pri uporabi takšnega koncepta je v Sloveniji 26 % učencev identificiranih kot nadarjenih (Juriševič, 2012). Visoko učno uspešni učenci Visoko učno uspešni učenci (angl. high achievers) so učenci, ki v šoli dobivajo visoke ocene. Običajno so dobro organizirani in znajo upravljati s časom. Na podlagi tega opravijo zahtevano delo in naloge pravočasno ter zelo 90 u. aram, n. jurinec, m. horvat, k. kosir ■ samopodoba in socialna sprejetost ... dobro. Za visoko učno uspešne učence je značilno tudi, da se pretežno ustrezno vedejo, se dobro prilagajajo šolskemu okolju, v učnih razpravah pa sodelujejo z navdušenjem. Visoko učno uspešni učenci niso nujno tudi nadarjeni učenci (Bainbridge, 2016). Podobno, kot velja za nadarjenost, je v raziskavah različno opredeljena tudi visoka učna uspešnost. Bain in Bell (2004) sta visoko učno uspešne učence določili na podlagi testa (Comprehensive Test of Basic Skills; McGraw-Hill, 1996), Ritchotte in sodelavci (2016) pa so med visoko uspešne učence uvrstili tiste, katerih povprečna ocena je bila enaka ali višja od 3,5. Socialne in čustvene značilnosti nadarjenih učencev Ugotovitve raziskav, ki so preučevale socialne in čustvene značilnosti nadarjenih učencev, niso enotne. Nekateri avtorji predpostavljajo, da so nadarjeni učenci psihično šibki in ranljivi oziroma da predstavlja oznaka učenca kot nadarjenega dejavnik tveganja (Neihart, 1999; Olszewski-Ku-bilius et al., 2015). Obenem so nekateri raziskovalci (Terman, 1925; Ter-man in Oden, 1947, 1959) že pred desetletji ugotovili, da so nadarjeni učenci uspešnejši od svojih vrstnikov ne samo na kognitivnem, temveč tudi na socioemocionalnem področju. Slednje so podprle številne študije (npr. Chan, 2010; Cross et al., 2004; Garland in Zigler, 1999; LoCicero in Ashby, 2000; Lee et al., 2012; Lopez in Sotillo, 2009; Mueller, 2009), kljub temu pa nekatere ugotovitve raziskav nakazujejo, da lahko predstavlja identificirana nadarjenost dejavnik tveganja za težave v psihološki prilagojenosti učenca (npr. Coleman in Cross, 1988; Cross et al., 1991; Cross, 1997; Pfeiffer, 2009). Neihart (1999, 2002) navaja prej omenjene nasprotujoče si ugotovitve o socialnem in čustvenem razvoju nadarjenih učencev ter jih povzema kot dva različna pristopa k razumevanju socioemocionalnih značilnosti nadarjenih: pristop odpornosti in pristop tveganja (angl. resilience and risk approach): 1) Pristop odpornosti pojmuje nadarjenost kot varovalni dejavnik: kljub morebitnemu doživljanju stresorjev iz okolja imajo nadarjeni učenci vrsto notranjih virov, ki jim omogočajo uspešno spoprijemanje s stresnimi dogodki. 2) V nasprotju s tem pa pristop tveganja predpostavlja, da oznaka učenca kot nadarjenega lahko predstavlja dejavnik tveganja. Tako na primer Freeman (2006) izpostavlja, da večja družbena pričakovanja in pritisk, povezan z učno uspešnostjo pri identificiranih nadarjenih učencih, lahko vodijo v doživljanje čustvene stiske. Znotraj tega pristopa avtorji predpostavljajo, da nekatere značilnosti nadar- 91 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 jenih učencev povečajo tveganje za težave v socialnem prilagajanju; asinhroni razvoj (Silverman, 2002) in višja čustvena intenzivnost ter občutljivost so najpogosteje omenjeni elementi, ki naj bi izpostavljali nadarjene učence višji stopnji stresa. Samopodoba nadarjenih in visoko učno uspešnih učencev Samopodoba je vse tisto, kar si posameznik misli o sebi, da je, in vse to, kar si želi pokazati, da je. Zajema naše predstave, zaznave, misli, pojmovanja in prepričanja o sebi, čustva do sebe ter vrednotenje samega sebe (Musek, 2005). Raziskav, ki bi se osredotočale izključno na primerjavo nadarjenih in visoko učno uspešnih učencev, ki niso bili identificirani kot nadarjeni, ni veliko. Bain in Bell (2004) sta primerjali identificirane nadarjene in visoko učno uspešne učence, ki niso bili identificirani kot nadarjeni, na štirih dimenzijah samopodobe: telesne sposobnosti, zunanji videz, odnosi z vrstniki in odnosi s starši. Ugotovili sta, da so učenci, ki so bili identificirani kot nadarjeni, dosegli višje rezultate od učencev, ki so visoko učno uspešni, a niso bili identificirani kot nadarjeni, na treh od štirih v raziskavo vključenih dimenzijah samopodobe: telesna sposobnost, zunanji videz in odnosi z vrstniki. Prav tako so nadarjeni učenci dosegli višji rezultat na lestvici splošne samopodobe, medtem ko na področju odnosov s starši statistično pomembnih razlik ni bilo. Poleg tega sta avtorici ugotovili, da so se nadarjeni učenci v primerjavi z učno uspešnimi učenci bolj nagibali k temu, da so svoj socialni uspeh, se pravi uspešnost v socialnih situacijah, prej pripisovali svojemu trudu in svojim sposobnostim kot pa sreči ali težavnosti situacije. Medtem ko sta se Bain in Bell (2004) pri raziskovanju osredotočili bolj na telesni in socialni vidik, so Ritchotte in sodelavci (2016) raziskovali učno samopodobo. Ugotovili so, da je učna samopodoba visoko učno uspešnih učencev, ki niso bili identificirani kot nadarjeni, primerljiva z učno samopodobo identificiranih nadarjenih učencev. Poleg tega so ugotovili, da je povprečna ocena visoko učno uspešnih učencev primerljiva s povprečno oceno identificiranih nadarjenih učencev. V nadaljevanju predstavljamo še nekaj ugotovitev iz raziskav, ki so se osredotočale na primerjavo samopodobe nadarjenih in nenadarjenih učencev. Metaanaliza, ki sta jo opravila Hoge in Renzulli (1993), je pokazala, da se nadarjeni učenci v primerjavi z vrstniki, ki niso bili prepoznani kot nadarjeni, zaznavajo kot uspešnejši na splošnem, učnem in vedenjskem področju. Na področju socialne samopodobe in samopodobe, povezane s telesnimi sposobnostmi, ni bilo statistično pomembnih razlik med nadarjenimi in nenadarjenimi učenci. Metaanaliza izpred nekaj let, ki sta 92 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... jo opravila Litster in Roberts (2011), pa je pokazala, da je največja razlika med nadarjenimi in nenadarjenimi učenci v učni samopodobi. To pomeni, da nadarjeni učenci dojemajo svoje učne sposobnosti višje kot njihovi vrstniki, ki niso bili identificirani kot nadarjeni. Nadarjeni učenci se zaznavajo tudi kot vedenjsko kompetentnejši, kar pomeni, da se v večji meri dojemajo kot osebe z lepim oziroma ustreznim vedenjem, poleg tega pa imajo tudi višjo splošno samopodobo. Na drugi strani pa so nadarjeni učenci nižje ocenili svoj zunanji videz in telesne sposobnosti. V zaznani socialni kompetentnosti ni bilo razlik med nadarjenimi in nenadarjenimi učenci. Litster in Roberts (2011) sta preverjala tudi, ali sta spol in starost moderatorja povezave med nadarjenostjo in različnimi področji zaznane kompetentnosti. Ugotovila sta, da so razlike med nadarjenimi in nenadarjenimi učenci v zaznanih učnih sposobnostih večje pri starejših učencih in pri dekletih. Vlogo spola je raziskovala tudi Ablard (1997), ki je ugotovila, da imajo nadarjena dekleta na učnem področju nižjo socialno samopodobo v primerjavi z nadarjenimi fanti. Preckel et al. (2008) pa so ugotovili razlike med nadarjenimi in povprečnimi učenci v učni samopodobi na področju matematike. Razlike so se pokazale le pri fantih, in sicer v prid nadarjenim. Na drugi strani pa obstajajo tudi študije, ki niso našle nobenih razlik med spoloma za področje učne, socialne in čustvene samopodobe pri nadarjenih učencih (npr. Bain in Bell, 2004; Chan, 2001; Cunningham in Rinn, 2007; Lee et al., 2012). Socialna sprejetost nadarjenih in visoko učno uspešnih učencev Odnos med socialno sprejetostjo in učno uspešnostjo učencev je bil v preteklosti pogosto preučevano področje: priljubljeni učenci so praviloma učno uspešnejši (npr. Hatzichtistou in Hopf, 1996; Košir et al., 2007; Wentzel, 1991, 1993; Wentzel in Asher, 1995), učenci z nižjim socio-metričnim statusom, predvsem zavrnjeni učenci, pa predstavljajo skupino z večjim tveganjem za učne težave in izpad iz šolskega sistema (Freder-ickson in Furnham, 2001; Hatzichtistou in Hopf, 1996; Ollendick et al., 1992). Socialna sprejetost učencev je bila v raziskavah operacionalizirana na različne načine. Bain in Bell (2004) sta socialno sprejetost učencev ocenjevali kot s strani učitelja zaznano socialno sprejetost posameznega učenca. Primerjali sta identificirane nadarjene učence in visoko učno uspešne učence, ki niso bili identificirani kot nadarjeni. Med nadarjenimi in visoko učno uspešnimi učenci nista ugotovili razlik v socialni sprejetosti. Pokazala pa se je statistično pomembna razlika v zaznani socialni sprejetosti 93 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 s strani učitelja glede na spol, in sicer so učitelji višje ocenili socialno spre-jetost fantov. V nadaljevanju navajamo še nekaj ugotovitev iz raziskav, ki so primerjale socialno sprejetost identificiranih nadarjenih učencev in učencev, ki niso bili identificirani kot nadarjeni. Lopez in Sotillo (2009) sta raziskovali socialno sprejetost nadarjenih učencev in učencev, ki niso bili prepoznani kot nadarjeni; za oceno socialne sprejetosti sta uporabili socio-metrično preizkušnjo in učiteljevo oceno socialne sprejetosti. Rezultati niso pokazali razlik med skupinama nadarjenih in nenadarjenih v priljubljenosti, zavrnjenosti, socialnem vplivu in socialni preferenčnosti. Poleg tega ni bilo razlik med obema skupinama v umestitvi učencev v socio-metrične skupine. Prav tako se učenci, ki so bili prepoznani kot nadarjeni, od svojih vrstnikov niso razlikovali v učiteljevi oceni socialne sprejetosti. O učinkovitejšem delovanju nadarjenih učencev v socialnih odnosih pa poročajo Cohen et al. (1994). Ugotovili so, da so bili učenci bolj sprejeti s strani svojih vrstnikov, pokazali so večje zavedanje vzajemnosti v prijateljskih odnosih in s strani vrstnikov so bili zaznani kot manj pogosto vključeni v agresivno vedenje ali manj pogosto označeni kot žrtve takšnega vedenja. Niso pa se pokazale razlike med obema skupinama v številu prijateljskih odnosov. Podobno je Peairs (2010) ugotovila, da imajo nadarjeni učenci višjo socialno preferenčnost (tj. razliko med standardiziranimi pozitivnimi in negativnimi sociometričnimi izbirami) in so pogosteje označeni kot priljubljeni v primerjavi z nenadarjenimi učenci. Prav tako je ugotovila, da je za nadarjene učence verjetneje, da so prijatelji in pripadniki vrstniških klik z nadarjenimi posamezniki. Poleg tega so bili s strani učiteljev zaznani kot socialno spretnejši kot njihovi nenadarjeni vrstniki. Tako se zdi, da so tudi medsebojni odnosi nadarjenih učencev primerljivi ali celo boljši v primerjavi z njihovimi vrstniki, ki niso bili identificirani kot nadarjeni. Rimm (2002) ob tem poudarja, da lahko nadarjeni učenci kljub s strani vrstnikov zaznani višji socialni sprejetosti poročajo o drugačnih občutjih; ne glede na objektivne pokazatelje njihove socialne sprejetosti lahko ti učenci doživljajo svojo nadarjenost kot negativni dejavnik za svoje medosebne odnose (Kerr et al., 1988). Poleg tega je Peairs (2010) ugotovila, da so negativne posledice zavrnitve izrazitejše pri nadarjenih učencih v primerjavi z njihovimi zavrnjenimi nenadarjenimi vrstniki. Nekatere raziskave tudi nakazujejo, da je spol moderator odnosa med nadarjenostjo in socialno sprejetostjo. Luftig in Nichols (1990) sta med raziskovanjem socialne sprejetosti med štirimi skupinami, ki so bile oblikovane na podlagi kognitivnih sposobnosti in spola, ugotovila, da so nadarjeni fantje najbolj priljubljeni, sledili so nenadarjeni fantje in nena- 94 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... darjena dekleta. Nadarjena dekleta so predstavljala najmanj priljubljeno skupino izmed omenjenih štirih. Ta ugotovitev nakazuje, da lahko nadarjenost vendarle predstavlja dejavnik tveganja za socioemocionalne izide pri dekletih. Poleg tega nekatere raziskave kažejo, da nadarjena dekleta v večji meri zanikajo svojo nadarjenost v primerjavi s fanti (Swiatek, 2001; Swiatek in Dorr, 1998) in pripisujejo večjo vrednost sprejemanju s strani vrstnikov (Chan, 2003, 2004). Nekatere raziskave pa na tem področju niso ugotovile razlik med spoloma (npr. Foust et al., 2006). Izhajajoč iz navedenih nekonsistentnih ugotovitev o razlikah v sa-mopodobi in socialni sprejetosti med prepoznanimi nadarjenimi učenci in učno uspešnimi učenci, ki niso bili prepoznani kot nadarjeni, je namen naše raziskave preveriti razlike v splošni, učni in samopodobi na področju odnosov z vrstniki ter socialni sprejetosti med nadarjenimi, visoko učno uspešnimi in ostalimi učenci v razredu. Dodatno želimo preveriti še, ali je spol dejavnik teh razlik znotraj omenjenih skupin. Na koncu nas zanima tudi, ali rezultati naše raziskave govorijo v prid hipotezi, ki predpostavlja večjo psihološko odpornost nadarjenih učencev (angl. »resilience hypothesis«), ali v prid hipotezi, ki nadarjenost opredeljuje kot dejavnik tveganja za težave v socialnem in emocionalnem razvoju nadarjenih učencev (angl. »risk hypothesis«). Metoda Udeleženci V raziskavi je sodelovalo 422 učencev, od tega 198 fantov (47,6 %), ter 25 njihovih učiteljev. Rezultati so bili zbrani v 25 oddelkih petih priložnostno izbranih slovenskih osnovnih šol. Učenci so bili stari od 11 do 15 let (M = 13,15, SD = 1,15). Vključeni so bili učenci od šestega do devetega razreda; 34 (8,1 %) jih je bilo iz šestega, 167 (39,8 %) iz sedmega, 87 (20,7 %) iz osmega in 132 (31,4 %) iz devetega razreda. 85 sodelujočih učencev je bilo identificiranih kot nadarjenih1 (20,1 %) in 117 učno uspešnih (27,7 %) učencev, ki niso bili prepoznani kot nadarjeni, njihova lanska ocena pri matematiki, slovenščini in tujem jeziku pa ni bila nižja od prav dobre. V skupini ostalih učencev, ki ne izpolnjujejo zgornjih dveh pogojev, je bilo 220 učencev (52,1 %). 173 učencev (41 %) je bilo iz šole v osrednjem delu Slovenije, preostali učenci, vključeni v raziskavo, so prihajali iz štirih osnovnih šol iz po-dravske regije. Pripomočki Sociometrična preizkušnja. Za pridobitev podatkov o socialni sprejetosti učencev v razredu je bila uporabljena sociometrična preizkušnja s pozi- 1 V skladu z dokumentom Koncept: Odkrivanje in delo z nadarjenimi učenci v devetletni osnovni šoli (1999). 95 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 tivnim (»Navedi tri sošolce, s katerimi se najraje družiš.«) in z negativnim kriterijem (»Navedi tri sošolce, s katerimi se najmanj rad/a družiš.«), pri čemer so bili učenci opozorjeni, da jim pri negativnem kriteriju ni treba navesti treh učencev. Po pristopu standardiziranih dosežkov Coiea, Dod-ga in Coppotellija (1982) je bila za učence v razredu na podlagi rezultatov sociometrične preizkušnje določena mera socialnepreferenčnosti oziroma všečnosti, ki je glavna mera socialne sprejetosti učenca. Definirana je kot razlika med standardiziranimi pozitivnimi in negativnimi izbirami. Učiteljeva ocena socialne sprejetosti učencev. Učitelji (običajno razredniki) so na štiristopenjski lestvici (1 - učenca sošolci sploh ne sprejemajo, 2 - učenca sošolci večinoma ne sprejemajo, 3 - učenca sošolci večinoma sprejemajo, 4 - učenca sošolci dobro sprejemajo) ocenili učenčevo stopnjo sprejetosti s strani sošolcev. Vprašalnik samopodobe - Self Description Questionnaire (SDQ-II). Vprašalnik samopodobe SDQ-II je uveljavljena in pogosto uporabljena mera samopodobe za mladostnike; osnovana je na modelu samopodobe Shavelsona, Hubnerjeve in Stantona (1976). Za potrebe raziskave so bile za pridobitev podatkov o samopodo-bi mladostnikov uporabljene tri dimenzije iz Vprašalnika samopodobe (Self-description questionnaire II, Marsh,i992): 1) samopodoba na področju odnosov z vrstniki (»Imam mnogo prijateljev.«), 2) učna samopodoba (»Pri večini šolskih predmetov se hitro učim.«) in 3) splošna samopodoba (»Nič, kar naredim, se ne posreči.«). Dimenzija samopodobe na področju odnosa z vrstniki meri samozaznano priljubljenost med vrstniki, zmožnost sklepanja prijateljstev in zaznano kvaliteto teh prijateljstev. Notranja zanesljivost lestvice je visoka (Cronbach a znaša od 0,84 do 0,86; Marsh, Parada in Ayotte, 2004; Cronbach a za naš vzorec 0,78). Dimenzija učne samopodobe meri mladostnikovo samozaznavo znanj, zmožnosti in interesa za šolsko delo na splošno. Notranja zanesljivost lestvice je visoka in znaša od Cronbach a = 0,88 do a = 0,90 (Marsh et al., 2004; Cronbach a na našem vzorcu 0,81). Dimenzija splošne samopodobe preverja mladostnikovo zaznavo sebe kot sposobnega in učinkovitega posameznika, ki je zadovoljen in ponosen nase. Notranja zanesljivost te lestvice se giblje od a = 0,82 do a = 0,85 (Marsh et al., 2004) ter a = 0,76 na našem vzorcu. Vse tri dimenzije vsebujejo po deset postavk, polovica postavk na vsaki dimenziji se vrednoti obrnjeno. Učenci so na 4-stopenjski lestvici (1 - nikoli ne velja, 2 - včasih velja, 3 - skoraj vedno velja in 4 -vedno velja) označili, v kolikšni meri navedena trditev drži zanje. Šolska uspešnost. Učenci so zapisali zaključene ocene treh predmetov (slovenščina, matematika in prvi tuji jezik) za preteklo šolsko leto. Ocene so bile seštete in oblikovane v mero šolske uspešnosti za vsakega učenca. 96 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... Nadarjenost in učna uspešnost. Podatke o identificiranih nadarjenih učencih smo pridobili od šolske svetovalne službe. V skupino visoko učno uspešnih učencev smo vključili vse učence, ki niso bili identificirani kot nadarjeni učenci, njihova zaključena ocena v preteklem šolskem letu pa ni bila nižja od prav dobre pri nobenem od predmetov (slovenščina, matematika, tuji jezik). Tretjo skupino - ostali učenci - pa sestavljajo tisti, ki niso identificirani niti kot nadarjeni niti kot visoko učno uspešni. Postopek Podatki so bili zbrani v šolskem letu 2013/2014. Sodelovali so le učenci, katerih starši so s podpisom potrdili, da soglašajo s sodelovanjem. Da bi zagotovili veljavnost podatkov, pridobljenih s sociometrično preizkušnjo, je bila raziskava izvedena v razredih, kjer je bilo vrnjenih več kot 70 % soglasij. Ker sociometrična preizkušnja ni anonimna mera, so učenci zaradi varovanja podatkov dobili šifre, ki so bile sestavljene iz prvih dveh črk imen/-a in prvih dveh črk priimka/-ov. Reševanje vprašalnika je potekalo skupinsko v razredu, učenci niso bili časovno omejeni, večina pa je vprašalnik rešila v 15 minutah. Medtem ko so učenci reševali vprašalnik, smo za oceno socialne sprejetosti vseh učencev, vključenih v raziskavo, prosili njihove učitelje. Podatke o nadarjenosti učencev smo pridobili od šolske svetovalne službe. Analiza podatkov Pozitivne in negativne nominacije sociometrične preizkušnje in učiteljeva ocena socialne sprejetosti učencev so bile standardizirane znotraj vsakega razreda. Analizo učinka nadarjenosti (nadarjeni, učno uspešni, ostali učenci) za različne mere socialne sprejetosti in lestvice splošne, socialne in učne samopodobe smo izvedli z multivariatno analizo variance. Nadaljnja analiza učinka skupine (nadarjeni, visoko učno uspešni, ostali učenci) in spola ter interakcij skupina x spol je bila opravljena z dvosmerno analizo variance. Post hoc analize za glavne učinke so bile opravljene z Bonffer-onijevim testom, analize interakcij pa s parnimi primerjavami (angl. simple effects). 97 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Rezultati Osnovne deskriptivne statistike Tabela 1: Mere opisne statistike in Pearsonov koeficient korelacije za uporabljene spremenljivke. M SD 1 2 3 4 5 6 7 8 1 SDQSplošna 28,96 5,43 -,08 -,18** 1 2 SDQSocialna 2979 5,86 -,09 ,00 ,69** 1 3 SDQUčna 28,07 6,03 -,01 -,38** 75** ,52** 1 4 Soc. pref. 0,00 1,64 ,04 -,18** ,01 ,14** ,06 1 5 Učit. ocena 3,34 0,62 ,03 -,15** ,10 ,I3* J5** ,42** 1 6 Ocene skupno 11,61 2,67 J5** -,75** ,24** ,01 43** ,24** ,23** 1 Opombe: SDQSplošna = splošna samopodoba; SDQSocialna = socialna samopodoba; SDQUčna = učna samopodoba; Učit. ocena = učiteljeva ocena socialne sprejetosti učenca; Ocene skupno = seštevek ocen treh predmetov v lanskem šolskem letu; * p < 0,05, ** p < 0,01. V Tabeli 1 so prikazane mere opisne statistike in Pearsonov koeficient korelacije za vse tri preverjane dimenzije samopodobe (učna, splošna in samopodoba na področju odnosov z vrstniki), socialno preferenčnost skupaj z učiteljevo oceno socialne sprejetosti in seštevek ocen pri treh predmetih (slovenščina, matematika in prvi tuj jezik). Splošna, učna in samopodo-ba na področju odnosov z vrstniki med seboj statistično pomembno in visoko pozitivno korelirajo, kar kaže na dobro povezanost poddimenzij konstrukta samopodobe. Nizka, vendar statistično pomembna pozitivna povezava se je pokazala med splošno samopodobo in povprečno oceno v preteklem šolskem letu. Socialna samopodoba se statistično pomembno pozitivno, vendar nizko povezuje z obema merama socialne sprejetosti -s socialno preferenčnostjo in učiteljevo oceno socialne sprejetosti učenca. Učna samopodoba nizko pozitivno, vendar statistično pomembno korelira z učiteljevo oceno socialne sprejetosti učenca; s seštevkom ocen v preteklem šolskem letu pa je srednje visoko statistično pomembno pozitivno povezana. Meri socialne sprejetosti sta med seboj srednje visoko pozitivno in statistično pomembno povezani, s seštevkom ocen iz preteklega leta pa zmerno in statistično pomembno pozitivno. Samopodoba nadarjenih, visoko učno uspešnih in ostalih učencev Multivariatna analiza variance splošne, učne in samopodobe na področju odnosov z vrstniki kot odvisnih spremenljivk ter spola in skupin učne uspešnosti kot neodvisnih spremenljivk je pokazala statistično pomemben učinek spola, V = 0,021, F(3, 408) = 2,97, p = 0,032, parcialna yf = 0,021 in skupine (nadarjen, učno uspešen, ostali), V = 0,205, F(6, 818) = 15,587, 98 u. aram, n. jurinec, m. horvat, k. kosir ■ samopodoba in socialna sprejetost ... p < 0,001, parcialna vf = 0,103. Učinek interakcije spol x skupina pa ni bil statistično pomemben, V = 0,016, F(6, 818) = 1,115, p = 0,351, parcialna n2 = 0,008. V nadaljevanju sledijo analize z dvosmernimi analizami varianc, kjer smo preverjali učinek spola, skupine učne uspešnosti in interakcije med njima za vsako izmed treh lestvic samopodobe posebej - za učno, splošno in samopodobo na področju odnosov z vrstniki. Samopodoba na področju odnosov z vrstniki Za lestvico samopodobe na področju odnosov z vrstniki smo ugotovili statistično pomemben učinek spola, F(i, 410) = 4,50, p = 0,03, parcialna yf = 0,011, ki nakazuje, da so fantje v povprečju dosegali višje rezultate na lestvici samopodobe na področju odnosov z vrstniki, medtem ko učinek skupine ni bil statistično pomemben, F(i, 410) = 0,01,p = 0,91. Pokazal se je statistično pomemben učinek interakcije med spolom in skupino, F(2, 410) = 3,08, p = 0,047, parcialna ^ = 0,015. Nadaljnje parne primerjave (angl. simple analysis) so pokazale, da imajo nadarjeni fantje statistično pomembno višjo samopodobo na področju odnosov z vrstniki kot nadarjena dekleta (p = 0,005). Ugotovili smo tudi, da imajo nadarjena dekleta statistično pomembno nižjo samopodobo na področju odnosov z vrstniki kot visoko učno uspešna dekleta (p = 0,038), ostale parne primerjave niso pokazale statistično pomembnih razlik (p > 0,05). Slika 1: Rezultati na lestvici samopodobe na področju odnosov z vrstniki, ločeni po spolu in skupinah učencev. 99 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Učna samopodoba Nadalje se je pri dvosmerni analizi variance lestvice učne samopodobe pokazal statistično pomemben učinek skupine, F(2, 410) = 36,86,p < 0,001, parcialna yf = 0,15. Post hoc analize so pokazale, da je učna samopodoba ostalih učencev statistično pomembno nižja kot učna samopodoba nadarjenih in učno uspešnih učencev (p < 0,001). Ostale razlike med skupinami niso bile statistično pomembne. Na podlagi dvosmerne analize variance učinek spola na lestvici učne samopodobe ni bil statistično pomemben, F(i, 410) = 0,330, p = 0,57, parcialna yf = 0,001. Prav tako tudi učinek interakcije spol x skupina ni bil statistično pomemben, F(i, 410) = 0,57,p = 0,57, parcialna yf = 0,003. Slika 2: Rezultati na lestvici učne samopodobe, ločeni po spolu in skupinah učencev. Splošna samopodoba Rezultati dvosmerne analize variance so pokazali, da je učinek spola pri splošni samopodobi statistično pomemben, F(i, 410) = 5,09, p = 0,03, parcialna yf = 0,012, in sicer imajo fantje višjo splošno samopodobo v primerjavi z dekleti. Nadalje se je kot statistično pomemben izkazal tudi učinek skupine, F(2, 410) = 8,56, p < 0,001, parcialna yf = 0,40; post hoc analize so pokazale, da imajo učenci, ki sodijo v skupino ostalih učencev, statistično pomembno nižjo splošno samopodobo kot učenci, ki sodijo v 100 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... skupino nadarjenih in učno uspešnih učencev (p < 0,01). Učinek interakcije spol x skupina ni bil statistično pomemben, F(2, 410) = 1,82,p = 0,16, parcialna yf = 0,009. Takšni rezultati nakazujejo, da imata skupina učne uspešnosti in spol močnejši učinek na splošno samopodobo kot interakcija med njima. Navkljub statistično nepomembni interakciji smo zaradi relativno velikih razlik v vrednostih opravili parne primerjave, ki razkrivajo, da je razlika med nadarjenimi fanti in dekleti statistično pomembna (p = 0,01), medtem ko razlika med fanti in dekleti pri visoko učno uspešnih učencih ni statistično pomembna (glej Sliko 2). Tako ne moremo sklepati, da gre za interakcijo med skupinami in spolom, prihaja pa do statistično pomembnih razlik znotraj nadarjenih učencev, ki nakazuje, da imajo nadarjeni fantje višjo splošno samopodobo v primerjavi z nadarjenim dekleti, medtem ko razlika med visoko učno uspešnimi fanti in dekleti ni statistično pomembna. Slika3: Rezultati lestvice splošne samopodobe, ločeni po spolu in skupinah učencev. Razlike v merah socialne sprejetosti Za ugotavljanje razlik v merah socialne sprejetosti smo izvedli multivari-atno analizo variance socialne preferenčnosti in učiteljeve ocene socialne sprejetosti učenca ter ugotovili, da obstaja statistično pomemben učinek skupine (nadarjen, učno uspešen, ostali), V = 0,06, F(4, 806) = 5,81,p < 101 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 0,001; učinek spola in učinek interakcije spol x skupina pa nista bila statistično pomembna. Sledijo rezultati dvosmernih analiz varianc za obe meri socialne spre-jetosti učencev. Socialna preferenčnost Rezultati dvosmerne analize variance so pokazali, da je bil glavni učinek skupine za socialno preferenčnost statistično pomemben, F(2, 405) = 10,40, p < 0,001, parcialna yf = 0,049. Post hoc analize so pokazale, da je socialna preferenčnost ostalih učencev statistično pomembno nižja v primerjavi z učenci iz skupine nadarjenih in učno uspešnih učencev (p < 0,01), medtem ko razlika med nadarjenimi učenci in učno uspešnimi učenci ni bila statistično pomembna. Učinek spola, F(i, 405) = 0,37,p = 0,55, parcialna yf = 0,001, in učinek interakcije spola ter skupine nista bila statistično pomembna, F(i, 405) = 0,30,p = 0,74, parcialna yf = 0,001. Slika 4: Rezultati socialne preferenčnosti učencev, ločenih po spolu in skupinah učencev. Učiteljeva ocena socialne sprejetosti učencev Kot druga mera socialne sprejetosti je bila analizirana učiteljeva ocena. Rezultati dvosmerne analize variance so pokazali statistično pomemben učinek skupine, F(i, 405) = 6,67,p = 0,001, parcialna yf = 0,032. Post hoc 102 u. aram, n. jurinec, m. horvat, k. kosir ■ samopodoba in socialna sprejetost ... analize so pokazale statistično pomembno razliko v učiteljevi oceni med skupino ostalih učencev in drugima dvema skupinama (nadarjeni in učno uspešni učenci),p < 0,05; učitelji so skupino ostalih učencev ocenili kot socialno slabše sprejeto v primerjavi z visoko učno uspešnimi in nadarjenimi učenci, medtem ko razlike med ostalima skupinama niso bile statistično pomembne. Učinek spola, F(i, 405) = 0,05,p = 0,82, parcialna yf < 0,001, in učinek interakcije spol x skupina pa nista bila statistično pomembna, F(2, 405) = 0,86, p = 0,43, parcialna yf = 0,004. Slika5: Rezultati učiteljeve ocene socialne sprejetosti učencev, ločenih po spolu in skupinah učencev. Diskusija Namen pričujoče raziskave je bil preučiti razlike v različnih vidikih sam-opodobe in socialni sprejetosti treh skupin osnovnošolcev: identificiranih nadarjenih učencev, visoko učno uspešnih učencev, ki niso bili identificirani kot nadarjeni, in ostalih učencev, pri čemer smo upoštevali še mod-eratorsko vlogo spola. Ugotovitve raziskav, ki so preučevale socialno-emo-cionalne značilnosti nadarjenih učencev, so nekonsistentne, kar lahko delno pripišemo tudi različnim opredelitvam nadarjenosti. Ker je večina teh raziskav narejenih na tujih vzorcih, posplošitve in primerjave s slovenskim vzorcem, tudi zaradi širše opredelitve nadarjenosti pri nas, niso 103 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 zanesljive, zato smo želeli raziskati, kako se omenjeni pojavi odražajo pri slovenskih osnovnošolcih. Samopodoba na področju odnosov z vrstniki Rezultati na lestvici samopodobe na področju odnosov z vrstniki, kjer so nadarjena dekleta dosegla pomembno nižji rezultat kot nadarjeni fantje in pomembno nižji rezultat kot visoko učno uspešna dekleta, nakazuje, da bi nadarjena dekleta utegnila biti rizična skupina, za katero oznaka nadarjenosti predstavlja dejavnik tveganja. Skladno z rezultati naše raziskave je podobno tudi Ablard (1997) ugotovila, da imajo dekleta, nadarjena na učnem področju, nižjo socialno samopodobo v primerjavi z nadarjenimi fanti. Nasprotno pa nekatere raziskave niso našle nobenih razlik med spoloma za področje učne, socialne in čustvene samopodobe pri nadarjenih učencih (npr. Bain in Bell, 2004; Chan, 2001; Cunningham in Rinn, 2007; Lee et al., 2012). Rezultati raziskav, ki so primerjali nadarjene in nenadarjene učence (Hoge in Renzulli,i993; Litster in Roberts, 2011), sovpadajo z našimi rezultati, saj ugotavljajo, da na področju socialne samopodobe med skupinama ni pomembnih razlik. Bain in Bell (2004), ki sta primerjali identificirane nadarjene in visoko učno uspešne učence, sta ugotovili, da nadarjeni učenci dosegajo višje rezultate (med drugim tudi) pri dimenziji samopodobe na področju odnosov z vrstniki, česar pa rezultati v naši raziskavi ne potrjujejo. Učna samopodoba Glede na kriterij razdelitve učencev v skupine po učni uspešnosti imajo nadarjeni učenci pričakovano najvišjo učno samopodobo, sledijo jim visoko učno uspešni učenci, najnižje pa so se uvrstili učenci iz skupine ostalih učencev, ki imajo na dotičnem področju pomembno nižjo samopodobo od ostalih dveh skupin učencev. Naši rezultati so skladni z raziskavo Ritchotte in sodelavci (2016), ki so primerjali identificirane nadarjene in visoko učno uspešne učence v samozaznavanju na področju učnih sposobnosti in ugotovili, da so njihove samozaznave primerljive. Če štejemo skupino visoko učno uspešnih učencev k nadarjenim, potem so naši rezultati skladni tudi z večino drugih raziskav in metaanaliz, ki so primerjale učno samopodobo med nadarjenimi in nenadarjenimi vrstniki ter ugotovile, da je učna samopodoba nadarjenih učencev višja (Hoge in Renzulli, 1993; Litster in Roberts, 2011; Preckel et al., 2008). Zanimiva je ugotovitev, da imajo nadarjena dekleta nižjo učno samopodobo kot nadarjeni fantje, čeprav imajo dekleta boljše ocene od fantov in bi lahko sklepali, da bodo imela posledično tudi višjo učno samopo- 104 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... dobo. V skupini visoko učno uspešnih učencev pa so rezultati na lestvici učne samopodobe skladnejši z ocenami - dekleta imajo boljše ocene in se tudi zaznavajo kot sposobnejše na učnem področju. Nižjo zaznavo učne samopodobe nadarjenih deklet bi morda lahko pripisali njihovim pričakovanjem do sebe, da morajo biti, zaradi označbe nadarjenosti in s tem višjih pričakovanj družbe, še boljše. Splošna samopodoba Rezultati na področju splošne samopodobe so pokazali, da ni pomembnih razlik med identificiranimi nadarjenimi in visoko učno uspešnimi učenci. Učenci iz skupine ostalih učencev imajo najnižjo splošno samopodobo - pomembno nižjo od ostalih dveh skupin; razlike med spoloma pa so pokazale, da imajo fantje višjo splošno samopodobo kot dekleta. Dobljeni rezultati sovpadajo z ugotovitvami metaanalize, v kateri sta Hoge in Renzulli (1993) primerjala nadarjene in nenadarjene učence; ugotovila sta, da se nadarjeni učenci na področju splošne samopodobe zaznavajo kot uspešnejši od nenadarjenih učencev. Bain in Bell (2004), ki sta primerjali identificirane nadarjene in visoko učno uspešne učence, ki niso bili identificirani kot nadarjeni, sta v svoji raziskavi ugotovili, da so nadarjeni učenci dosegli višji rezultat na lestvici splošne samopodobe kot visoko učno uspešni, kar se v naši raziskavi ni pokazalo. Socialna sprejetost Kot eno izmed mer socialne sprejetosti smo uporabili socialno prefer-enčnost. Skupini nadarjenih učencev in visoko učno uspešnih učencev sta imeli statistično pomembno višjo socialno preferenčnost kot skupina ostalih učencev. Dobljeni rezultati sovpadajo z ugotovitvami raziskave Peairs (2010), ki je ugotovila, da imajo nadarjeni učenci višjo socialno preferenčnost. O učinkovitejšem delovanju nadarjenih učencev v socialnih odnosih poročajo tudi Cohen et al. (1994). Podrobnejše analize primerjav med skupinami in spoloma (ki niso bile statistično pomembne) pa nakazujejo podobnost z raziskavo Luftiga in Nicholsa (1990), ki sta skupine oblikovala na podlagi kognitivnih sposobnosti in spola ter ugotovila, da so nadarjeni fantje najbolj priljubljeni, sledili so nenadarjeni fantje in nenadarjena dekleta. Nadarjena dekleta so predstavljala najmanj priljubljeno skupino izmed omenjenih štirih. Če iz naše raziskave izvzamemo skupino ostalih učencev in primerjamo skupini nadarjenih in učno uspešnih, tudi naši rezultati kažejo, da so nadarjena dekleta v socialni preferenčnosti ocenjena najslabše. Ta ugotovitev nakazuje, da lahko nadarjenost vendarle predstavlja dejavnik tveganja za razvoj socialno-emocionalnih težav pri dekletih. Poleg tega nekatere raziskave 105 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 kažejo, da nadarjena dekleta v večji meri zanikajo svojo nadarjenost v primerjavi s fanti (Swiatek, 2001; Swiatek in Dorr, 1998) in pripisujejo večjo vrednost sprejemanju s strani vrstnikov (Chan, 2003, 2004). Nasprotno z rezultati naše raziskave pa rezultati, ki sta jih dobili López in Sotillo (2009), niso pokazali razlik med skupinama nadarjenih in nenadarjenih učencev v socialni preferenčnosti. Kot drugo mero socialne sprejetosti smo uporabili učiteljevo oceno socialne sprejetosti učencev. Rezultati so pokazali, da so učitelji ocenili nadarjene in učno uspešne učence kot pomembno bolj sprejete od ostalih učencev. Če združimo skupini nadarjenih in učno uspešnih učencev, ki sta dosegli primerljive ocene učiteljev v socialni sprejetosti, rezultati pričujoče raziskave skladno z rezultati Peairs (2010) kažejo, da učitelji zaznavajo nadarjene učence kot spretnejše od njihovih nenadarjenih vrstnikov. Nasprotno pa López in Sotillo (2009) nista ugotovili razlik v učiteljevi oceni socialne sprejetosti med nadarjenimi in nenadarjenimi učenci. Če povzamemo rezultate mer socialne sprejetosti, lahko trdimo, da so rezultati identificiranih nadarjenih in visoko učno uspešnih učencev podobni; pomembne razlike se pokažejo, če ti dve skupini primerjamo s skupino ostalih (nenadarjenih) učencev, saj so slednji, sodeč po rezultatih, slabše socialno sprejeti. Če pojmujemo socialno preferenčnost in učiteljevo oceno sprejetosti kot objektivna pokazatelja socialne sprejetosti, potem bi lahko predpostavljali, da se bodo rezultati samopodobe na področju odnosov z vrstniki skladali z rezultati socialne sprejetosti, vendar rezultati samopodobe na področju odnosov z vrstniki niso pokazali pomembnih razlik med nobeno izmed treh skupin. Presenetijo pa rezultati nadarjenih deklet, ki se kljub dobri socialni sprejetosti s strani vrstnikov zaznavajo pomembno nižje kot nadarjeni fantje in visoko učno uspešna dekleta, ki naj bi bila nadarjenim dekletom izmed vseh ostalih skupin v našem vzorcu najbolj podobna. Raziskave (Cillessen, 2011; Pittinsky in Carolan, 2008) sicer kažejo, da podatki, pridobljeni s samooceno ali učiteljevo oceno, niso nujno skladni s podatki, ki jih pridobimo s sociometrično preizkušnjo. Zaključek Na podlagi podatkov te raziskave lahko sklenemo, da identificirani nadarjeni učenci in visoko učno uspešni učenci dosegajo primerljive rezultate pri vseh treh preverjanih dimenzijah samopodobe (splošna, učna in samopodoba na področju odnosov z vrstniki) kakor tudi v obeh merah socialne sprejetosti (socialna preferenčnost in učiteljeva ocena socialne sprejetosti učenca). Pomembne razlike pa so se pokazale med omenjenima skupinama in skupino ostalih učencev v vseh v raziskavo vključenih spremenljivkah, razen v samopodobi na področju odnosov z vrstniki, kjer ni 106 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... bilo razlik med skupinami. Skupini identificiranih nadarjenih in visoko učno uspešnih učencev sta si torej kljub široki opredelitvi nadarjenosti bolj podobni kot različni. Očitno je učna uspešnost skupni imenovalec, ki prinaša pozitivnejšo samopodobo in socialno sprejetost, ne glede na druge različnosti in posebnosti posameznikov. Nadarjena dekleta imajo izmed vseh primerjanih skupin najnižjo samopodobo na področju odnosov z vrstniki, kljub temu da se v obeh merah socialne sprejetosti pomembno ne razlikujejo od drugih skupin nadarjenih in visoko učno uspešnih učencev. Na podlagi rezultatov pričujoče študije in obravnavane literature lahko predpostavljamo, da oznaka nadarjenosti za nadarjena dekleta predstavlja dejavnik tveganja, medtem ko za nadarjene fante predstavlja varovalni dejavnik. Omejitve raziskave in predlogi za nadaljnje raziskovanje Pričujoča raziskava ima nekaj pomanjkljivosti, na katere je treba opozoriti pri interpretaciji in posplošitvi naših rezultatov. Ena izmed ključnih pomanjkljivosti je priložnostno vzorčenje in velikost vzorca, vključenega v raziskavo. Da bi rezultate, pridobljene v raziskavi, lahko posplošili na celotno slovensko osnovnošolsko populacijo zadnje triade, bi bilo treba zajeti vzorec, proporcionalen s slovenskimi učenci zadnje triade glede na spol, starost in regijo. Vse zaključke, vezane na primerjanje skupin identificiranih nadarjenih učencev in učno uspešnih učencev, je treba sprejeti z zadržki, saj je razdelitev vzorca vprašljiva 1) zaradi neizenačenosti in heterogenosti skupine nadarjenih učencev, ki je posledica zelo široke slovenske opredelitve nadarjenosti, in zaradi 2) določitve kriterija za visoko učno uspešne učence, ki je postavljena na podlagi samoporočanih ocen. Prav tako za učence iz skupine visoko učno uspešnih učencev nismo imeli podatkov, ali so bili v preteklosti v postopku identifikacije nadarjenosti in v postopku niso bili prepoznani ali pa preprosto niso bili evidentirani kot potencialno nadarjeni učenci. Obstaja možnost, da so v skupini učno uspešnih učencev tudi učenci, ki bi bili v postopku identifikacije prepoznani kot nadarjeni učenci, če bi bili evidentirani. Nadaljnjo omejitev predstavlja tudi veljavnost podatkov, pridobljenih s samoporočanjem, kar še posebej velja za podatke o učni uspešnosti učencev. Kot slabost oziroma omejitev raziskave lahko navedemo tudi majhno variabilnost v učiteljevi oceni socialne sprejetosti učencev. Večina učiteljev je učencem namenila eno izmed dveh najvišjih ocen sprejetosti (»učenci sošolca večinoma sprejemajo« in »učenci sošolca dobro sprejemajo«). Dodatno bi lahko postopek pridobivanja učiteljevih ocen objek- 107 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 tivizirali z vključitvijo več učiteljev za ocenjevanje posameznega razreda in kot končno mero uporabili povprečno učiteljevo oceno za vsakega učenca. V nadaljnjih raziskavah bi bilo smiselno dodatno kvalitativno preučiti skupino nadarjenih deklet, ki se je v naši raziskavi izkazala kot rizična, saj bi na ta način pridobili dragocen vpogled v delovanje dotične skupine deklet. Glede na široko zastavljeno slovensko opredelitev nadarjenosti bi bilo zanimivo analizirati tudi različne programe za nadarjene, ki jih predpisuje Koncept o delu z nadarjenimi učenci v osnovni šoli. Literatura Ablard, K. (1997) Self-perceptions and needs as a function of type of academic ability and gender. RoeperReview20 (2), pp. 110-115. Bain, S. K., Bell, S. M. (2004) Social self-concept, social attributions, and peer relationships in fourth, fifth, and sixth graders who are gifted compared to high achievers. Gifted Child Quarterly 48 (3), pp. 167178. Bainbridge, C. (2016) https://www.verywell.com/high-achiever-1449168 (12. 5. 2016). Carman, C. A. (2013) Comparing apples and oranges: Fifteen years of definitions of giftedness in research. Journal of Advanced Academics 24 (1), pp. 52-70. Chan, D. W. (2001) Global and specific self-concepts of gifted adolescents in Hong Kong. Journal for the Education of the Gifted 24 (4), pp. 344-364. Chan, D. W. (2003) Dimensions of emotional intelligence and their relationships with social coping among gifted adolescents in Hong Kong. Journal of Youth and Adolescence 32 (6), pp. 409-418. Chan, D. W. (2004) Social coping and psychological distress among Chinese gifted students in Hong Kong. Gifted Child Quarterly 48 (1), pp. 30-41. Chan, D. W. (2010) Perfectionism among Chinese gifted and nongifted students in Hong Kong: The use of the Revised Almost Perfect Scale. Journalfor the Education for the Gifted34 (1), pp. 68-98. Cillessen, A. H. N. (2011) Sociometric Methods. V Rubin, K. H., Bu-kowski, W. M., Laursen, B. (eds.). Handbook of Peer Interactions, Relationships, and Groups. New York: A Division of Guilford Publications, Inc., pp. 82-99. Cohen, R., Duncan, M., Cohen, S. L. (1994) Classroom peer relations of children participating in a pull-out enrichment program. Gifted Child Quarterly 38 (1), pp. 33-37. 108 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... Coie, J. D., Dodge, K. A., Coppotelli, H. A. (1982) Dimensions and types of social status: A crossage perspective. Developmental Psychology 18, pp. 557- 570. Coleman, L. J., Cross, T. L. (1988) Is being gifted a social handicap? Jour-nalfor the Education of the Gifted 11 (4), pp. 41-56. Cross, T. L. (1997) Psychological and social aspects of educating gifted students. Peabody Journal of Education 72 (3&4X pp. 180-200. Cross, T. L., Adams, C., Dixon, F., Holland, J. (2004) Psychological characteristics of academically gifted adolescents attending a residential academy: A longitudinal study. Journal for the Education of the Gifted28 (2), pp. 159-181. Cross, T., Coleman, L., Terhaar-Yonkers, M. (1991) The social cognition of gifted adolescents in schools: Managing the stigma of giftedness. Jo-urnalfor the Education of the Gifted 15 (1), pp. 44-55. Cunningham, L. G., Rinn, A. N. (2007) The role of gender and previous participation in a summer program on gifted adolescents' self-concepts over time. Journal of the Education of the Gifted30 (3), pp. 326352. Foust, R. C., Rudasill, K. M., Callahan, C. M. (2006) An investigation into the gender and age differences in the social coping of academically advanced students. Journal of Advanced Academics 18 (1), pp. 60-80. Frederickson, N. L., Furnham, A. F. (2001) The long-term stability of soci-ometric status classification: A longitudinal study of included pupils who have moderate learning difficulties and their mainstream peers. Journal of Child Psychology and Psychiatry 42 (5), pp. 581-592. Freeman, J. (2006) Giftedness in the long term. Journal for the Education of the Gifted29 (4), pp. 384-403. Garland, A. F., Zigler, E. (1999) Emotional and behavioral problems among highly intellectually gifted youth. Roeper Review 22 (1), pp. 41-44. Hatzichristou, C., Hopf, D. (1996) A multiperspective comparison ofpeer sociometric status groups in childhood and adolescence. Child Development 67 (3), pp. 1085-1102. Hoge, R. D., Renzulli, J. S. (1993) Exploring the link between giftedness and self-concept. ReviewofEducational Research 63 (4), pp. 449-465. Juriševič, M. (2012) Nadarjeni učenci v slovenski šoli. Ljubljana: Pedagoška fakulteta. Juriševič, M. (2009) Odkrivanje in delo z nadarjenimi učenci v šoli - stanje in perspektive. Psihološka obzorja 18 (4), pp. 153-168. Kerr, B. A., Colangelo, N., Gaeth, J. (1988) Gifted adolescents' attitudes toward their giftedness. Gifted Child Quarterly 32 (2), pp. 245-247. 109 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Košir, K., Sočan, G., Pečjak, S. (2007) The role of interpersonal relationships with peers and with teachers in students' academic achievement. Review of Psychology 14 (1), pp. 43-58. Lee, S.-Y., Olszewski-Kubilius, P., Turner Thomson, D. (2012) Academically gifted students' perceived interpersonal competence and peer relationships. Gifted Child Quarterly 56 (2), pp. 90-104. Litster, K. in Roberts, J. (2011) The self-concepts and perceived competencies of gifted and non-gifted students: A meta-analysis. Journal of Research in Special Educational Needs 11 (2), pp. 130-140. LoCicero, K. A., Ashby, J. S. (2000) Multidimensional perfectionism in middle school age gifted students: A comparison to peers from the general cohort. RoeperReview 22 (3), pp. 182-185. López, V., Sotillo, M. (2009) Giftedness and social adjustment: Evidence supporting the resilience approach in Spanishlspeaking children and adolescents. High Ability Studies 20 (1), pp. 39-53. Luftig, R. L., Nichols, M. L. (1990) Assessing the social status of gifted students by their age peers. Gifted Child Quarterly 34 (3), pp. 111-115. Marsh, H. W. (1992) SelfDescriptionQuestionnaire, II. Sydney, New South Wales, Australia: University of Western Sydney. Marsh, H. W., Parada, R. H., Ayotte, V. (2004) A Multidimensional Perspective of Relations Between Self-Concept (Self Description Questionnaire II) and Adolescent Mental Health (Youth Self-Report). Psychological Assessment 16 (1), pp. 27-41. McGraw-Hill. (1996) Comprehensive Test of Basic Skills, Monterrey, CA: CTB/McGraw-Hill. Mueller, C. E. (2009) Protective factors as barriers to depression in gifted and nongifted adolescents. Gifted Child Quarterly 53 (1), pp. 3-14. Musek, J. (2005) Psihološke in kognitivne študije osebnosti. Ljubljana: Znanstveni inštitut Filozofske fakultete. Neihart, M. (1999) The impact of giftedness on psychological well-being: What does the empirical literature say? Roeper Review 22 (1), pp. 1017. Neihart, M. (2002) Risk and resilience in gifted children: A conceptual framework. In Neihart, M., Reis, S., Robinson, N. M., Moon, S. M. (eds.). The social and emotional development of gifted children: What do we know?. Waco, TX: Prufrock Press, pp. 113-122. Ollendick, T. H., Weist, M. D., Borden, M. C., Greene, R. W. (1992) So-ciometric status and academic, behavioral, and psychological adjustment: A five-year longitudinal study. Journal of Consulting and Clinical Psychology 60 (1), pp. 80-87. 110 u. ARAM, N. JURINEC, M. HORVAT, K. KOŠIR ■ SAMOPODOBA IN SOCIALNA SPREJETOST ... Olszewski-Kubilius, P., Subotnik, R. F., Worrell, F. C. (2015) Conceptualizations of giftedness and the development of talent: Implications for counselors. Journal of Counseling Development 93 (2), pp. 143-152. Peairs, K. F. (2010) The social world of gifted adolescents: Sociometric status, friendship and social network centrality (Neobjavljena disertacija). Pridobljeno iz ProQuest dissertation thesis (9. 5. 2016). Pfeiffer, S. I. (2009) The gifted: Clinical challenges for child psychiatry. Journal of the American Academy of Child Adolescent Psychiatry 48 (8), pp. 787-790. Pittinsky, M., Carolan, B. V. (2008) Behavioral versus cognitive classroom friendship networks. Social Psychology of Education 11 (2), pp. 133147. Preckel, F., Goetz, T., Pekrun, R., Kleine, M. (2008) Gender differences in gifted and average-ability students comparing girls' and boys' achievement, self-concept, interest, and motivation in mathematics. Gifted Child Quarterly 52 (2), pp. 145-159. Renzulli, J. S. (1977) The enrichment triad model: A guidefor developing defensible programs for the gifted and talented. Wethersfield, CT: Creative Learning Press. Rimm, S. (2002) Peer pressures and social acceptance of gifted students. In Neihart, M., Reis, S., Robinson, N. M., Moon, S. M. (eds.). The social and emotional development of gifted children: What do we know? Waco, TX: Prufrock Press, pp. 13-18. Ritchotte, J. A., Suhr, D., Alfurayh, N. F., Graefe, A. K. (2016) An Exploration of the Psychosocial Characteristics of High Achieving Students and Identified Gifted Students: Implications for Practice. Journal of Advanced Academics 27 (1), pp. 23-38. Robinson, N. M. (2002) Introduction. In Neihart, M., Reis, S., Robinson, N. M. in Moon, S. M. (eds.). The social and emotional development of gifted children. What do we know?. Waco, TX: Prufrock Press, pp. xi-xxiv. Shavelson, R. J., Hubner, J. J., Stanton, G. C. (1976) Self-concept: Validation of construct interpretations. Rewiew of Educational Research 46 (3), pp. 407-441. Silverman, L. K. (2002) Asynchronous development. In Neihart, M., Reis, S., Robinson, N. M., Moon, S. M. (eds.). The social and emotional development of gifted children. What do we know? Waco, TX: Prufrock Press, pp. 31-37. Strokovni svet RS za splošno izobraževanje (1999) Koncept: Odkrivanje in delo z nadarjenimi učenci v devetletni osnovni šoli. 111 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Swiatek, M. A. (2001) Social coping among gifted high school students and its relationship to self-concept. Journal of Youth and Adolescence 30 (1), pp. 19-39. Swiatek, M. A., Dorr, R. M. (1998) Revision of the social coping questionnaire: Replication and extension of previous findings. Journal of Advanced Academics 10 (1), pp. 252-259. Terman, L. M. (1925) Genetic studies of genius: Vol. 1. Mental and physical traits of a thousand gifted children. Stanford, CA: Stanford University Press. Terman, L. M., Oden, M. H. (1947) The gifted group grows up: Twenty five years'follow-up of a superior group. Genetic studies of genius, Vol. 4. Stanford, CA: Stanford University Press. Terman, L. M., Oden, M. H. (1959) The gifted group at mid-life:3s years'follow up of the superior child. Genetic studies of genius, Vol. 5. Stanford, CA: Stanford University Press. Wentzel, K. R. (1991) Relations between social competence and academic achievement in early adolescence. Child Development 62 (5), pp. 1066-1078. Wentzel, K. R. (1993) Does being good make the grade? Social behavior and academic competence in middle school. Journal of Educational Psychology 8s (2), pp. 357-364. Wentzel, K. R., Asher, S. R. (1995) The academic lives of neglected, rejected, popular, and controversial children. Child Development 66 (3), pp. 754-763. 112 Motivational Goals and Academic Performance from the Perspective of Students' Perceived Quality of Relationship with Their Class Teachers at the Start of the Upper Secondary Education Level Klaudija Sterman Ivancic and Melita Puklek Levpuscek Student's Transition to the Secondary Education Level he transition to the secondary education level represents a new edu- cational and social environment for an adolescent. Each adolescent meets peers and teachers previously unknown to them, and in addition, the upper secondary education leveli environment is much more academically oriented and focused on educational achievements in comparison with the primary education. Gutman and Eccles (2007) stated that this transition to the new form of education is one of the most important life changes for the adolescent beside puberty, cognitive development and changes in family and friendship relationships which all play an important role in adolescents' further development. Various studies show that adolescents - upon entering a new school environment - are likely to deal with lower levels of confidence in establishing new relationships and perceive a poorer social support from teachers and peers (Bronfenbrenner, 1979; Eccles & Midgley, 1989; Wentzel, 1998); lower motivation for learning and lower educational achievements (Barber & Olsen; 2004; Eccles et al., 1993; Eccles, 2004; Gutman & Midgley, 2000; Wig-field & Eccles, 1994); as well as can have poorer endeavor for attending classes (Elias, Gara & Ubriaco, 1985). Beside, some authors (e.g. Bronfenbrenner, 1979; Eccles & Midgley, 1989; Wentzel, 1998) point out that interpersonal relationships play an important role in an adolescent's adaptation to the 1 In Slovenia the education system consists of uniformed 9-year primary school and secondary school. This means that secondary education level often described in the literature is equivalent to grades 6-9 ofSlovenian primary school and not to Slovenian upper secondary education level. This is why we make a distinction between secondary and upper secondary education level. 113 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 new school environment, as they affect motivation for learning, academic achievements and adapted learning behaviour. Authors (Maulana, Opdenakker, Den Brook & Bosker, 2012; Went-zel, 1994) who studied changes in relationship of adolescents with their teacher at the transition to the secondary education level, have found that adolescents and teachers establish a certain level of mistrust at the start of the new educational journey, as both of them only get to know each other; and that adolescents in general perceive less opportunities for establishing more confident relationships with their teachers as they had on the primary school level. The authors see this initial mistrust in the relationship between adolescents and teachers as one of the most important reasons for the decrease in the motivation for learning and academic achievements. Results of some studies (e.g. Rueger, Malecki & Demaray, 2010; Sawyer, Pfeiffer & Spence, 2009) also show that girls are more sensitive to the transition to the upper secondary education level than boys and perceive this transition as more stressful. The Quality of Student-Teacher Relationship In this study, we wanted to explore the quality of teacher-student relationship from point of socio-emotional support as perceived by students2. We used Weiss' definition of social support (1988) which includes all important elements of the above mentioned features of social support, i.e. the function of socialising, and the emotional, instrumental and informational aspects of a social support: a) stable attachment which gives us the feeling of emotional security and closeness (emotional support); b) social integration or sense of belonging and closeness with the similarly thinking individuals (emotional support and socialising); c) altruism or the need to care for others (emotional support); d) providing a reciprocal approval, including a mutual stimulation of confidence and affirmation of one's own worth (emotional support); e) a reliable alliance which is related to the reliable availability of a social environment if help at resolving everyday problems is needed (or instrumental help); and f) guidance in case of any stressful developments or danger in the form of advice, direction or efficient strategies in resolving problems (informational support). According to Weiss, an individual will perceive appropriate support from the environment and will not feel isolated if they has an access to all six above mentioned sources to fulfil their social needs in his social environment (Cutrona & Russell, 1987; Furman & Buhrmester, 1985; Greg-go, 2008). 2 Since we discuss the relationship between adolescents and teachers (i.e. in educational context), in the following sections adolescents are referred to as students. 114 K. STERMAN IVANCIC, M. PUKLEK LEVPUSCEK ■ MOTIVATIONAL GOALS ... However, different studies show that students more often seek for the instrumental and informational support of the teacher and less often for the emotional one (Darling, Hamilton & Niego, 1994; Furman & Buhrmester, 1985; Lempers & Clark-Lempers, 1992). Boys are especially keener to seek informational support in their relationship with teachers whereas girls more often turn to teachers for emotional support and also tend to report of higher levels of such support in their relationship (Kerr, Preuss & King, 2006; Rueger, Malecki & Demaray, 2010; Sawyer, Pfeiffer & Spence, 2009; Sontag & Graber, 2010). What we are interested here is, to what extent is the teacher's so-cio-emotional support present in the first year of an upper secondary school level education and how it relates to students' academic motivation and achievement. The Quality of Student-Teacher Relationship and Academic Achievement In his theoretical model of a relationship between a teacher and a student,3 Pianta (Pianta, Hamre & Stuhlman, 2003) defines a good-quality relationship as a key factor contributing to the student's academic achievements, and stresses the importance of the teacher's socio-emotion-al support. Such support stimulates the individual's socio-emotional development and is of particular importance for students with learning and behavioural difficulties. Pianta also pointed out that the importance of a good-quality relationship with the teacher does not diminish with student's growing up, and that the teacher's socio-emotional support is important in the times of transition to the higher levels of education. An upper secondary school teacher, who shows emotional warmth and acceptance and is there for their students, stimulates students' learning interests which in turn is shown in better academic achievements and vice versa. Students who report higher levels of conflict and negative interactions in their relationships in their school environment on average report of lower academic achievements (Berndt & Keefe, 1996; Pianta, Hamre & Stuhlman, 2003; Wentzel & Caldwell, 1997; Wentzel, 2012). An interesting study done by Košir (2013) revealed that even young teachers in Slovenia still believe in a stereotype that teacher's warm, supportive and caring relationships with his students result in poorer achievements of educational goals. However, studies from the past decade show the opposite. Various Slovenian and foreign researchers (e.g. Connell & 3 Peklaj and Pečjak (2015) state that we find only Pianta's theoretical model for explaining a relationship between a teacher and a pupil in the field of educational psychology, despite the importance of this relationship. 115 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 Wellborn, 1991; Deci, 1992; Gregory & Weinstein, 2004; Magajna, Ka-vkler, Čačinovič-Vogrinčič, Pečjak & Bregar, 2008; Pianta & Walsh, 1996; Wentzel, 1997) established - their conclusions are based on the findings from the studies exploring the quality of students' relationships in educational context - that a student's perception of socio-emotional support by their teacher is very important for achieving their learning goals. A teacher's emotional support hence plays an important role in improving students' learning adjustment. The Quality of Student-Teacher Relationship and Motivation for Learning In their model of academic motivation, Darling and Steinberg (1993) pointed out that the teacher's emotional support was an important motivational factor in an educational situation. In other words, it is important that the teacher offers the student an opportunity for a supportive and mutual relationship beside rules and expectations the teacher has for an individual. Teacher's expectations regarding academic achievements and behaviour are the most effective if expressed within emotionally positive and thoughtful relationship (Wentzel, 1997, 2002, 2003, 2012). Pupils thus find it easier to identify with the teacher's values which gradually become their own motivation for learning. In addition, the socio-emo-tional support by the teacher strengthens the young person's sense of connectedness and belonging to the school, and as such stimulates their motivation for learning that consequently contributes to better academic achievements. The motivation for learning stretches beyond the contextual factors in a school and classroom - it is mainly a result of a successful socialization processes, including the good-quality relationships between teachers and students. Recent studies (e.g. Crosnoe, Johnson & Elder. 2004; Gregory et al., 2010; Murdock & Miller; 2003; Wentzel, 1997; Wentzel, 2012, Wentzel, Russell & Baker, 2015) confirmed the importance of the following aspects of a relationship with a teacher which contribute to a student's better motivation for learning: a confidential relationship with students, aspiration to link the school curriculum with students' interests, and a balance between awarding achievements and emphasizing the importance and value of a learning experience. The results further show that the perception of a teacher to be supportive to some extent depends on the students' school curriculum and students' learning abilities. Students from a general upper secondary school perceive a supportive teacher as a person who encourages them to tackle new challenges and to cooperate within the class, while for students from vocational schools a teacher has to be above all kind and 116 K. STERMAN IVANCIC, M. PUKLEK LEVPUSCEK ■ MOTIVATIONAL GOALS ... just; be able to explain subject matter clearly and to maintain order in the classroom (Daniels & Araposthatis, 2005). Student's Motivational Orientation Since we are interested in the effects of teacher-student relationship on students' motivational orientation in the context of students' transition to the secondary school level education, we used the achievement goal theory to investigate motivational goals in academic environment (Elliot, McGregor & Gable, 1999; Harackiewicz & Elliot, 1993; Middleton & Midgley, 1997). According to this theory, it is the motivation goals and not final results that give a meaning to active performance in a certain learning situation for an individual (Maehr & Zusho, 2009). The two most important motivational goals within the framework of this theoretical concept (Elliot et. al, 1999; Harackiewicz & Elliot, 1993; Middleton & Midgley, 1997) are: i) mastery goals and the development of one's own abilities, and ii) performance goals of showing and comparing one's own abilities. The latter represent self-presentation goals which can be further divided into performance-approach goals and performance-avoidance goals. Student's Motivational Orientation and Academic Achievement Various studies in Slovenia and abroad show that the correlation between students' motivational goals and academic achievements is ambiguous. Peklaj and her colleagues (Peklaj et al., 2009) studied the effects of motivational orientations on learning achievements (Mathematics, Slovene language and final academic achievement). The study included Slovenian pupils (grade eight of primary school) and secondary school students (third year of upper secondary school). The results show that mastery goal orientation is importantly related to higher marks in Mathematics in the former group (similar finding in the study of Puklek Levpušček & Zupančič (2009)), and to higher marks in the Slovene language and Mathematics in the latter group. Performance goal orientation of students in secondary school was connected with better achievements in both subjects. The same holds for the individual's performance-avoidance goal orientation in Mathematics. The authors therefore point out that correlations between motivational orientation and achievements show significant differences according to the learning context (e.g., more achievement oriented secondary school environment). These findings are supported by some other studies (e.g., Linnenbrink-Garcia, Tyson & Patall, 2008; Zusho, Karabenick, Bonney & Sims, 2008) where the researchers established that the correlation between mas- 117 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 tery goals and academic achievement does not differ from the correlation between performance goal and educational achievements in the school contexts where marks serve as a criterion for measuring academic success. Although the correlations between mastery goals and academic achievements are in general positive, regression analyses do not show that such motivational orientation would be a clear indicator of learning achievements, especially if the researchers took the individual's previous learning achievements into account. This of course poses a question (Maehr & Zusho, 2009), to what extent are the individuals with mastery goals orientation even motivated for achieving high marks and to what extent can marks even be a real indicator of an individual's abilities. In their literature review of the effects of individual's motivational orientation on learning achievements, Maehr and Zusho (2009) establish that the majority of studies used marks as a criterion for evaluating individual's abilities. They further state that such an approach ignores the contextual factors that influence the individual's motivational orientation. They - similar to the authors of the above mentioned studies - establish that the results of the studies exploring correlations between the individual's motivational goals, other aspects of motivation and academic achievements, are rather unified regarding an individual's mastery goals and performance-avoidance goals: the individual's mastery orientation strengthens his interest for a certain learning field and meta-cog-nitive learning strategies use (e.g. Elliot et al., 1999; Pintrich, 2000), and while it is true that the performance goals orientation has a positive effect on the individual's motivation and academic achievements, students with this kind of motivational orientation generally express a higher level of anxiety in learning situations and a poorer interest for the subject (e.g. Church, Elliot & Gable, 2001; Elliot & Church, 1997). As Grolnick and her colleagues (Grolnick, Friendly & Bellas, 2009) claim, we witness a significant difference between those with the mastery goals orientation and those with the performance goals orientation which is expressed in the fact that the latter lacks the autonomous motivation in the sense of interest for an in-depth learning of a certain subject, although they show similar positive learning self-image and on average achieve good academic performance. Aims and Hypotheses In the present study we want to examine students' perceived socio-emo-tional support of their class teacher because the effects of interpersonal relationships on students' learning adjustment during the transition to upper secondary school has not been profoundly researched yet. We assume 118 K. STERMAN IVANCIC, M. PUKLEK LEVPUSCEK ■ MOTIVATIONAL GOALS ... that interpersonal relationships are an important factor in the student's adaptation to a new educational environment, as they are connected to the student's motivation for learning and consequently to their learning achievements and adapted learning behaviour (e.g. Bronfenbrenner, 1979; Eccles & Midgley, 1989; Wentzel, 1998). Different authors have already confirmed the effects of emotional closeness, connection and affection in the student-teacher relationship on the student's motivation for learning and the learning achievements (e.g. Armenta, Knight, Carlo & Jacob-son, 2011; Brittain et al., 2013; Gregory & Weinstein, 2004; Spera, 2006; Wentzel et al., 2015). However, in most research, the motivation for learning is explored mainly from the aspect of the student's interest for learning. Students that receive more affection, support and positive attitude from a teacher, have a greater interest in learning and feel more competent, and consequently show better learning achievements (Midgley, Feld-lauffer & Eccles, 1989; Roeser, Eccles & Sameroff, 1998; Ryan & Grolnick, 1986). Some studies (e.g. Bouffard, Boileau & Vezeau, 2001; Eccles, Midgley & Adler, 1984) also show significant changes in student motivational orientation when entering secondary school, especially negative changes in mastery goal orientation. The results show that one of the reasons for such changes lie in student's social experiences encouraged by systematic changes in their school environment that are in contrast with student's increased competency and social maturity, e.g. more closed, controlled teacher-dominated and formal school environment, teaching practices that provide students with lower sense of autonomy and control, and external reward system. However, there is less research examining students' perceptions of teacher's support when entering secondary educational level and their motivational orientation in this period. We are also interested in investigating the predictive power of student's perceived socio-emo-tional support and negative interactions in their relationship with the class teacher when predicting student's motivational orientation and academic performance, while controlling for educational programme and gender. Based on previous research we formed the following hypotheses: i) students will report relatively low levels of perceived socio-emotional support in relationship with their class teacher and relatively high levels of performance goal orientation, ii) there are significant differences in perceived socio-emotional support and motivational orientation between students from different educational programmes, iii) there are significant gender differences in perceived socio-emotional support and motivational orientation, iv) perceived teacher socio-emotional support is an impor- 119 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 tant predictor of student's motivational orientation and academic performance when controlling for enrolled educational programme and gender. Method Participants Students that participated in this research also participated in the PISA 2012 research. The PISA research in Slovenia includes all secondary school programmes; within the programme among all the 15-year-olds approximately 30 students are randomly chosen. We kept the sample of randomly chosen students within educational programmes and randomly selected 56 programmes to participate in our study. 47 of the 56 invited educational programmes decided to participate. The study included 602 students enrolled in their first grade of different upper secondary school programmes. Their mean age was 15.5 years. A little less than half of the sample were female students (N = 272; 45.2%) and more than half were male students (n = 330; 54.8%). Students attended technical education programmes (n = 260; 43.2%), general gymnasium (n = 139; 23.1%), vocational education programmes (n = 95; 15.8%), gymnasia specialist (n = 75; 12.5%), while the least of participating students were from short-term vocational education programmes (n = 33; 5.5%). Instruments Quality of the Student-Teacher Relationship For establishing the quality of student and class teacher relationships, we used the Network of Relationship Inventory questionnaire (NRI, Furman and Buhrmester, 1985). The questionnaire consists of 33 items to which students answered according to a five-point Likert-type scale (from 1 - a little or not at all, to 5 - mostly), and measures the quality of the student's relationship with their parents, their best friend and their teacher. For the purposes of this research only the student's evaluation of the relationship with the teacher were used. The questionnaire includes 11 sub-scales that describe socio-emotional support in a relationship: i) socialising, referring to the frequency of the adolescent's socialising with a certain person (e.g. How much of your free time do you spend with this person?), ii) intimate disclosure in a relationship, referring to the degree of the student's trusting intimate information to a class teacher (e.g. How much do you tell this person?), iii) instrumental aid, referring to the degree of help the student feels they receive from the class teacher (e.g. How much does this person help you figure out or fix things?), iv) nurturance (e.g. How much do you protect and look out for this person?), v) approval, referring to how much the student feels approved, respected and admired by the class teacher or how 120 k. sterman ivancic, m. puklek levpuscek ■ motivational goals much they feel their actions are approved, respected and admired (e.g. How much does this person treat you like you're admired and respected?), vi) reliable alliance, referring to the student's perception of relationship stability and durability (e.g. How sure are you that your relationship will last in spite of fights?), vii) affection that the student feels he receives from the class teacher (e.g. How much does this person really care about you?), viii) satisfaction with a relationship, referring to the student's general evaluation of satisfaction with a relationship with a class teacher (e.g. How satisfied are you with your relationship with this person?), ix) antagonism, referring to the student's perception of tension in a relationships (e.g. How much does this person punish you?) x) conflict, referring to the student's evaluation of contradiction, conflict and quarrel frequency in a relationship (e.g. How much do you and this person disagree and quarrel?), and xi) relative power, referring to the student's perception of his autonomy and subordination within the relationship (e.g. Who tells the other person what to do more often, you or this person?). Each subscale contains three items. In the research literature, we did not identify any information referring to psychometric properties of the already mentioned subscales in a student-teacher relationship. According to the preliminary study results we decided to eliminate the subscale that refers to the student's socialising with the class teacher, because of low inter-item correlations in our sample. After reviewing the content of items we established that the items did not reflect the student-teacher relationship in Slovenian cultural environment. All of the other scales show moderately good internal consistency in our sample, that is the coefficient alpha values range between a = 0.73 and a = 0.87. Furman and Buhrmester (1985) found that the aforementioned subscales constitute two higher-order factors: socio-emotional support and negative interactions in a relationship. The authors did not identify the psychometric properties for those two scales in a student-teacher relationship. In our study these two scales showed moderately good internal consistency with coefficient values a = 0.91 and a = 0.77. The results on individual subscales were calculated for each student by adding the values of all three items that form the scale. For the purpose of comparing student results on different subscales, we calculated the average values of each scale. Missing values in individual items were substituted by the average value of the other two items in the same scale, according to the instructions given by the authors (Furman and Buhrm-ester, 1985), based on the condition that at least two values were available. In case they were not available, the student's answers on that scale were not taken into account. 121 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 The Student's Motivational Goals Students' achievement goals were measured by the Patterns of Adaptive Learning Scales questionnaire (Midgley et al., 2000; Slovenian translation and adaptation Puklek Levpušček & Zupančič, 2009). In the study we used the following three subscales: i) mastery goal orientation (e.g. "It's important that in this school year I gain a lot of knowledge in this subject"), ii)performance-approach goal orientation (e.g. "One of my goals is to show others how successful I am at schoolwork."), and iii) performance-avoidance goal orientation (e.g. One of my goals is to show my classmates that I am not doing poorly at school."). According to authors (Midgley et al., 2000) internal consistency coefficients for those three scales are good: a = 0.86 for mastery goal orientation subscale, a = 0.86 for performance-approach orientation subscale, and a = 0.75 for performance-avoidant orientation subscale. The questionnaire consists of 14 items to which students answered on a five-point Likert-type scale (from 1 - very untrue of me, to 5 - very true of me). The first two subscales consist of five items and the last one consists of four. The results on individual subscales were calculated for each student by adding the values of all the items that form the scale. For the purpose of comparing student results on different subscales we calculated average values of each subscale. Student's Academic Achievement Student's general academic record at the end of the school year was used as a measure of his academic achievement (i-unsufficient, 2-sufficient 3-aver-age, 4-good, 5-excellent). Procedures School headmasters were the first to be invited to the research, and based on their decision to participate they chose a research coordinator. Included in the letter were also consent forms for parents to sign and to confirm the students' voluntary participation in the research. The participation was anonymous; each student was assigned a code by the school. The questionnaires were filled in during class meetings under the supervision of the school psychologist who gave students directions for filling in the questionnaires. Students were given an hour to fill in the questionnaires. The content of the questionnaires and the research process were examined by the Ethics Commission at the Department of Psychology of the University of Ljubljana. 122 K. STERMAN IVANCIC, M. PUKLEK LEVPUSCEK ■ MOTIVATIONAL GOALS ... Results In the results, we first list the descriptive parameters for subscales of the students' socio-emotional support and motivational goals in the whole sample. Then, based on the analysis of variance (ANOVA), we investigate the effects of gender, educational programme, and academic performance on students' perceived socio-emotional support of their class teacher and their achievement goals. Finally we test the predictive model of the effects of gender, educational programme, and perceived socio-emotional support on the students' achievement goals and academic performance by using a multiple regression method. Table 1. Descriptive statistics for the perceived socio-emotional support, achievement goals, and academic achievement N M SD Skewness Kurtosis Statistic Statistic Statistic Statistic SE Statistic SE NRI subscales Conflict 569 1.64 .81 1.61 .I0 2.47 .I9 Instrumental aid 569 2.71 1.01 .31 .I0 -.67 .I9 Satisfaction 569 3.19 1.15 -.06 .I0 -.99 .I9 Intimate disclosure 569 1.65 .75 1.44 .I0 2.°5 .I9 Nurturance 569 2.17 .96 .76 .I0 -.°3 .I9 Affection 569 2.53 I.0I .34 .I0 -.62 .I9 Antagonism 569 2.00 .89 1.24 .I0 L55 .I9 Reassurance ofworth 569 2.74 .99 .23 .I0 -.57 .I9 Relative power 569 2.81 .99 .I9 .I0 -.55 .I9 Reliable alliance 569 2.67 1.09 .4I .I0 -.52 .I9 NRI higher-order factors Socio-emotional support 569 2.49 .8I .33 .I0 -.35 .I9 Negative interactions 569 1.62 .8I I.75 .I0 3.25 .I9 Achievement goals Mastery goals 565 3.80 .79 -.23 .I0 -.35 .20 Performance-approach goals 565 3.05 .95 .II .I0 -.36 .20 Performance-avoidance goals 565 3.43 .887 -.057 .I00 -.403 .200 123 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 N M SD Skewness Kurtosis Statistic Statistic Statistic Statistic SE Statistic SE Academic achievement Final academic success 569 3.27 .91 -.259 .102 .22 .20 Valid N (listwise) 565 Table 1 shows that the students in general evaluated teacher's so-cio-emotional support as relatively low, while the presence of negative interactions in a relationship was, on average, evaluated even lower. They perceived higher reassurance of worth, class teacher instrumental aid, reliable alliance and their affection in a relationship than intimate disclosure. Regarding achievement goal orientation, the students assessed their mastery goals the highest, followed by performance-avoidance goals, while the performance-approach goals were the least present. Perceived Socio-emotional Support of the Class Teacher In the next step, we sought to discover differences in the perceived so-cio-emotional support and negative interactions of the class teacher (the two higher-order NRI factors) among the students from different educational programmes. We excluded students from the short-term vocational programmes due to its poor representativeness (n = 33; 5.5%). We examined the differences in perceived socio-emotional support and negative interactions of student's according to educational programme by one-way ANOVA. The results did not show any significant differences in the perception of socio-emotional support by the class teacher in relation to the student's educational programme: F(3, 565) = 2.143, p = 0.094, MSE = 1.377, partial q2 = 0.016, i-fi = 0.546, but showed significant differences in student's perceived negative interactions in relation to educational programme: F(3, 565) = 6.617,p = 0.000, MSE = 4.251, partial q2 = 0.034, i-fi = 0.973. We have established similar results regarding differences in the perceived socio-emotional support in relation to gender, where girls on average state a somehow higher perceived socio-emotional support from their class teachers than boys (M =2.51, SD = 0.808 vs. M =2.45; SD =0.801), but the differences are nevertheless not statistically significant: F(i, 567) = 0.932, p = 0.335, MSE = 0.603, partial q2 = 0.002, i-fi = 0.161. However, the results point to statistically significant differences amongst boys and girls in perceived negative interactions with their class teacher: F(i, 567) = 47.678, p = 0.000, MSE = 29.14, partial q2 = 0.078, i-fi = 1.00 124 K. STERMAN IVANCIC, M. PUKLEK LEVPUSCEK ■ MOTIVATIONAL GOALS Further, we have also established statistically significant differences in the students' perceived negative interactions with regard to their level of academic performance: F(4, 564) = 17.318, p = 0.000, MSE = 10.275, partial 92 = 0.109, i-fi = 1.000. Again, we did not establish significant differences in perceived socio-emotional support in relation to academic performance: F(4, 564) = 1.598,p = 0.173, MSE = 1.029, partial 92 = 0.011, i-fi = 0.494 Students' Achievement Goals Table 2. Descriptive statistics of the students' achievement goals accord- ing to the educationa prog ramme Mastery goal orientation Performance-approach goal orientation Performance-avoidance goal orientation M SD M SD M SD Gymnasia general 3.83 075 2.87 0.91 3.33 0.89 Gymnasia specialist 3.52 0.82 2.81 0.86 3.23 0.86 Technical education programmes 3.82 0.76 3.02 °.95 3.42 0.89 Vocational programmes 3.81 0.84 3.35 0.90 3.60 0.83 Total 378 0.78 3.01 a93 3.40 0.88 The results in Table 2 show that the students from all educational programmes perceive themselves as highly mastery-goal oriented. This is followed by performance-avoidance goal orientation, while their performance-approach goals were the least present. The results of one-way ANOVA show statistically significant differences in students' mastery goal orientation according to the type of educational programme, albeit the differences are small: ^(3,56) = 3.19, p = 0.02 MSE = 1.93, partial q2 = 0.02, i-fi = 0.74. From all four educational programmes included in the comparison, students from the gymnasia specialist perceive themselves as the least mastery-goal oriented (M = 3.52, SD = 0.82). The results also show statistically significant differences in the students' performance goal orientation: ^(3,52) = 6.53, p = 0.00; MSE = 5.53, partial q2 = 0.03, i-fi = 0.97. The most prominent are differences between gymnasia and vocational educational programmes, where the students from the latter, on average, express higher performance goal orientation (M = 3.02; SD = 0.95 and M = 3.35; SD = 0.90 vs. M = 2.81; SD = 0.86 and M = 2.87; SD = 0.91). The results of one-way ANOVA also indicate statistically significant differences among educational programmes regarding the performance-approach goal orientation: ^(3,56) = 2.98, p = 0.03; MSE = 2.29, partial q2 = 0.02, i-fi = 0.704. The highest perfor- 125 ŠOLSKO POLJE, LETNIK XXVII, ŠTEVILKA 1-2 mance-avoidance goal orientation was expressed by the vocational secondary educational programmes' students, followed by the technical education programmes' students. Table3. Correlations among studied variables Performance-avoid-ance goal orientation , ¿a 1 § c^'G 0 ¡5 (g .2 •T « ^ s s 0 ^ 7 13 41 63 89