Adolescents as Delinquent Actors and as Targets of
VARSTVOSLOVJE, Journal of Criminal Justice and Security year 15
pp. 439-459
Preventive Measures
Anabel Taefi, Thomas Görgen, Benjamin Kraus
Purpose:
The article aims at examining the prevalence of deviance and delinquency in a sample of students, at explaining property and violent offending via risk factors and examining students' experiences with and views on preventive approaches. Design/Methods/Approach:
Data stem from a school survey conducted among 2186 13-17 year old male and female students, who attended school in an urban or a rural northwest area of Germany. The instrument was developed on the basis of the new ISRD-3 questionnaire and included newly developed questions on students' experiences with and views on preventive actors and approaches. Descriptive as well as multivariate methods are applied. Findings:
Deviance and delinquency were found to be widespread, but mostly of low severity. Different predictors for violent and property offending can be found. Groups of students with differential involvement in delinquency show clearly differentiated profiles with regard to risk factors. Evaluations of preventive actors and approaches are very similar across groups of students with differential delinquent involvement. Research Limitations/Implications:
Special schools have been excluded from the sample. Generally, school surveys may fail at including high risk individuals, such as students who skip school. Practical Implications:
Findings hint at the importance of including peers and family in preventive approaches. Originality/Value:
Extension of a self-report study among youngsters as targets of prevention with questions on their experiences and evaluations of preventive approaches may give implications on differential receptiveness of young people for preventive approaches.
UDC: 343.91-053.6
Keywords: self-reports, deviance and delinquency, prevention, youth violence, substance abuse
439
Adolescents as Delinquent Actors and as Targets of Preventive Measures
Mladostniki kot prestopniki in kot ciljne skupine preventivnih ukrepov
Namen prispevka:
Namen prispevka je proučiti razširjenost deviantnosti in prestopništva v izbranem vzorcu dijakov, pojasniti kršitve na področju premoženjske kriminalitete in nasilništva v luči dejavnikov tveganja in proučiti izkušnje dijakov s preventivnimi pristopi ter njihova mnenja o slednjih. Metode:
Podatki izvirajo iz ankete, opravljene v šolah in izvedene na vzorcu 2.186 dijakov obeh spolov, starih med 13-17 let, ki so obiskovali šolo v mestnem in podeželskem okolju severozahodne Nemčije. Instrument temelji na novem vprašalniku ISRD-3 in vključuje nova vprašanja o izkušnjah in pogledih dijakov na dejavnike preprečevanja in zadevne pristope. Uporabljene so opisne in multivariatne metode. Ugotovitve:
Ugotovitve kažejo, da sta deviantnost in prestopništvo zelo razširjena, vendar večinoma v blagih oblikah. Napovedi glede nasilništva in premoženjske kriminalitete je mogoče utemeljiti na različnih dejavnikih. Skupine dijakov so v prestopništvo vpletene prek različnih oblik in kažejo na različne profile glede na dejavnike tveganja. Ne glede na vrsto in težo prestopkov so si vrednotenja dejavnikov preprečevanja in z njo povezanih pristopov v vseh omenjenih skupinah zelo podobna.
Omejitve/uporabnost raziskave:
Šole s posebnim programom so bile iz vzorca izključene. Ankete prav tako ne vključujejo posameznikov z visokim tveganjem, npr. dijakov, ki izostajajo od pouka.
Praktična uporabnost:
Ugotovitve kažejo na pomembnost vključitve vrstnikov in družine v preventivne pristope. Izvirnost/pomembnost prispevka:
Obseg študije, v kateri mladi kot ciljne skupine prevencije prek samonaznanitvenih vprašalnikov odgovarjajo na vprašanja o svojih izkušnjah in oceni preventivnih pristopov, lahko opozori na nekatere vidike vzrokov za razlike glede dovzetnosti mladih za preventivne pristope.
UDK: 343.91-053.6
Ključne besede: samonaznanitev, deviantnost, prestopništvo, preprečevanje, mladoletniško nasilje, zloraba substanc
1 INTRODUCTION
In multiple ways, adolescents are specifically connected to topics of delinquency: Compared to older adults, they are highly involved in many types of delinquent behaviour, both as offenders and as victims. At the same time, they are the main target group of prevention programmes and measures. Based on juveniles' self-
440
Anabel Taefi, Thomas Görgen, Benjamin Kraus
reports, this study has a dual focus both on juvenile offending and on youngsters as persons addressed by prevention efforts. 2186 8th to 10th grade students1 in German schools have been surveyed with a standardized self-report instrument between December 2011 and March 2012. This article presents data on deviance and victimization with special emphasis on students' differential involvement in delinquency. It analyses predictors of violence and property offences and focuses on connections between juvenile involvement in offending on the one hand, and young persons' experiences with prevention and their perceptions of preventive actors and approaches on the other.
2 CURRENT STATE OF RESEARCH
In the last couple of decades, self-report studies (e.g. Baier, Pfeiffer, Simonson, & Rabold, 2009; Enzmann et al., 2010; Junger-Tas et al., 2010) have become an indispensable complement to police and court statistics on crime. Surveys on self-reported delinquency, mostly conducted among easily accessible populations of adolescents and young adults, have multiple strengths that are important for criminological research. First, they go beyond offences reported to law enforcement agencies and provide information on prevalence and incidence of delinquent behaviour. Second, they provide insight into relationships between reported and unreported offences. Third, they have produced substantial knowledge on risk factors and protective factors related to delinquent behaviour.
Among the findings based on self-report studies are the following:
-	Rule-breaking is widespread in adolescence and is part of the process of growing up.
-	The majority of all offences committed by juveniles are of low severity.
-	For most adolescents, delinquent behaviour is a transient phenomenon that does not develop into criminal careers.
-	A small group of juveniles commits a large proportion of all offences, and this concentration is especially strong for serious offences.
-	Risk factors for persistent and serious offending are connected to socialization and family processes, neighbourhood characteristics and social structure, lifestyle and peer behaviour, but also to personal characteristics such as self-control, norms and morality.
-	Serious juvenile offending is rarely an isolated phenomenon but usually connected to other types of problem behaviour such as drug and alcohol abuse, school absenteeism and other forms of deviance.
-	There is a considerable overlap between victim and offender characteristics and populations.
While instruments used in self-report studies touch upon a broad range of topics - criminal behaviour, other types of deviance, family, school, leisure time activities, peers, attitudes and values, victimization, etc. - and also include contacts with police and law enforcement agencies, topics of involvement of youngsters
1 In Germany, these grades cover mainly the age group of 13-17 year old boys and girls.
441
Adolescents as Delinquent Actors and as Targets of Preventive Measures
in crime prevention measures and their perceptions of preventive approaches targeted at them have remained outside the focus of this strand of research.
Prevention is always "prevention as perceived" by the target group. This makes the way juveniles perceive attempts to reduce delinquent behaviour an important topic in studies on juvenile delinquency. The research presented here adds questions on prevention to an advanced type of self-report instrument. It measures juveniles' experiences with preventive approaches on one hand, and the way they think about prevention on the other. Whom do young people consider to be influential, what kind of approaches do they regard as potentially successful, how do they judge the potential impact that school can have on substance abuse and violent behaviour? These aspects are relevant to assess the way in which prevention customers perceive and evaluate measures and actors and can be linked to differential involvement in delinquency.
3 METHODS
The survey was conducted in two neighbouring areas in the federal state of North Rhine-Westphalia. The city of Muenster (290 000 inhabitants, 303 km2) was chosen as an urban area, the county of Warendorf (280 000 inhabitants, 1318 km2) as a rural region. Muenster, the administrative centre of the surrounding region, is home to a large university and multiple other institutions of higher education. More than 80% of the workforce is employed in the tertiary sector. In the county of Warendorf, the secondary sector (mainly mechanical engineering and metal processing) is equally important as the tertiary sector. The percentage of citizens who are not German nationals is slightly lower in the city (6.8%) than in the rural area (7.0%). Nonetheless, if first and second generation migrants are combined, 26.6% of Muenster's and 19.8% of Warendorf's population have a migration background (Ministerium für Arbeit, Integration und Soziales des Landes Nordrhein-Westfalen, 2012).
In the German school system, secondary education is stratified and leads to three divergent qualification levels ("Hauptschule", "Realschule", "Gymnasium"). Additionally, there are schools that teach children in stratified courses or classes under one roof ("Gesamtschule").2
In order to gain access to the sample, approval had to be obtained at multiple levels. The first step was to provide information on the aims and structure of the survey to headmasters/headmistresses of schools, and ask for their willingness to participate in the study. In case of approval, each class teacher decided whether he or she agreed to participate. Afterwards, written consent from parents of underage students needed to be obtained. Participation in the survey was voluntary and anonymous. In total, 19 out of 34 schools agreed to participate. The overall response rate of students in the classes that were participating was 65.9%. Non-response occurred due to lack of parental consent, students being absent at the time of the survey (because of illness or truancy), or students' refusal to participate.
2 Furthermore, there are schools that prepare for vocational training after grade 10 ("Fachoberschule") as well as several types of special schools ("Foerderschule") for children with learning disabilities or other handicaps.
442
Anabel Taefi, Thomas Görgen, Benjamin Kraus
The relative proportions of these different backgrounds of non-response cannot be quantified.
The survey was conducted as a paper and pencil survey in class, and the questionnaire was based on instruments used in the International Self-Reported Delinquency Study (see Enzmann et al., 2010; Junger-Tas et al., 2010, on the second wave ISRD-2, and Junger-Tas, Marshall, & Ribeaud, 2003, on the first wave). Via contacts with the ISRD Steering Committee for the third wave, the newly developed ISRD-3 instrument could be used.3 It was adapted for the specific purposes of the YouPrev survey, and the instrument was shortened in order be applicable in one lesson. Additionally, a section on experiences with and attitudes towards preventive measures and preventive actors was included (http://youprev.eu/pdf/ YouPrev_Instrument_SchoolSurvey_English.pdf).
4 SAMPLE DESCRIPTION
Characteristic	Urban % (n)	Rural % (n)	Total % (n)
Total sample	47.2 (1.031)	52.8 (1.155)	100 (2.180)
Sex: female	46.4 (478)	48.5 (559)	47.5 (1037)
Mean age (in y.)	14.79	14.75	14.77
Students' place of residence			
Large city	69.9 (698)	1.6 (18)	33.5 (716)
Small town	16.2 (162)	56.3 (642)	37.6 (804)
Village	13.8 (138)	42.1 (480)	28.9 (618)
School type			
Gymnasium (high school)	45.9 (473)	29.0 (335)	37.0 (808)
Realschule (junior high school)	26.0 (268)	50.0 (578)	38.7 (846)
Hauptschule (lower secondary school)	28.1 (290)	21.0 (242)	24.3 (532)
Grades			
8th	33.4 (344)	34.9 (403)	34.2 (747)
9th	35.6 (367)	36.2 (418)	35.9 (785)
10th	31.0 (320)	28.9 (334)	29.9 (654)
Migration background			
Migr. backgr. 1st & 2nd generation	31.2 (313)	20.3 (230)	25.4 (543)
Among those: language spoken at home not German	39.0 (115)	21.5 (47)	31.5 (162)
Table 1: Sample characteristics by region, 2138 < n < 2186
Of the 2 186 respondents, 52.8% were attending school in the rural, and 47.2% in the urban area. Nearly half of the students were female (47.5%), mean age was
3 Special thanks to the ISRD Steering Committee and Dr. Dirk Enzmann.
443
Adolescents as Delinquent Actors and as Targets of Preventive Measures
14.77 years (SD = 1.11, Min = 124, Max = 19, n = 2170). Table 1 gives an overview on sample characteristics subdivided by the area in which the school was situated.
Most of the students who attended school in Muenster also lived there (see Table 1), while 30% lived outside of the city in smaller towns or villages. Percentages of students attending Gymnasium and Realschule differed between urban and rural area. Most of the students in Muenster (73.6%) and Warendorf (79.8%) lived with both parents (or a stepparent) and siblings (urban = 79.0%, rural = 86.2%). Over 30% of the students attending school in the city were first or second generation migrants; in the rural area this was only true for around 20% of the sample. The average age of migration to Germany among first generation migrants was six years (SD = 5.074, n = 132). As indicated above, the rate of students with a migration background approximately concurs with the overall figures for this German region.
The majority of students with a migrant background had their origins in states belonging to the former Soviet Union or Poland, another large share came from other European countries, including Turkey. Around 20% of migrant students stemmed from the Middle East, from South/South East Asia or Sub-Saharan Africa. Very few students came from North or Latin America, North Africa or Australia.
5 SELF-REPORTED DELINQUENCY
In the questionnaire, delinquency was surveyed via 16 questions relating to different types of offending and asking for students' life-time as well as twelve-month prevalence of offending. The 16 offence types relate to several forms of property and violent offences, vandalism, illegal downloading and drug-trafficking. Two of the items, carrying a firearm5 and carrying other weapons or weapon-like objects, relate to acts which are not in every case illegal according to German law.
Illegal downloading of music or movies is the most widespread type of offence. 47.3% of all boys and 33.2% of all girls reported at least one illegal download during the past twelve months. Life-time prevalence was 55.3% for boys and 39.0% for girls. 21.1% of all students reported illegal downloading as their only offence during the last twelve months.
The overall rate of self-reported offending is high (cf. Table 2) with a lifetime prevalence of 59% and a 12-month prevalence of 51.6% for the total group of students. No significant differences of life-time and twelve-month prevalence could be found between the two regions, but rates for boys and girls differ significantly.
444
4	Relates to students who attended school at the age of 5 and have skipped another year.
5	In the legal case this could refer e.g. to airsoft guns (some are even legal for minors), air rifles or gas pistols, which are mainly accessible for adults.
Anabel Taefi, Thomas Görgen, Benjamin Kraus
Self-reported offending	Urban % (n)	Rural % (n)	X2	df	P
All offences: life-time prev.	57.5 (576)	60.3 (689)	1.72	1	.190
All offences: 12-month prev.	49.4 (437)	53.5 (548)	3.16	1	.075
Self-reported offending	Boys % (n)	Girls % (n)	X2	df	P
All offences: life-time prev.	67.9 (759)	49.4 (506)	76.35	1	***
All offences: 12-month prev.	60.2 (583)	42.8 (402)	57.89	1	***
Table 2:
Overall rates of self-reported offending by region and sex
Figure 1: Twelve-month prevalence of boys' and girls' self-reported delinquency
Figure 1 gives an overview on the prevalence of offences (except illegal downloading), subdivided by students' sex. Among the offence types shown in Figure 1, the ones reported most often by students were those which supposedly are of lower severity; vandalism was reported by 9.2% of the sample for the last twelve months, among the property offences, shoplifting and bicycle theft were the most prevalent ones (11.1% vs. 7.0%), and among violent offences, participation in a group fight (6.9%) was most wide spread. The share of students who reported having committed more serious offences such as assault (2.4%) or robbery and extortion (1.3%) was comparably small. In total, 19.1% reported they had committed at least one property offence during the last year, and 8.5% stated they committed a violent offence (participation in a group fight, assault, robbery and extortion, hate crime). Apart from shoplifting where differences are not significant, all offences were committed by more boys then by girls.
The overall offence rates demonstrate the ubiquitous nature of juvenile delinquency. Spraying graffiti and other forms of vandalism, shoplifting, theft of a bicycle, carrying a weapon or weapon-like object and illegal downloading may be considered as petty offences, while burglary, theft of personal belongings, of a car or motorbike or from a car, robbery and extortion, participation in group fights, carrying a firearm, assault, hate crimes and drug-trafficking constitute a more severe category. If this distinction is applied to the number of offences reported for_
445
Adolescents as Delinquent Actors and as Targets of Preventive Measures
the last twelve months prior to the survey, 92.3% of all reported incidents belong to the group of petty offences, while only 7.7% were characterized as serious ones.
Students were also asked whether they ever had contact with the police because they did something illegal. 19.6% of all respondents reported such police contacts at least once in their life. Boys (24.9%) experienced this significantly more often than girls (13.9%, x2 = 40.933, df = 1, p < .001), and differences between rural and urban areas were not significant. 83.1% of those who had contact with the police because of doing something forbidden (also) had an encounter during the last twelve months prior to participating in the survey. Of the 424 students who reported a police contact, 392 of them also indicated why they had been in contact with the police (which was asked in an open question format). Table 3 presents an overview on the most frequent reasons. In nearly one third of the cases (31.4%), police contacts were linked to minor traffic offences, committed by riding the bicycle on the wrong side of the road or without a light, for example. To some extent, this may be specific for the regions where the survey was conducted. Situated in the North German Plain, bicycle use is very common in all age groups - and the police are known for frequently controlling bicycle traffic.
Table 3: Most frequent reasons for students' last contact with police, n = 392
Offence	%
Minor traffic offence (bicycle)	31.4
Theft	18.6
Violent offence	11.0
Violation of youth protection regulations	6.6
Vandalism	6.4
Possession of drugs	2.8
Trespassing / breaking and entering	2.3
Shooting with airsoft guns	2.3
Theft - in many cases shoplifting - was the reason for 18.6% of the police contacts. 11% of students' last contacts with the police were linked to violent offences. Another considerable group of students were in contact with the police because of violations of youth protection statutes (6.6%). In most cases, this meant they were approached by the police because of underage drinking. Vandalism was the reason for 6.4% of the contacts with police officers, only a small amount of youngsters were caught possessing drugs (2.8%), trespassing/breaking and entering (2.3%) or shooting with airsoft guns. The 18.6% which are missing in Table 3 contain offences which were only named seldom, such as arson, harassment, cyberbullying, fare evasion or driving without a license.
6 DIFFERENTIAL INVOLVEMENT IN OFFENDING
For all further analyses, three groups with different levels of self-reported delinquency (during the last twelve months) were distinguished. Offence types
446
Anabel Taefi, Thomas Görgen, Benjamin Kraus
accounted for were property offences, violent offences, vandalism, drug-trafficking, carrying of weapons or weapon-like objects. Illegal downloading was excluded; consequently, students who reported illegal downloading as their only offence were classified as non-offenders. The first group consisted of frequent violent offenders (FVO, 2.8%, n = 59), i.e. students who reported five or more violent offences for the period of the last twelve months. The second group was composed of all other offenders (25.9%, n = 556); the third group were non-offenders who did not report any of the offences given in the questionnaire or indicated illegal downloading as their only offence (71.4%, n = 1561).
Differential involvement in delinquency may be linked to differences in exposition to risk factors. Some - mostly familial - factors, were indicators for the presence of social problems and characteristics of own deviant behaviour are displayed subdivided by offender types (cf. Table 4).
	> 5 violent	all offenders	no offence (n = 1561)
Characteristic	offences (FVO) (n = 59)	except FVO (n = 566)	
Male	74.6%	63.4%	47.7%
Age	15.2 y.	14.9 y.	14.7 y.
Attending "Hauptschule" (lower level secondary school)	57.6%	29.9%	21.1%
Mother does not live at home	23.2%	10.0%	9.3%
Father does not live at home	28.6%	23.7%	19.0%
Father has a steady job/is self-employed	80.0%	91.7%	93.4%
Migration background	44.6%	31.2%	22.6%
Language spoken at home not German	22.4%	9.4%	6.7%
Truancy (> 3 entire days during last 12 m.)	42.9%	13.5%	2.3%
> 2x heavily drunk during last 30 days	62.5%	21.4%	4.9%
Cannabis use, last 30 days	40.8%	15.6%	2.3%
Drug use (life-time prev., w/o cannabis)	61.9%	35.1%	8.8%
Deviant Peers	91.5%	86.6%	47.8%
Violent Peers	74.9%	35.0%	9.6%
Table 4: Characteristics and risk factors of offender types (last twelve months)
Offenders are more strained than non-offenders and frequent violent offenders are more strained than other offenders. Not having a mother or father around in one's everyday life may be an indicator of "broken homes"; having an unemployed father can be an indicator of a low socio-economic status. Attending the lowest school type and having a migration background should be seen as constructs that often come along with social marginalization and exclusion and disadvantaged conditions for socialization. Deviant behaviour such as excessive consumption of alcohol and drug use has a much higher prevalence among the two groups of offenders, as well as prevalence of deviant and violent peers, who are also known to be main risk factors for delinquency (see e.g. Farrington, 2008).
447
Adolescents as Delinquent Actors and as Targets of Preventive Measures
Since victimization and offending are closely related, especially at a young age, there should be differences found between the three groups. Figure 2 shows that victimization rates increase with the level of involvement in delinquency. For all victimization types given in the questionnaire, frequent violent offenders have the highest victimization rates. Group differences are very pronounced for violent victimization, except for dating violence, which was mainly reported by girls. Non-offenders have lower rates of victimization than the other offenders. Overall, 74.1% of FVO stated they had become a victim of one the offences during the last 12 months; so did 52.8% of the group of other offenders and 32.1% of the non-offenders.
Figure 2: Self-reported victimization during the last twelve months by offender types
Overall, both offending and frequent violent offending clearly appear to be linked to certain characteristics and risk factors. In the following section, predictors of offending will be examined systematically by multivariate analysis.
7 PREDICTORS OF OFFENDING
In order to determine the effects of predictors of offending, binary logistic regression models were applied. As possible risk factors for deviance, some attitudes and external characteristics have been included in the ISRD questionnaire. Five scales6 on personality, family and neighbourhood risk factors were included in the analyses. The self-control scale was introduced by Grasmick, Tittle, Bursik, and Arneklev (1993, shortened version) in order to test Gottfredson and Hirschi's general theory of crime (Gottfredson & Hirschi, 1990), and has high reliability
448
6 Parental supervision scale had five, the other scales had four answer categories.
Anabel Taefi, Thomas Görgen, Benjamin Kraus
(a = .833).7 The morality scale features "pro-social values" and "shaming" (Wikstrom & Butterworth, 2006; Wikstrom & Svensson, 2010) which are core aspects of Wikstrom's Situational Action Theory of Crime Causation. Reliability is high with a = .778. The adherence to violence-legitimizing norms of masculinity scale was developed by Enzmann & Weteels (2002) on the basis of culture of honour theory (Nisbett & Cohen, 1996). Reliability is high with a = .792. Parental supervision and perception of neighbourhood ("social disorganization") scales have been constructed by the ISRD research group. The parental supervision and control8 scale consists of twelve items that address parental knowledge, child disclosure, parental supervision and whether parents set a time. Reliability of the scale is high with a = .848. Regarding social disorganization of the neighbourhood, three items were added to the ISRD-3 scale; they address perceived security of the neighbourhood, presence of police and possibilities for leisure time activities for youngsters. Reliability was high with a = .760. Further classical risk factors have been included in the regression models.
Since the two main types of offending - property and violent offences - differ in their phenomenology and may be influenced by divergent predictors, two separate explanatory models were tested.
Table 5 gives an overview of the analysis of predictors of violent offending. The regression model on 12 months prevalence of violent offending has a very good model fit with an R2-value of 0.423.
Predictor	P	Exp(B)
Region	.878	.967
Sex (ref. = male)	***	.411
Age	.613	.950
Violence-legitimizing norms of masculinity	**	1.706
Morality / acceptance of norms	**	.476
Low self-control	*	1.501
Social disorganization of neighbourhood	.361	1.263
Low parental supervision	.624	1.087
Drunk (last month)	***	1.526
Cannabis (last month)	**	1.411
Deviant peers (drugs, property offences)	*	2.266
Violent peers	***	5.412
Constant	.079	.032
Table 5: Binary logistic regression on 12-month-prevalence of violent offending, n = 1891, R2 (Nagelkerke) = 0.423
The strongest predictor for violent offending is contact with violent peers (cf. Table 5). High impact of peers who are engaged in violent activities themselves
7	The Cronbach's a-values were calculated on the basis of the YouPrev-dataset.
8	Abbreviated below as "parental supervision scale".	_
449
Adolescents as Delinquent Actors and as Targets of Preventive Measures
may have a special meaning in this sample, as the most prevalent violent offence is participation in group fights. Being male and the frequency of getting drunk also show to be strong predictors. Adherence to violence-legitimizing norms of masculinity, morality, cannabis consumption, self-control and deviant peers are further significant predictors. In this model, offending is not predicted by attending school in an urban or rural area. Age, social disorganization of the neighbourhood and parental supervision do not have a significant impact on whether a student belongs to the group of self-reported violent offenders or not.
The model explaining self-reported property offending during the last twelve months (cf. Table 6) also has a very good model fit with R2 = 0.358.
Table 6: Binary logistic regression on 12-month-prevalence of property offending, n = 1819, R2 (Nagelkerke) = 0.358
Predictor	P	Exp(B)
Region	.169	1.224
Sex (ref. = male)	.656	1.070
Age	.814	1.016
Violence-legitimizing norms of masculinity	.297	1.146
Morality / acceptance of norms	***	.282
Low self-control	.183	1.192
Social disorganization of neighbourhood	.281	1.192
Low parental supervision	**	1.497
Drunk (last month)	***	1.387
Cannabis (last month)	*	1.224
Deviant peers (drugs, property offences)	***	3.434
Violent peers	***	2.254
Constant	.392	.309
Predictors of property offences are morality, contact with violent and with deviant peers, low parental supervision, and consumption of alcohol and drugs. Compared to the model explaining violent offending, sex and adherence to violence-legitimizing norms of masculinity lose their influence. There were only minor differences between property offending of boys and girls, and the adherence to VLNM is conceptually linked to violent offending but not to property offences. Self-control, which may be especially important for violent offences that more often occur on the spur of the moment, is not a significant predictor. Property offending is strongly connected to acceptance of norms as well as to parental supervision and enforcement of norms. Alcohol and drug use, as well as belonging to a deviant and delinquent peer group, are key predictors in both models.
450
Anabel Taefi, Thomas Görgen, Benjamin Kraus
8 STUDENTS' VIEWS ON PREVENTIVE MEASURES AND APPROACHES
In addition to the questions adapted from ISRD, the study included a section addressing students' experiences with and views on preventive measures, especially in a school context. Students were asked what approaches would work in preventing juvenile delinquency and who is important as a preventive actor. Results are displayed in Table 7 and Table 8.
Again, the columns shows mean values of students who, based on their self-reports, were categorized as frequent violent offenders, other offenders or nonoffenders.
All of the approaches given are ranked rather positively. However, while students do not completely oppose punitive approaches, they prefer those approaches which are directed at the improvement of individual problems and reduction of societal marginalization. For all items, support for preventive measures decreases with level of involvement in delinquency.
Approach	M (SD): > 5 violent offences (52 < n < 56)	M (SD): all other offenders (546 < n < 554)	M (SD): no offence (1517 < n < 1522)
Improve their prospects to get a job.	1.87 (.912)	1.79 (.776)	1.74 (.741)
Listen to their sorrows and problems.	2.00 (.934)	1.87 (.834)	1.75 (.767)
Provide good opportunities for leisure time activities.	2.26 (.915)	1.89 (.842)	1.81 (.758)
Provide training for better social behaviour.	2.25 (.998)	2.09 (.883)	1.92 (.786)
Give them a good general education.	2.30 (.972)	2.13 (.903)	1.97 (.791)
Give information on possible consequences.	2.42 (.937)	2.17 (.921)	2.01 (.858)
Provide counselling to their parents.	2.69 (1.058)	2.40 (.974)	2.11 (.851)
Punish them severely when caught.	2.66 (1.116)	2.39 (.984)	2.13 (.876)
Table 7: Students' perceived efficacy of preventive approaches (4-point scale from 1 = works very good to 4 = is rather harmful; items sorted by means of the total sample)
Table 8 shows that in accordance with criminological findings, students perceive their parents and friends to be the most important persons who can keep them away from doing forbidden things. While values for friends and parents given by non-offenders and the group of other offenders are rather similar, larger differences can be found regarding the estimates given by frequent violent offenders. Again, rank orders are similar across groups, but in most cases, offenders perceive the potential influence on their behaviour as smaller than non-offenders. With regard_
451
Adolescents as Delinquent Actors and as Targets of Preventive Measures
to professionals, relatively strong influence is assigned to police; but again, both groups of offenders show lower values. The group of frequent violent offenders attributes a higher influence to sports coaches than the other two groups, while little influence is ascribed to teachers. This may be due to a general age-related opposition to this profession, students' everyday experience of teachers' limited influence on their behaviour, and the perception of teachers as being responsible for education but not for prevention of delinquency.
Table 8: Students' views on the importance of preventive agents (4-point scale from 1 = very important to 4 = unimportant, items sorted by means in the general sample)
	M (SD):	M (SD):	M (SD):
Agent	> 5 violent offences	all other offenders	no offence
	(46 < n < 53)	(522 < n < 545)	(1449 < n < 1510)
Friends	1.43 (.910)	1.42 (.749)	1.35 (.669)
Parents	1.92	1.58	1.42
	(1.152)	(.832)	(.689)
Police	2.80 (1.241)	2.13 (1.031)	1.96 (.910)
Social workers	2.98 (1.120)	2.62 (.944)	2.39 (.926)
Sports coaches	2.67 (1.226)	2.81 (1.047)	2.73 (.986)
Teachers	3.33 (.967)	3.06 (.932)	2.68 (.903)
When asked about the overall influence school can have on keeping students away from substance use and violent behaviour, students in general and especially offenders assign limited influence to school (cf. Table 9). Offenders and nonoffenders are somewhat more positive in their views of school's potential influence on violence than on use and abuse of alcohol and drugs.
Influence of school on ...	M: > 5 violent offences (n = 53)	M: all other offenders (538 < n < 550)	M: no offence (1506 < n < 1508)
Substance consumption	3.68	3.70	3.25
Violent behaviour	3.32	3.38	2.93
Table 9: Students' perceptions of school's potential influence on substance use and violence
(5-point scale Experiences with and Perceptions of Drug Prevention Measures
from 1 = very
strong influence a majority of students (72%) reported having been provided with information influence at all) on alcohol, drugs and other harmful substances during the last twelve months (cf.
Table 10), mainly in school or by their parents. Also, some students stated they received information on the internet or from social workers.
452
Anabel Taefi, Thomas Görgen, Benjamin Kraus
Information on substance use provided?	% (2069 < n < 2096)
No	28.0
Yes, in school by a teacher	43.2
Yes, in school by another person	26.2
Yes, by parents	30.3
Yes, on the internet	19.9
Yes, in a youth centre	3.9
Table 10: Students' experiences with provision of information on substance use during the last 12 months (multiple answers possible)
Those students who were given information on substance abuse were asked for their views on the information provided and on the effects this information had on them. Table 11 gives an overview of students' evaluations of substance abuse prevention measures, subdivided by offender types.
Answer categories	> 5 violent offences (42 < n < 43)	Other offences (399 < n < 412)	No offence (1002 < n < 1028)
I learned new facts about alcohol and drugs.	53.5	54.4	65.8
I learned new facts about the health effects of alcohol and drugs.	61.9	54.0	71.6
It made me curious about some drugs.	35.7	28.2	9.1
It was nothing new to me.	54.8	50.4	35.4
I learned new facts about how to keep away from alcohol and drugs.	35.7	38.6	45.3
I learned new facts on how to help my friends staying / getting away from drugs.	52.4	37.7	38.1
Table 11: Students' views on information provided on substance use/abuse during the last twelve months
(multiple answers possible), percentages of students who responded with
While most students stated they learned new facts about substances and their health effects, a considerable number also said it was nothing new to them. The overlap between both groups may be due to divergent evaluations of multiple sources of information used during the last twelve months. The most interesting results can be found focussing on the answers of both offender groups when contrasted with non-offenders. Around one fourth to one third of them indicated that the information provided made them curious about some drugs; this percentage is considerably higher than in the group of non-offenders. It could be assumed that information that made youngsters curious was less often provided by professionals and more often by friends or via internet, but this is not the case: Students who became more curious mainly received their information by_
453
Adolescents as Delinquent Actors and as Targets of Preventive Measures
professionals and parents. Among all who reported having been more curious after the intervention, male students and those from the rural region were somewhat overrepresented and there is a very high share of persons who report having used multiple substances. Given the question formats used and the cross-sectional nature of the data, it cannot be determined whether the information on substance abuse provided to them really had stimulating effects on substance use. However, it appears that for a minority of students substance abuse prevention via provision of information (i.e. using a mainly cognitive approach) may either be ineffective or rather have detrimental effects. What makes this finding worrisome is the fact that this 15% (overall rate among all students) minority is characterized by a relatively high level of deviant behaviour both in the fields of substance use and violence.
While they show the highest share of persons with unwanted effects of receiving information, more than half of the frequent violent offenders also stated that they learned how to keep their friends away from drugs; when compared to the other two groups, this proportion is very high. Given the perceived influence students attribute to their friends, these skills may be very valuable especially for the highly strained group of frequent violent offenders who have both a high prevalence of substance use and a high ratio of deviant friends who might consume substances themselves.
Experiences with and Perceptions of Violence Prevention Measures
One quarter of all students reported they had participated in a measure aiming at the prevention of violence during the last twelve months. These measures mainly were realized in school (Table 12).
Table 12: Students' participation in violence prevention measures during the last 12 months (multiple answers possible)
Participation in violence prevention measures	% (1886 < n < 2042)
Yes	25.6
No	74.4
Yes, outside of school	4.6
Yes, in school ...	21.9
Training against bullying at school	10.2
Training on how to settle conflicts without violence	18.7
Other activities	4.9
The majority of students evaluated the activities they participated in as helpful and useful (Table 13).9 A questionable effect seems to be that 90% of frequent violent offenders who participated in prevention measures reported that they learned how to protect themselves from attacks (see Table 13), and 78% said they learned about what to do if under attack. This suggests that violence prevention measures in the
454
9 Results have to be interpreted with consideration of the very small number of FVO who responded
to the questions.
Anabel Taefi, Thomas Görgen, Benjamin Kraus
school context may be adapted by highly delinquent youths in an instrumental manner to boost their potential for defence. Given the elevated victimization rates among FVO, this group can be considered to be especially receptive for such messages and interpretations.
Most of the students agreed with the statement that the measure showed them the negative consequences of violence for themselves and others. Nearly half of the students stated they felt more secure after the measure. The frequent violent offenders may have the highest ratio of persons who agreed to the statement that their way of thinking about violence had been changed, but at the same time they have the lowest values of consent to the statement that they learned to solve problems without use of force or that they learned about the consequences of their actions.
Another unintended effect of participation in a violence prevention measure may be a risen feeling of insecurity. Large differences between the three groups can be found regarding this effect; especially the difference between non-offenders (14.3%) and frequent violent offenders (47.4%) is high. The increased feeling of insecurity of frequent violent offenders may be based on perceived better defence skills by their peers and the bigger awareness of possible punishments. Types of measures were similar across the three groups.
Answer categories	% > 5 violent offences (18 < n < 19)	% All other offenders (134 < n < 138)	% no offence (364 < n < 375)
Helpful to protect myself from attacks by others.	89.5	59.9	70.0
Taught me how to intervene when I see violence against others.	73.7	68.8	69.0
Changed my way of thinking about violence.	57.9	44.4	42.9
Taught me how to resolve problems without violence.	55.6	65.7	74.5
Taught me what to do if somebody tries to attack me.	77.8	59.9	59.7
Provided information on where to turn to when I am under threat by others.	68.4	63.2	73.0
Made me feel more secure.	47.4	48.5	49.7
Made me feel more insecure.	47.4	20.7	14.3
Made me more aware of how violence harms people.	63.2	54.7	71.7
Made me more aware of possible punishments and other consequences.	52.6	62.7	66.5
Table 13: Students' views on effects of violence prevention measures they participated in during the last twelve months (multiple answers possible), percentages of students who responded with
455
Adolescents as Delinquent Actors and as Targets of Preventive Measures
Students were asked for their own ideas on how to prevent substance use and violence if they were themselves teachers. Especially with regard to the prevention of violence, bandwidth and heterogeneity of suggested measures were large. Students suggested providing information on effects of violence on the victim's side and consequences on the offenders' side, as well as providing information on alternative modes of conflict resolution. Many students also suggested encouraging talks, communication and mediation between persons involved in violent incidents. A smaller number of students stated they would clarify norms, rules, and collectively ban violence in the school context. Strengthening relevant skills and resources via training, e.g. anger control training for offenders, self-defence classes for victims or conflict mediation courses were concrete approaches named by some students. Some boys and girls indicated that in case of violence there should be strict sanctions (judicial ones as well as sanctions in the school context). Other students pleaded for involving the parents of offenders, and some pointed out that teachers should talk to students involved in violent incidents in order to understand the causes and problems underlying their behaviour and provide support. Furthermore, some students recommended improving the atmosphere in class and building trust between students and teachers, but some also suggested non-intervention, as teachers can do nothing about violent offending of students in any case. Interestingly, responses of students who have been categorized as frequent violent offenders show nearly the same variety, and their ideas on how to prevent violence coincide with the approaches non-offenders suggest. All three groups recommended providing information about the effects of violence, deterrence by clear sanctioning, communication with the conflict parties and strengthening skills for peaceful conflict resolution. As opposed to the other two groups, frequent violent offenders pointed out that possibilities for interventions are limited and teachers cannot do anything against the violence among students.
Diversity of suggested approaches to reduce substance use was somewhat smaller: The main measures suggested were to provide information on substances and the possible consequences of substance use for health and social development, to deter students via negative examples (e.g. by inviting former substance abusers to school) and to talk with substance abusing students in order to understand the underlying causes and problems and to be able to provide support. Furthermore, students suggested sanctions and drug and alcohol controls in schools if rules are broken. Again, they recommended involving the parents but also to provide leisure time and sports activities offered by the school. Some students had different ideas, and pointed out that school should refrain from prevention and intervention in this field - either because these school measures are regarded as inefficient and drug abuse prevention is not seen as teachers' business, or because students claim a right to self-harm.
9 CONCLUSIONS
Limitations of the study mainly relate to the problems all school surveys have to face: Even though they reach a large number of participants they may miss
456
Anabel Taefi, Thomas Görgen, Benjamin Kraus
information from students who skip school or of those whose parents did not allow their participation. Also, special schools have been excluded from the sample. Furthermore, the sample is not a nationally representative one but was recruited in two neighbouring regions in the Northwest of Germany. The similarity of urban and rural data must be interpreted with regard to proximity and similarity of the urban and the rural region and the fact that a considerable number of students regularly commute between both spaces.
Most findings from the German school survey are in accordance with what is known from other recent self-report studies: Juvenile delinquency is widespread and mostly of low severity. Overall, offences are mainly committed by youngsters, but girls are highly engaged in property offences, especially theft. There is a small group of violent offenders which feature many risk factors for persistent offending and they are accountable for the majority of all reported offences. Victimization rates of students are quite high; especially theft, cyberbullying and dating violence have often been experienced.
The YouPrev school survey showed that self-report studies can be extended to include young people's views of and experiences with prevention. Juveniles hold relatively elaborated concepts of preventive approaches to be initiated by school. These concepts partially mirror approaches endorsed and taken in prevention practice.
The survey results indicate significant links between offending and the way preventive measures and approaches are seen and interpreted by juveniles. On one hand, perspectives of delinquent and non-delinquent youths have much in common. They regard parents and peers as the most important prevention agents, and are not completely opposed to prevention via deterrence but they consider measures targeted at a delinquent person's individual problems and at reduction of social marginalization to have stronger beneficial effects. Offenders and nonoffenders share the view that school may have more influence on violence than on students' use of legal and illegal substances.
On the other hand, skepticism towards potential preventive effects increases with involvement in delinquency. Offenders, and especially frequent violent offenders, attribute less influence to preventive measures than non-offenders do; these results are similar for all different kinds of potential measures that were given in the survey. In the same way, offenders in comparison with non-offenders attribute less importance to actors who might have preventive influence on them. Peers appear to be the most important preventive actors in the eyes of juveniles. Preventive strategies should consider that juveniles are socialized in peer groups to a large extent; approaches that aim at peer groups could be promising alternatives and supplements to preventive efforts which are mainly targeted at individual problematic youngsters.
Finally, the results point to possible differential effects of preventive efforts for groups with different levels of involvement in delinquency. While most self-reported effects of prevention measures in which youngsters have participated are in an intended direction, smaller groups of participants report unwanted or critical effects such as a heightened level of curiosity towards drugs or an increase
457
Adolescents as Delinquent Actors and as Targets of Preventive Measures
in knowledge on how to defend against attacks among youngsters classified as frequent violent offenders.
The study shows that surveying youngsters about their perceptions of preventive measures can provide valuable results that preventive strategies may take into account. Juveniles' answers give hints as to their accessibility for prevention, as well as on the fact that preventive measures are understood and utilized by youngsters in the light of their own experiences and needs and not necessarily in the way intended by those designing and implementing the programmes or measures.
REFERENCES
Baier, D., Pfeiffer, C., Simonson, J., & Rabold, S. (2009). Jugendliche in Deutschland als Opfer und Täter von Gewalt - Erkenntnisse einer deutschlandweiten Repräsentativbefragung [Juveniles in Germany as victims of violence and as violent offenders - Findings from a representative nationwide German survey]. Zeitschrift für Jugendkriminalrecht und Jugendhilfe, 20(2), 112-119. Enzmann, D., & Weteels, P. (2002). Jugenddelinquenz und Ethnizität: Die Rolle gewaltlegitimierender Männlichkeitsnormen [Juvenile delinquency and ethnicity: The role of violence legitimizing norms of masculinity]. In M. Osterheider (Ed.), Forensik 2001. Innovative Konzepte. 16. Eickelborner Fachtagung (pp. 14-33). Dortmund: PsychoGen Verlag. Enzmann, D., Marshall, I. H., Killias, M., Junger-Tas, J., Steketee, M., & Gruszczynska, B. (2010). Self-reported youth delinquency in Europe and beyond: First results of the Second International Self-report Delinquency Study in the context of police and victimization data. European Journal of Criminology, 7(2), 159-183. Farrington, D. P. (Ed.). (2008). Integrated developmental & life-course theories of offending.
New Brunswick: Transaction Publishers. Gottfredson, M., & Hirschi, T. (1990). A general theory of crime. Stanford: Stanford University Press.
Grasmick, H. G., Tittle, C. R., Bursik, R. J., & Arneklev, B. J. (1993). Testing the core empirical implications of Gottfredson and Hirschi's general theory of crime. Journal of Research in Crime and Delinquency, 30(1), 5-29. Junger-Tas, J., Marshall, I. H., & Ribeaud, D. (2003). Delinquency in international perspective: The International Self-reported Delinquency Study (ISRD). Monsey; The Hague: Criminal Justice Press & Kugler Publications. Junger-Tas, J., Marshall, I. H., Enzmann, D., Killias, M., Steketee, M., & Gruszczynska, B. (Eds.). (2010). Juvenile delinquency in Europe and beyond: Results of the Second International Self-report Delinquency Study (ISRD2). New York: Springer. Ministerium für Arbeit, Integration und Soziales des Landes Nordrhein-Westfalen. (2012). Zuwanderungsstatistik Nordrhein 2011 [Migration statistics North Rhine-Westphalia]. Retrieved from http://www.mais.nrw.de/08_PDF/003_ Integration/003_zuwanderung/zuwanderungsstatistik_nrw-2011.pdf Nisbett, R. E., & Cohen, D. (1996). Culture of honor: The psychology of violence in the South. Boulder: Westview Press.
458
Anabel Taefi, Thomas Görgen, Benjamin Kraus
Wikström, P.-O. H., & Butterworth, D. (2006) Adolescent crime: Individual differences
and lifestyles. Cullompton: Willan. Wikström, P.-O. H., & Svensson, R. (2010). When does self-control matter? The interaction between morality and self-control in crime causation. European Journal of Criminology, 7(5), 395-410.
About the Authors:
Anabel Taefi, sociologist, is a research associate at German Police University (Muenster). Her main fields of research are juvenile delinquency, youth correctional facilities, developmental and life-course criminology, domestic violence in later life. E-mail: anabel.taefi@dhpol.de
Thomas Görgen, Ph.D., is a professor at German Police University (Muenster) and head of the Department of Criminology and Crime Prevention. His research interests include topics of juvenile delinquency, victimization in later life, domestic violence, homicide, and crime and violence prevention. E-mail: thomas.goergen@ dhpol.de
Benjamin Kraus, sociologist, is a research associate at German Police University (Muenster). His main fields of research are juvenile delinquency, victimization in later life, homicide, and crime and violence prevention. E-mail: benjamin.kraus@ dhpol.de
459