Anthropos 56 (1): 213–225 | issn0587-5161 | e-issn 2630-4082
Effect of Cognitive Reflection
on Escalation of Commitment
Petra Rekar
phd student at the University of Ljubljana, Slovenia
petra.rekar@gmail.com
Mitja Perat
Inštitut za razvoj človeških virov, Slovenia
mitjaperat@institut.burnout.si
© 2024 Petra Rekar and Mitja Perat
Abstract. Sunk costs are known to be one of the drivers of escalation of
commitment. One explanation for the effect is the overgeneralization
of the waste not rule, or the utilization of automatic Type 1 processing,
when themore deliberate Type 2 would be needed. The objective of the
studywas to test if the cognitive reflection test (crt), as ameasure of an
individual’s tendency to engage in analytical thinking, is related to the
escalation of commitment. We hypothesized that individuals scoring
lower on the testwill exhibit the tendency to honour sunk costs.Using a
continuation with investment problemwe found thatcrt is not related
to decision making which includes sunk cost. We can infer that crt is
not the only or the main predictor of the escalation of commitment.
Key Words: decision making, cognitive reflection, escalation of com-
mitment, sunk costs, analytical thinking
Učinek kognitivne refleksije na stopnjevanje zavezanosti
Povzetek.Poslovno nepovratni stroški so dejavnik, ki vpliva na stopnje-
vanje zavezanosti k izbiri. Ena od razlag za nastanek učinka je preti-
rana posplošitev pravila, da se z denarjem ne razmetava, oz. uporaba
avtomatiziranih hevristik prvega reda, takrat ko bi bila potrebna upo-
raba analitičnega razmišljanja. Namen raziskave je preverjanje, ali se
test kognitivne refleksije, kot mera posameznikove nagnjenosti k ana-
litičnemu razmišljanju, povezuje z zavezanostjo k določeni izbiri. Pred-
videvali smo, da bodo posamezniki, ki bodo dosegli nižjo mero kogni-
tivne refleksije, v večji meri upoštevali nepovratne stroške pri investi-
cijskih odločitvah. Pri nalogi nadaljevanja z investicijo smo ugotovili,
da se test ne povezuje z odločanjem, ki vsebuje nepovratne stroške. Na
osnovi tega sklepamo, da testa kognitivne refleksije ne moremo upo-
rabiti kot prediktorja za ugotavljanje stopnjevanja zavezanosti.
https://doi.org/10.26493/2630-4082.56.213-225
Petra Rekar and Mitja Perat
Ključne besede: odločanje, kognitivna refleksija, stopnjevanje zaveza-
nosti, poslovno nepovratni stroški, analitično razmišljanje
Literature Review
Honouring sunk costs is a well-documented effect, where an individual
continues with an investment or endeavour once an initial amount of as-
sets (money, time or effort) have been allocated to it (Arkes and Blumer
1985). This type of behaviour goes against the standard economic theory
(VonNeumann andMorgenstern 1947), which is normative in its essence
and provides a norm that is the correct course of action. The normative
approach to decisionmaking provides standards of how decisions should
be made, with the aim to assist individuals to maximize the outcome. Ac-
cording to economic theory, the normatively correct decision is that only
future (or incremental) revenues and costs should be considered when
evaluating an investment (or project). A sunk cost has already occurred,
is irrecoverable and as such is not affected by the decision, hence it should
not be considered when evaluating an investment. However, research in-
dicates this is not the case and that sunk costs commonly drive the es-
calation of commitment (Schmidt and Calantone 2002; Sleesman et al.
2012; Soman and Gourville 2001) in finance (Tan and Yates 2002), cus-
tomer loyalty (Jang, Mattila, and Bai 2007; Liang, Lee, and Tung 2014), or
the continuation of an activity (Åstebro, Jeffrey, and Adomdza 2007). Fa-
miliarity with economic theory or with rational decision making alone
does not always reduce or eliminate this effect (Fennema and Perkins
2007; Roth, Robbert, and Straus 2015), nor is it reduced by one’s cognitive
ability (Haita-Falah 2017; Stanovich and West 2008). On the contrary, it
has been argued that being more analytical might actually amplify the ef-
fect, as individuals who make logical arguments for a prior investment
might be more inclined to allocate additional resources (Wong, Kwong,
and Ng 2008). The effect is well researched from the angles of instruc-
tion (Tan andYates 1995), frame (Klaczynski 2001; Salter and Sharp 2001),
the level of sunk costs as the percentage of the total investment (Garland
1990), and the completion of an investment (Boehne and Paese 2000), as
well as some differences in the characteristics of the individual. The ef-
fect is strengthened by state orientation (van Putten, Zeelenberg, and van
Dijk 2010), agreeableness and conscientiousness (Fujino et al. 2016), and
younger age (Bruine de Bruin, Strough, and Parker 2014; Eberhardt, Bru-
ine de Bruin, and Strough 2019). It is believed that experience reduces the
effect in older individuals; however, the type of sunk cost appears to play
214
Effect of Cognitive Reflection on Escalation of Commitment
a key role. When the cost is monetary, experience plays a positive role;
however, experienced inventors who receive a negative review of their
idea (costs were effort and time) continue to invest more assets (money
and time) in further development than their more inexperienced coun-
terparts (Åstebro, Jeffrey, and Adomdza 2007). Domain-specific knowl-
edge is believed to be helpful, but only when the situation cues its use.
In one of the studies, individuals with knowledge of standard economic
theory (cpas, mbas, and accounting students) made more normatively
correct decisions when dealing with a clearly economic decision problem
(Fennema and Perkins 2007). In a similar manner, providing individuals
with training in the sunk cost rule positively affected normative decision
making (Larrick, Morgan, and Nisbett 1990).
There appears to be several explanations of the causes, ranging from
mental accounting (Thaler 1999), loss aversion (Arkes and Blumer 1985),
commitment to project completion (Boehne and Paese 2000), effort jus-
tification (Cunha and Caldieraro 2009), agency (Harrison and Harrell
1993), or the waste not rule (Arkes and Ayton 1999). Although depending
on circumstances, the type of sunk costs (money, time, effort), or indi-
viduals’ characteristics, any of these mechanisms could explain the ef-
fect. In the case of a monetary sunk cost, an individual acquainted with
standard economic theory should be able to make a normatively correct
decision, assuming they either reflect on the task (engage in analytical
thinking), or they obtained sufficient relevant experience, which would
facilitate a correct automatic response. Arkes and Ayton (1999) propose
that the sunk cost effect is a result of the overgeneralization of the ‘do
not waste’ rule, where ceasing the project would be considered a waste
of money already spent. Essentially, they posit that the effect is a result
of the so-called Type 1 processing. In cognitive science, there is a con-
sensus that individuals switch between two qualitatively different modes
of processing, namely the autonomous and fast Type 1 and the more ef-
fortful and slow Type 2. By default, individuals have a tendency to utilize
Type 1, as it is cognitively less costly; however, this type of thinking can
lead to ‘irrational’ decisions (Evans and Stanovich 2013; Toplak,West, and
Stanovich 2014), as it is dependent on the cues from the environment, as
well as intrinsic factors (such as motivation or mood) (Klaczynski 2001).
The cognitive reflection test (crt) is a popularmeasure to distinguish be-
tween individuals more prone to Type 1 processing (automatic, fast) and
those who aremore analytical. In spite of its wide use, it is stable over time
(Stagnaro, Pennycook, and Rand 2018) and robust to multiple exposures
215
Petra Rekar and Mitja Perat
(Bialek and Pennycook 2017). The test is composed of 3 logical problems,
which cue incorrect ‘intuitive’ answers that need to be suppressed to de-
liberate on the correct answer (Frederick 2005). The higher the number
of correct answers, the more analytical an individual is. The test seems to
be a good predictor of performance in tasks that require engagement in
more effortful deliberation, such as heuristics and biases (Frederick 2005;
Toplak, West, and Stanovich 2011). Additionally, it is related to impres-
sionmanagement, with colourmanipulations affecting judgements of less
analytical (more impulsive) individuals (Cardoso, Leite, and de Aquino
2018). Stanovich (2012) proposes that deviations fromnormative decision
making may be the result of individual differences in thinking disposi-
tions. To successfully perform on a variety of heuristics and biases tasks,
an individual needs to first detect that there is a need to override Type 1
processing and inhibition, but they also need to possess the right mind-
ware, including knowledge (Stanovich 2018; Stanovich and West 2008).
As crt requires an individual to suppress their initial (intuitive, though
incorrect) response, Ronayne, Sgroi and Tuckwell (2021) showed that the
capacity for reflection predicts sunk cost effect, where the ‘cost’ is effort
exerted.
To elicit the sunk cost effect, two types of tasks are commonly used –
utilization tasks, where an individual needs to select between two equally
appealing options – and progress decisions, where an individual needs to
decide whether to escalate a commitment. To measure the effect of mon-
etary sunk cost, utilization decisions could be potentially problematic,
as they might not be recognized as economic decisions and individuals
might not apply domain-specific knowledge. In the case of progress (or
the escalation of commitment) decisions, individuals should be aware of
the type of decision they are making and apply the necessary knowledge
(Roth, Robbert, and Straus 2015), assuming the context cues relevant do-
main knowledge.
The aim of the study was to inspect whether more analytical individ-
uals would more often reach a normatively correct decision in an esca-
lation of commitment task. We hypothesized that if the sunk cost effect
stems from the overgeneralization of the waste not rule, individuals with
a higher propensity for analytical thinking should not succumb to sunk
costs. We used the cognitive reflection test as a measure of propensity for
analytical thinking, as it is believed to be associated with good decision
making, when deliberation is paramount, such as on heuristics and biases
tasks. However contrary to research so far, we failed to find an associa-
216
Effect of Cognitive Reflection on Escalation of Commitment
tion betweencrtand sunk costs, which is a classic heuristic in the finan-
cial sphere. We demonstrated that simply being more analytical (scoring
higher on thecrt), even when possessing the right knowledge, is not suf-
ficient to overcome the fallacy.
Participants
A total of 188 participants were included in the study (127 women; 61
men;mean age = 19.54, sd= 4.24). A convenience sample was used, where
the majority of the participants (n = 164) came from a pool of 4th year
students of an economics high school, while a minority (n = 24) were
professionals working in finance with degrees in economics or a simi-
lar subject. In the case of the students, we were allowed to modify their
curriculum by including information regarding the valuation of invest-
ments (which included the treatment of sunk costs); however, no partic-
ular stress was placed on this information. This information was given
roughly 5–6 weeks before the testing took place to ensure that it was not
too fresh in their memory and that they could make the connection. As
the valuation of projects is a part of the standard curricula at the uni-
versity level, it was assumed that professionals had this knowledge. All
the materials were in the Slovene language as all of the participants were
native speakers.
Methodology
For the students, testing took place during their regular economics class.
Upon being informed about the study they were asked to give their
consent. The experiment was a paper and pencil one and consisted of
a short financial assignment, a crt test and a demographics question-
naire. Firstly, the participants were presentedwith an investment problem
which involved sunk costs, modelled after the airline company problem
by Arkes and Blumer (1985), where they had to decide whether to con-
tinue with an investment of the development of a virtual reality headset
when the competition had just launched a similar product, while their
company would need 3 more years to complete it. A certain amount of
sunk costs related to the development so far had already been incurred,
which was clearly stated. We included the information on expected rev-
enues and the uncertainty associated with them, as this information
would normally be available. Moreover, the expected revenues are vi-
tal information as it enables individuals to set a budget without using
subjective expectations about the value of the investment (Heath 1995).
217
Petra Rekar and Mitja Perat
Table 1 Results of Statistical Analysis
Item Whole sample Women Men Students Employees
n     
Mean . . . . .
Std. error mean . . . . .
Median . . . . .
Standard deviation . . . . .
Variance . . . . .
Minimum     
Maximum     
Skewness . . . . –.
Std. error skewness . . . . .
Kurtosis –. . –. . –.
Std. error kurtosis . . . . .
Table 2 Chi-Square Statistics for Sunk Cost Task by Gender
Sunk cost task Women Men χ2()
n  n 
Incorrect answer  .  . .
Correct answer  .  .
notes p = 0.870.
The participants had to decide whether they would continue with this
investment. The correct solution was to continue with the project (since
the projected revenues surpassed the projected costs) and disregard the
sunk costs. The use of calculators was not allowed as only a very sim-
ple calculation was needed (adding/subtracting digits up to 10). Upon
completion, they were given a 3-item cognitive reflection test (Freder-
ick 2005), which was translated into Slovene using forward/backward
translation. Although some have questioned the reliability of the test on
adolescents, as it could be too difficult (Primi et al. 2016), we deemed it
would be appropriate given that all participants were at least 18 years old
at the time of taking the test. Lastly, we collected demographic data.
Results
Firstly, we inspected for any differences in the sample, where we found
no differences between participants in reaching the correct decision. Em-
ployees did not perform any better than students (χ2(1, n = 188) = 1.086,
p = 0.297). Additionally, we also found no difference with respect to gen-
218
Effect of Cognitive Reflection on Escalation of Commitment
Table 3 Chi-Square Statistics for Sunk Cost Task by Participants
Sunk cost task Students Employees χ2()
n  n 
Incorrect answer  .  . .
Correct answer  .  .
notes p = 0.297.
Table 4 Chi-Square Statistics Split by crtScore
Sunk cost task crt_Score χ2()
   
Incorrect answer     .
Correct answer    
notes p = 0.691.
Table 5 Chi-Square Statistics for Participants with Different Levels of Cognitive
Reflection
Item Incorrect Correct χ2() p
answer answer
High_analytical  
. .
Rest  
Low_analytical  
. .
Rest  
Mid_analytical  
. .
Rest  
der (χ2(1, n = 188) = 0.027, p = 0.870). To test our assumption that more
analytical individuals would more often reach normatively correct solu-
tions (would not honour sunk costs), we started by calculating the par-
ticipants’ crt score. In line with other reports (Campitelli and Labollita
2010; Frederick 2005), 46.3 of the participants correctly solved at least
one problem. The average number of correct answers was 0.84 (sd =
1.073), with 21.3 solving one problem correctly, 12.2 solving two prob-
lems correctly and 12.8 solving all three problems correctly.We did find
some differences in how well the two groups scored on the test; employ-
ees performed better, correctly solving 1.92 tasks (students 0.68), which
was statistically significant (χ2(3, n = 188) = 35.694, p = 0.000). We ap-
plied aChi-Square test (where we separated individuals based on number
of points on the test), which rejected our hypothesis. We found that crt
is not related to the escalation of a commitment in the presence of sunk
219
Petra Rekar and Mitja Perat
costs, (χ2(3, n = 188) = 1.463, p = 0.691). To confirm the results, we made
a distinction proposed by Frederick (2005) on highly analytical partici-
pants (who solved all 3 problems correctly), medium (those who solved
1 or 2 problems correctly) and low analytical participants (who did not
solve any problem correctly). Dividing the participants in this way (by
separating the extremes) corroborated our initial results. For highly an-
alytical participants, we obtained (χ2(1, n = 188) = 0.028, p = 0.867), for
low analytical ones (χ2(1, n = 188) = 0.382, p = 0.537), and for participants
classified as medium analytical (χ2(1, n = 188) = 0.595, p = 0.441).
General Discussion
Wewanted to inspectwhether the disposition towards analytical thinking
can explain the results on a monetary sunk cost task in a group of indi-
viduals acquainted with standard economic theory. With respect to sunk
cost fallacy being related to the age of participants, our result is contrary
to other studies, where older participants were less likely to continue with
the commitment to a failing plan (Bruine de Bruin, Strough, and Parker
2014; Eberhardt, Bruine de Bruin, and Strough 2019; Karns 2012). Our
results did not find a connection between reaching normatively correct
decisions and acrtscore, implying that analytical thinking is neither the
main nor only factor preventing the sunk cost effect. These results are in
line with some previous research, which shows that the sunk cost fallacy
is not related to measures of executive function processes (Del Missier,
Mäntylä, and Bruine De Bruin 2012). Our findings are unlike those of
Ronayne and colleagues (2021); however, this could be due to a differ-
ence in the task used. The propensity for analytical thinking might be a
good predictor when sunk cost is in the form of effort expended, where
it would be sufficient for an individual to pause and consider the expen-
diture of effort so far vis-à-vis alternatives before continuing. However,
in the case of money spent, reflection (switching to analytical thinking)
alonemight not be sufficient, even when the task is clearly from a domain
the participants are familiar with and they should apply knowledge-based
decision making. It is likelier that the interplay of several factors deter-
mines whether an individual will behave normatively correctly on such
tasks. In our sample, even though the participants would have learned
about sunk costs before, we cannot rule out the lack of appropriatemind-
ware. Individuals first need to recognize the task and then apply the nec-
essary knowledge, implying a certain level of the inhibition of impulse.
As crt is a measure of inhibition, we only controlled at the level when
220
Effect of Cognitive Reflection on Escalation of Commitment
an individual already needs to apply knowledge, whereas we did not con-
trol for how the participants interpreted the task. Although instructions
pointed out that the assignment had a correct answer, we did not empha-
size that it should be solved based on the knowledge they had received in
class or during schooling, to keep it in line with decision making in real-
ity. It might be that participants interpreted the assignment as if we were
asking for their preference and they had not applied the necessary knowl-
edge in the first place. Although thematerials were pilot tested andwe did
not receive any ambiguity regarding the interpretation of the questions,
such an explanation cannot be ruled out. In the cases where participants
did not have much practical experience treating (monetary) sunk cost,
a problem of task recognition might arise, which results in the applica-
tion of the wrong decision rules, indicating that the use of knowledge in
less experienced individuals needs to be prompted by the environment to
prevent the sunk cost fallacy.
Furthermore, some researchers suggest that either general or statistical
numeracy has greater predictive power in superior decisionmaking than
crt(Cokely et al. 2018; Sobkow,Olszewska, andTraczyk 2020). Although
we collected the average grade in mathematics class, we did not control
specifically for this type of numeracy, which is a limiting factor in our
research.
Certain design features also limit the ecological validity. Firstly, in real-
ity sunk costs would normally be known to a decisionmaker and not nec-
essarily explicitly stated in the analysis. By explicitly stating sunk costs,
the saliency of this information might override the application of rele-
vant knowledge. Secondly, although the use of hypothetical scenarios is
common in such settings, it might be problematic. The stakes are clearly
not real, which might reduce the motivation that would be present in a
real world setting. Participants need to position themselves in the role
of a ceo; moreover, the amounts of money (millions) can be too ab-
stract. However, making the assignment more familiar (e.g. by using a
smaller amount, investment in common items) can also be problematic,
as it might be unduly influenced by recent experiences.
The biggest hurdle to the generalizability of our results is the sample
size and composition. Although students are commonly used and they
were supplemented with participants who work in the field of finance,
our samplewas relatively young and perhaps lacking some experience. To
counterbalance this, the sample would need to include older individuals
with more work experience in the field of finance. The task also required
221
Petra Rekar and Mitja Perat
amoderate cognitive effort, meaning participants needed to bemotivated
to engage in solving it, though we do not deem this as particularly prob-
lematic, as participation was voluntary and they could withdraw at any
point.
Although the area of the escalation of commitment received signifi-
cant interest, it might be prudent to inspect the effect by using new in-
struments that would provide a slightly different view and contribute to
a better understanding of the phenomena.
References
Arkes, Hal R., and Peter Ayton. 1999. ‘The Sunk Cost and Concorde Effects:
Are Humans Less Rational Than Lower Animals?’ Psychological Bulletin
125 (5): 591–600.
Arkes, Hal R., and Catherine Blumer. 1985. ‘The Psychology of Sunk Cost.’ Or-
ganizational Behavior and Human Decision Processes 35 (1): 124–140.
Åstebro, Thomas, Scott A. Jeffrey, and Gordon K. Adomdza. 2007. ‘Inventor
Perseverance after Being Told to Quit: The Role of Cognitive Biases.’ Jour-
nal of Behavioral Decision Making 20 (3): 253–272.
Bialek, Michael, and Gordon Pennycook. 2017. ‘The Cognitive Reflection Test
Is Robust to Multiple Exposures.’ Behavior Research Methods 50:1953–1959.
Boehne, DonnaM., and Paul W. Paese. 2000. ‘DecidingWhether to Complete
or Terminate an Unfinished Project: A Strong Test of the Project Comple-
tion Hypothesis.’ Organizational Behavior and Human Decision Processes
81 (2): 178–94.
Bruine de Bruin,Wändi, JoNell Strough, andAndrewM. Parker. 2014. ‘Getting
Older Isn’t All That Bad: Better Decisions and CopingWhen Facing “Sunk
Costs.”’ Psychology and Aging 29 (3): 642–647.
Campitelli, Guillermo, and Martín Labollita. 2010. ‘Correlations of Cognitive
Reflection with Judgments and Choices.’ Judgment and Decision Making 5
(3): 182–191.
Cardoso, RicardoLopes, Rodrigo deOliveira Leite, andAndréCarlos Busanelli
de Aquino. 2018. ‘The Effect of Cognitive Reflection on the Efficacy of Im-
pressionManagement: An Experimental Analysis with Financial Analysts.’
Accounting, Auditing & Accountability Journal 31 (6): 1668–1690.
Cokely, Edward T., Adam Feltz, Saima Ghazal, Jinan N. Allan, Dafina Petrova,
and Rocio Garcia-Retamero. 2018. ‘Decision Making Skill: From Intelli-
gence to Numeracy and Expertise.’ In Cambridge Handbook of Expertise
and Expert Performance, edited by K. Anders Ericsson, Robert R. Hoff-
man, Aaron Kozbelt, and A. MarkWilliams, 476–505. 2nd. ed. Cambridge:
Cambridge University.
Cunha, Jr., Marcus, and Fabio Caldieraro. 2009. ‘Sunk-Cost Effects on Purely
Behavioral Investments.’ Cognitive Science 33 (1): 105–113.
Del Missier, Fabio, Timo Mäntylä, and Waendi Bruine De Bruin. 2012. ‘Deci-
sion-Making Competence, Executive Functioning, and General Cognitive
Abilities.’ Journal of Behavioral Decision Making 25 (4): 331–351.
222
Effect of Cognitive Reflection on Escalation of Commitment
Eberhardt,Wiebke,Wändi Bruine de Bruin, and JoNell Strough. 2019. ‘AgeDif-
ferences in Financial Decision Making: The Benefits of More Experience
and Less Negative Emotions.’ Journal of Behavioral Decision Making 32 (1):
79–93.
Evans, Jonathan St. B. T., and Keith E. Stanovich. 2013. ‘Dual-Process Theories
of Higher Cognition: Advancing the Debate.’ Perspectives on Psychological
Science 8 (3): 223–241.
Fennema,M. G., and Jon D. Perkins. 2007. ‘Mental Budgeting VersusMarginal
Decision Making: Training, Experience and Justification Effects on Deci-
sions Involving Sunk Costs.’Journal of Behavioral Decision Making 21 (3):
225–239.
Frederick, Shane. 2005. ‘Cognitive Reflection and DecisionMaking.’ Journal of
Economic Perspectives 19 (4): 25–42.
Fujino, Junya, Shinsuke Fujimoto, FumitoshiKodaka, Colin F. Camerer, Ryosa-
ku Kawada, Kosuke Tsurumi, Shisei Tei, Masanori Isobe, Jun Miyata, and
Genichi Sugihara. 2016. ‘Neural Mechanisms and Personality Correlates of
the Sunk Cost Effect.’ Scientific Reports 6:33171.
Garland, Howard. 1990. ‘Throwing GoodMoney After Bad: The Effect of Sunk
Costs on the Decision to Escalate Commitment to an Ongoing Project.’
Journal of Applied Psychology 75 (6): 728–731.
Haita-Falah, Corina. 2017. ‘Sunk-Cost Fallacy and Cognitive Ability in Indi-
vidual Decision-Making.’ Journal of Economic Psychology 58:44–59.
Harrison, Paul D., and Adrian Harrell. 1993. ‘Impact of “Adverse Selection” on
Managers’ Project Evaluation Decisions.’ Academy of Management Journal
36 (3): 635–643.
Heath, Chip. 1995. ‘Escalation and De-Escalation of Commitment in Response
to Sunk Costs: The Role of Budgeting in Mental Accounting.’ Organiza-
tional Behavior and Human Decision Processes 62 (1): 38–54.
Jang, Dongsuk, Anna S. Mattila, and Billy Bai. 2007. ‘Restaurant Membership
Fee andCustomer Choice: The Effects of SunkCost and Feelings of Regret.’
International Journal of Hospitality Management 26 (3): 687–967.
Karns, Tara E. 2012. ‘Age, Expertise, and the Sunk-Cost Fallacy.’ Master’s thesis,
University of West Virginia.
Klaczynski, Paul A. 2001. ‘Framing Effects on Adolescent Task Representa-
tions, Analytic and Heuristic Processing, and Decision Making.’ Journal
of Applied Developmental Psychology 22 (3): 289–309.
Larrick, Richard P., James N. Morgan, and Richard E. Nisbett. 1990. ‘Teaching
the Use of Cost-Benefit Reasoning in Everyday Life.’ Psychological Science
1 (6): 362–370.
Liang, Austin Rong-Da, Chia-Ling Lee, andWei Tung. 2014. ‘The Role of Sunk
Costs in Online Consumer Decision-Making.’ Electronic Commerce Re-
search and Applications 13 (1): 56–68.
Primi, Caterina, Kinga Morsanyi, Francesca Chiesi, Maria Anna Donati, and
Jayne Hamilton. 2016. ‘The Development and Testing of a New Version
of the Cognitive Reflection Test Applying Item Response Theory (irt).’
Journal of Behavioral Decision Making 29 (5): 453–469.
223
Petra Rekar and Mitja Perat
Putten,Marijke van,Marcel Zeelenberg, and Eric vanDijk. 2010. ‘WhoThrows
Good Money after Bad? Action vs. State Orientation Moderates the Sunk
Cost Fallacy.’ Judgment and Decision Making 5 (1): 33–36.
Ronayne, David, Daniel Sgroi, and Anthony Tuckwell. 2021. ‘Evaluating the
Sunk Cost Effect.’ Journal of Economic Behavior & Organization 186 (c):
318–327.
Roth, Stefan, Thomas Robbert, and Lennart Straus. 2015. ‘On the Sunk-Cost
Effect in Economic Decision-Making: A Meta-Analytic Review.’ Business
Research 8 (1): 99–138.
Salter, Stephen B., and David J. Sharp. 2001. ‘Agency Effects and Escalation of
Commitment: Do Small National Culture Differences Matter?’ The Inter-
national Journal of Accounting 36 (1): 33–45.
Schmidt, Jeffrey B., and Roger J. Calantone. 2002. ‘Escalation of Commitment
during New Product Development.’ Journal of the Academy of Marketing
Science 30 (2): 103–118.
Sleesman, Dustin J., Donald E. Conlon, Gerry McNamara, and Jonathan E.
Miles. 2012. ‘Cleaning up the Big Muddy: A Meta-Analytic Review of the
Determinants of Escalation of Commitment.’ Academy of Management
Journal 55 (3): 541–562.
Sobkow, Agata, Angelika Olszewska, and Jakub Traczyk. 2020. ‘Multiple Nu-
meric Competencies Predict DecisionOutcomes beyond Fluid Intelligence
and Cognitive Reflection.’ Intelligence 80:101452.
Soman, Dilip, and John T. Gourville. 2001. ‘Transaction Decoupling: How
Price Bundling Affects the Decision to Consume.’ Journal of Marketing
Research 38 (1): 30–44.
Stagnaro, Michael, Gordon Pennycook, and David G. Rand. 2018. ‘Perfor-
mance on the Cognitive Reflection Test Is Stable across Time.’ Judgment
and Decision Making 13 (3): 260–267.
Stanovich, Keith E. 2012. ‘On the Distinction between Rationality and Intelli-
gence: Implications for Understanding Individual Differences in Reason-
ing.’ In The Oxford Handbook of Thinking and Reasoning, edited by Keith J.
Holyoak and Robert G. Morrison, 433–455. New York: Oxford University
Press.
———. 2018. ‘Miserliness in Human Cognition: The Interaction of Detection,
Override and Mindware.’ Thinking & Reasoning 24 (4): 423–444.
Stanovich, Keith E., and Richard F.West. 2008. ‘On the Relative Independence
of Thinking Biases and Cognitive Ability.’ Journal of Personality and Social
Psychology 94 (4): 672–795.
Tan, Hun-Tong, and J. Frank Yates. 1995. ‘Sunk Cost Effects: The Influences
of Instruction and Future Return Estimates.’ Organizational Behavior and
Human Decision Processes 63 (3): 311–319.
———. 2002. ‘Financial Budgets and Escalation Effects.’Organizational Behav-
ior and Human Decision Processes 87 (2): 300–322.
Thaler, Richard H. 1999. ‘Mental Accounting Matters.’ Journal of Behavioral
Decision Making 12 (3): 183–206.
Toplak, Maggie E., Richard F. West, and Keith E. Stanovich. 2011. ‘The Cogni-
224
Effect of Cognitive Reflection on Escalation of Commitment
tive ReflectionTest as a Predictor of Performance onHeuristics-and-Biases
Tasks.’Memory & Cognition 39 (7): 1275–1289.
———. 2014. ‘Assessing Miserly Information Processing: An Expansion of the
Cognitive Reflection Test.’ Thinking & Reasoning 20 (2): 147–168.
Von Neumann, John, and Oskar Morgenstern. 1947. Theory of Games and Eco-
nomic Behavior. 2nd rev. ed. Princeton, nj: Princeton University Press.
Wong, Kin Fai Ellick, Jessica Yuk Yee Kwong, and Carmen K. Ng. 2008. ‘When
Thinking Rationally Increases Biases: The Role of Rational Thinking Style
in Escalation of Commitment.’ Applied Psychology 57 (2): 246–171.
Appendix
Assume you are a director of a company ‘Innovation Ltd.’ Your company
embarked on a newr&dproject – avrheadset which would enable an im-
proved experience of several popular computer games. However, at this
point in time one of your competitors announced a launch of their ownvr
headset, which has better characteristics than yours and is cheaper.
The project is two years underway and your team informed you theywould
need another three years to complete it. Due to aging of technology, your
team believes the bulk of the sales revenues would occur in five years after
introduction, should you decide to continue with the project. Your mar-
keting and development team provided you with the following data:
Costs incurred to date: 7 million eur.
Costs/sales Years
1 2 3 4 5
Expected future costs 3 2 2
Expected sales 1 3 3 2 1
Note. In million eur.
Cognitive reflection test (crt):
1. A bat and a ball cost $1.10 in total. The bat costs $1.00 more than the
ball. How much does the ball cost? cents
2. If it takes 5 machines 5 minutes to make 5 widgets, how long would it
take 100 machines to make 100 widgets? minutes
3. In a lake, there is a patch of lily pads. Every day, the patch doubles in
size. If it takes 48 days for the patch to cover the entire lake, how long
would it take for the patch to cover half of the lake? days
225