The effects of anticipated regret on risk preferences of social and problem gamblers

Text-only Preview

Judgment and Decision Making, Vol. 4, No. 3, April 2009, pp. 227–234
The effects of anticipated regret on risk preferences of social and
problem gamblers
Karin Tochkov?
Texas A&M University-Commerce
Anticipated regret is an important determinant in risky decision making, however only a few studies have explored
its role in problem gambling. This study tested for differences in the anticipation of regret among social and problem
gamblers and examined how these differences affect risk preferences in a gambling task. The extent of problem gambling
was assessed using the South Oaks Gambling Screen and participants were randomly assigned to one of two conditions.
In the risky feedback condition, the feeling of regret was avoided by choosing the risky gamble, whereas in the safe
feedback condition the safe gamble was the regret-minimizing option. Problem gambling was associated with the choice
of the risky gamble in both conditions indicating less sensitivity to anticipated regret. It was also associated with risk
seeking across feedback conditions when the stakes of winning and loosing were higher. These ?ndings suggest that
less regret or the poor anticipation of regret might contribute to excessive gambling and thus need to be addressed in
cognitive treatments of problem gambling.
Keywords: anticipated regret; risk taking; problem gambling.
1 Introduction
1999; Zeelenberg et al., 2000). However, relatively few
studies have examined anticipated regret in the context
Cognitive distortions have been identi?ed in the research
of gambling. Wolfson and Briggs (2002) investigated the
literature as a key contributing factor to the instigation
intentions of 485 lottery players to participate in a newly
and maintenance of problem gambling (Ladouceur et al.,
introduced lottery in the United Kingdom and found that
2002; Sharpe, 2002). Numerous studies have shown that
38% were willing to purchase tickets because they antici-
gamblers hold irrational beliefs and suffer from cogni-
pated feeling regret if their numbers came up. This share
tive biases such as the illusion of control, the gambler’s
increased to 54% for those who played regularly with the
fallacy, and the attribution bias, which lead to inaccu-
same set of numbers.
rate inferences about outcome probabilities (Delfabbro &
In a comprehensive study of the Dutch postcode lot-
Wine?eld, 2000; Gaboury & Ladouceur, 1989; Sundali &
tery, Zeelenberg and Pieters (2004) showed that the re-
Croson, 2006; Toneatto et al., 1997). The resulting over-
gret people expected to feel if they decided not to play
estimation of the chances to win and the underestimation
and discovered that their neighbor had won signi?cantly
of possible losses propel individuals to place more risky
contributed to their intention to participate in the lottery
bets and encourage persistent gambling.
in the near future. Rae and Haw (2005) assessed the ef-
Besides erroneous beliefs about probabilities, gam-
fects of anticipated regret, disappointment, and elation,
bling persistence and risk taking may also be affected
on the persistence in gambling of 93 gamblers and found
by the anticipation of emotions associated with gambling
that anticipated emotions did not predict gambling persis-
outcomes. The role of anticipated emotions is well docu-
tence. Although all three studies examine the relationship
mented in the literature on decision making, with antici-
between anticipated regret and gambling, their results do
pated regret receiving particular attention (Mellers et al.,
not allow to draw conclusions about the clinical implica-
tions of regret for excessive gambling. This is due to the
?The author would like to thank Jonathan Baron, two anonymous
fact that none of the studies screened the participants for
reviewers, and conference participants at the 2008 ABCT Annual Con-
gambling problems making it impossible to assess how
vention and the 2009 meeting of the British Psychological Society for
helpful comments and suggestions on an earlier draft of the paper. Tina
many of them actually met the criteria for problem or
Clark provided excellent research assistance. Financial support from the
pathological gamblers.
Faculty Research Enhancement Grant by the Graduate School at Texas
The present study examined the role of anticipated re-
A&M University-Commerce is gratefully acknowledged. Address: De-
partment of Psychology and Special Education, P.O. Box 3011, Com-
gret in problem gambling by focusing on two issues.
merce, TX 75429. Email: [email protected]
First, it estimated the effect of gambling preference on

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
the anticipation of regret. Previous research has sug-
els of regret anticipation. Following Zeelenberg et al.
gested that although erroneous beliefs and distorted cog-
(1996), two experimental conditions were created, each
nitions are common among all types of gamblers, patho-
of which contained a partial and a complete feedback op-
logical gamblers seem to suffer more cognitive illusions
tion. In the risky feedback condition, partial feedback
(Baboushkin et al., 2001), endorse more irrational beliefs
was associated with the risky gamble, whereas in the safe
(Joukhador et al., 2003) and to be more convinced in
feedback condition the regret-minimizing option was the
their irrational beliefs than social gamblers (Ladouceur,
safe gamble. I hypothesized that a weaker gambling pref-
2004). In a recent study, Tochkov (2009) showed that
erence would trigger regret avoidance and thus a prefer-
high-frequency gamblers were less able to anticipate re-
ence for the partial feedback which would lead to risk
gret than low-frequency gamblers indicating that inaccu-
seeking or risk aversion depending on the feedback con-
rately anticipated regret is a possible contributing factor
dition. In contrast, a stronger gambling preference would
to persistent gambling. Participants were asked to choose
be related to risk seeking regardless of the expected feed-
gambles with ?ctitious monetary outcomes and imagine
back. Moreover, I examined risk preferences for gambles
how they would feel once they learn the outcome of their
with low and high variation in outcome probabilities to
decision. A week later, the same participants were asked
test for the robustness of the results with respect to the
to play the same gambles for real with actual monetary
different stakes involved in winning and losing.
wins and losses and rate their regret. Tochkov (2009) re-
ported that the difference between anticipated and actual
regret was signi?cantly larger for high-frequency gam-
2 Method
blers as compared to that of low-frequency gamblers.
These ?ndings suggest that when gamblers do not an-
2.1 Participants
ticipate the negative feelings of regret they might expe-
Participants (n=108) were undergraduate psychology stu-
rience once they learn the outcome of their bet, they are
dents who responded to a recruiting message seeking in-
more tempted to continue gambling. In contrast, the an-
dividuals who like to gamble. Sixty-two of the partici-
ticipation of regret about gambling outcomes could serve
pants (57%) were female, and 46 (43%) were male. The
as a natural inhibitor to continuous gambling. This line
majority (85%) were between the ages 18–25, 10% were
of reasoning is supported by evidence from a number of
between 26–30, and 5% were older than 30.
studies which found that the anticipation of regret de-
creases the intentions of engaging in risky and potentially
addictive behaviors (Richard et al., 1996; Van Empelen et
2.2 Measures
al., 2001). In a recent study, Fernandez-Duque and Lan-
The South Oaks Gambling Screen (SOGS; Lesieur &
ders (2008) showed that individuals experienced more re-
Blume, 1987) was used to assess the extent of gambling
gret than they had anticipated, which in turn made them
problems and preferences. The SOGS is a 20-item ques-
more risk averse in a subsequent gambling task.
tionnaire based on the DSM-III criteria for pathological
The present study explores the relationship between
gambling and is one of the most widely used measures
anticipation of regret and gambling preferences by giv-
in studies on gambling for both clinical and non-clinical
ing gamblers the choice between two gambles, one of
populations. It has well-established psychometric prop-
which promised to reveal the outcome of the selected op-
erties, including high internal reliability and a 1-month
tion only (partial feedback) and the other the outcome of
test-retest reliability. SOGS scores range from 0 to 20,
both the selected and the rejected options (complete feed-
with a score below 3 indicating no gambling problems, a
back). It was hypothesized that a stronger gambling pref-
score of 3 or 4 indicating a problem gambler, and a score
erence would be adversely associated with regret antici-
of 5 or higher suggesting a probable pathological gam-
pation and would thus result in the more frequent choice
bler. For the purposes of the current study, participants
of the gamble with partial feedback.
with a SOGS score of 1 or 2 were classi?ed as social
The second goal of the study was to assess the effect of
gamblers, and those with scores of 3 and higher as prob-
gambling preferences on risk attitudes. Previous studies
lem gamblers. Higher levels of gambling preference were
have demonstrated that problem gamblers exhibit higher
expected to be negatively associated with regret anticipa-
degrees of risk taking than normal subjects because they
tion and to contribute to more risk seeking.
suffer from cognitive biases that make risky bets more
The Gambling Attitudes and Beliefs Survey (GABS;
attractive (Gaboury & Ladouceur, 1989; Toneatto, 1999)
Breen & Zuckerman, 1999) was used to evaluate cog-
or because they feel overcon?dent in their skills (Goodie,
nitive biases, irrational beliefs, and positive values re-
2005). Accordingly, it was hypothesized that the risk
garding gambling. It consists of 35 items including state-
attitude associated with a stronger gambling preference
ments such as “If I have been lucky lately, I should press
would be skewed towards risk seeking due to lower lev-
my bets” or “People who make big bets can be very

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
sexy.” Responses are recorded on a 4-point Likert-type
were also shown beneath each gamble, along with a mes-
scale. High scores on GABS suggest that participants
sage informing participants that they would learn only the
view gambling as exciting and socially meaningful, and
outcome of their chosen gamble or the outcome of both
that luck and strategies (even illusory ones) are important,
gambles, depending on whether the gamble was associ-
which in turn was hypothesized to impede the feelings of
ated with partial or complete feedback, respectively.
regret and regret anticipation and encourage risk taking.
The Competitiveness Index — Revised (CI-R; Hous-
ton et al., 2002) is a 14-item measure designed to assess
the desire to win in interpersonal situations. Respondents
2.4 Feedback conditions
rate the extent to which they agree with the items using a
5-point Likert scale. Competitiveness as measured by the
Participants were randomly assigned to one of two exper-
CI-R has been shown to be an important risk factor for
imental conditions designed by Zeelenberg et al. (1996).
pathological gambling (Parke et al., 2004), and was thus
In the risky feedback condition participants learned the
hypothesized to encourage risk seeking.
outcome of the risky gamble regardless of whether they
The Eysenck Impulsiveness Questionnaire (EIQ;
preferred the safe or the risky gamble. In contrast, the
Eysenck et al., 1985) is a 63-item measure with 3 sub-
outcome of the safe gamble was revealed only if it was
scales targeting different aspects of impulsivity. Only the
the preferred one. The partial feedback on the risky gam-
?rst two subscales focusing on impulsiveness and ven-
ble thus ensured that regret was not possible, whereas the
turesomeness were used as it was assumed that these
complete feedback associated with the choice of the safe
traits affect the risk preference of participants. The re-
gamble guaranteed that regret was felt given the possibil-
sponses of the two subscales were combined into one
ity of comparing the outcomes of the two gambles. In
measure with each given equal weight. Impulsivity is a
the safe feedback condition, it was the outcome of the
common trait of problem gamblers and was expected to
safe gamble that was disclosed regardless of the choice
result in more risk taking as it impedes the assessment of
of gamble. On the other hand, participants learned the
anticipated emotions and probabilities.
outcomes of both gambles only if they preferred the risky
gamble. In this condition, only the choice of the risky
2.3 Computerized gambling task
gamble could result in regret as it was associated with
complete feedback.
The task involved 20 pairs of gambles displayed on a
computer screen in a sequence randomized separately for
each subject. Each gamble had two possible payoffs with
different probabilities. Probabilities were either .3 and .7
2.5 Procedure
(low variance) or .1 and .9 (high variance). Within each
pair of gambles, only gambles with the same variance
Participants were told that the experiment involved
were included in order to examine whether variation in
choices between pairs of gambles with real monetary
the probability of a payoff across gambling pairs affected
wins and losses and that their payment would be the
risk preference. Each pair of gambles included one rel-
sum of their earnings over all trials. After providing in-
atively safe (small payoff with high probability) and one
formed consent, participants were randomly assigned ei-
risky gamble (high payoff with low probability).1 Pos-
ther to the risky or the safe feedback condition and were
sible monetary outcomes involved wins and losses rang-
asked to complete the questionnaire package including
ing between $0 and $15 and were designed so that when
SOGS, GABS, CI-R, and EIQ. Next, they were presented
gambles were paired they would have the same expected
with the computerized gambling task. The experimenter
value. Each gamble displayed on the screen was repre-
demonstrated how to play the gambles on the computer
sented by a pie chart, which visualized the different prob-
and informed them about the partial and complete feed-
abilities. Payoffs and their corresponding probabilities
back associated with each gamble. Participants were told
1The pairs of gambles included in the task were:
that in each trial they should chose one gamble and then
([.3/$10;.7/$0], [.3/$0;.7/$4]),
([.3/$2;.7/$6], [.3/$15;.7/$2]),
rate the strength of their preference for that gamble on a
([.3/$4;.7/$1], [.3/$10;.7/$1]),
([.3/$6;.7/$0], [.3/$15;.7/$0]),
scale of 1 (very weak preference) to 12 (very strong pref-
([.3/$1;.7/$5], [.3/$10;.7/$1]),
([.3/$10;.7/$2], [.3/$2;.7/$5]),
([.3/$2;.7/$8], [.3/$15;.7/$2]),
([.3/$10;.7/$2], [.3/$2;.7/$3]),
erence). The experiment lasted approximately 45 min-
([.3/$15;.7/$1], [.3/$1;.7/$7]),
([.3/$15;.7/$1], [.3/$1;.7/$6]),
utes and everyone was paid $10 regardless of their actual
([.1/$2;.9/$0], [.1/$15;.9/$2]),
([.1/$1;.9/$0.20], [.1/$10;.9/$1]),
winnings. Lastly, participants were debriefed about the
([.1/$10;.9/$2], [.1/$2;.9/$3]),
([.1/$0;.9/$1.70], [.1/$15;.9/$0]),
purpose of the experiment and the actual likelihood of
([.1/$10;.9/$0], [.1/$0;.9/$1]),
([.1/$2;.9/$3.40], [.1/$15;.9/$2]),
([.1/$15;.9/$1], [.1/$1;.9/$0.80]),
([.1/$10;.9/$2], [.1/$2;.9/$0]),
winning, and were provided with information on where
([.1/$15;.9/$1], [.1/$2.50;.9/$1]),
([.1/$1;.9/$2], [.1/$10;.9/$1]).
to seek help with gambling problems if needed.

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
Table 1: Descriptive statistics of the entire sample and by
Table 2: Results of the multiple regression analysis for
gambler type.
variables predicting risk preference.
Safe Feedback
Gambling preference
Risky Feedback 52
Feedback condition
Lagged risk
* p<0.05 ** p<0.01 *** p<0.001
2.6 Data analysis overview
condition and the risky gamble in the risky feedback con-
The data were analyzed using linear regression. The de-
dition. In the risky feedback condition, the risky gamble
pendent variable was the strength of risk preference and
was the preferred option by both types of gamblers with
was modeled after Zeelenberg et al. (1996) by converting
the preference scores for the selected gambles to a scale
about two thirds of all gambles. Although this share was
evenly spaced around zero. For this purpose the pref-
higher for problem gamblers as compared to social gam-
erence scores were multiplied by 1 for the risky gamble
blers (66.3 vs. 64.7), the difference was not statistically
and ?1 for the safe gamble and then 1/2 was subtracted
signi?cant. In the safe feedback condition, the safe gam-
to create a scale that is evenly spaced around zero. The
ble was chosen more often than the risky gamble for both
risk preference variable thus ranged from ?11.5 (strong
categories of gamblers. Again, the share of risky gambles
risk aversion) to 11.5 (extreme risk seeking). The inde-
selected by problem gamblers was higher than for social
pendent variables included the strength of gambling pref-
gamblers (47.2 vs. 41.6), however this time the differ-
erence as measured by the SOGS score and feedback con-
ence was statistically signi?cant (p<.05). Furthermore,
dition (safe vs. risky feedback). The interaction between
in the safe feedback condition the difference between the
strength of gambling preference and feedback condition
shares of safe and risky gambles for social gamblers was
was included in the regression equation along with addi-
signi?cantly larger (p<.001) than for problem gamblers,
tional control variables such as competitiveness, impul-
indicating that problem gamblers are less sensitive to the
sivity, gender, and the level of gambling biases and dis-
effects of anticipated regret.
tortions concerning gambling (as measured by GABS).
The results of the regression analysis examining the ef-
Lastly, a lag variable was introduced to estimate the ef-
fect of gambling preference on risk attitudes are shown
fect of the risk preference indicated on the previous trial.
in the ?rst column of Table 2. The constant represented
the average risk preference in the safe feedback condition
and was statistically signi?cant. The negative sign of the
3 Results
coef?cient indicated that participants exhibited risk aver-
sion in the safe feedback condition. The feedback vari-
Descriptive statistics of the sample are displayed in Table
able which had also a signi?cant coef?cient denoted the
1. Out of a total sample of 108 participants, 66% were
difference in the levels of risk preference between the two
classi?ed as social gamblers (1?SOGS?2) and 34% as
feedback conditions. The sum of the feedback coef?cient
problem gamblers (SOGS?3). To test for differences in
and the constant indicated the presence of risk seeking in
the anticipation of regret, I examined the choices made
the risky feedback condition.
by social and problem gamblers in each of the two feed-
The gambling preference variable measured the
back conditions. To avoid regret, participants were ex-
marginal effect of SOGS scores on risk preference. In the
pected to choose the safe gamble in the safe feedback
safe feedback condition an increase in the SOGS score by

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
one led to an increase in risk seeking by 0.56. The inter-
All Gambles
action between gambling preference and feedback con-
dition represented the additional marginal effect in the
risky feedback condition. Accordingly, a one point in-
crease in the SOGS score resulted in a 0.26 (0.56–0.82)
points decrease in risk seeking (or alternatively increase
in risk aversion). Higher SOGS scores were correlated
with risk seeking in the safe feedback and risk aversion in
the risky feedback. This is the opposite of what should be
expected if regret minimization was a primary concern.
These ?ndings suggest that individuals with a stronger
gambling preference are less susceptible to the effects
Social gamblers
of anticipated regret. Among the control variables, only
Problem gamblers
gender and impulsivity showed a signi?cant and positive
effect suggesting that higher levels of impulsivity and be-
ing male contribute to more risk seeking.
Safe feedback
Risky feedback
The estimation was also performed separately for gam-
Low?variance Gambles
ble pairs with a relatively low (.3 vs. .7) and a relatively
high (.1 vs. .9) variation in outcome probabilities to test
whether an increase in the riskiness of gambles affected
regret minimization. The estimates are shown in the sec-
ond and third column of Table 2, respectively. In the low-
variance model the statistical signi?cance and the signs
of the coef?cients were largely identical to the estimates
of the model that included all gambles. The only excep-
Social gamblers
tion was that the magnitude of the decrease in risk pref-
Problem gamblers
erence in the risky feedback condition was much smaller
as compared to the overall model. In the high-variance
model participants did not exhibit a signi?cant risk aver-
sion in the safe feedback condition. Moreover, the effect
of gambling preference on risk attitudes did not differ sig-
Safe feedback
Risky feedback
ni?cantly across the two feedback conditions, suggesting
that higher SOGS scores were associated with increases
High?variance Gambles
in risk seeking regardless of the possibility to experience
Figure 1 illustrates the average risk preference levels
for each of the four experimental groups. In the model
Social gamblers
including all gambles, risk preferences for the two groups
Problem gamblers
of gamblers had the same direction within each feedback
condition: risk aversion in the safe feedback condition
and risk seeking in the risky feedback condition. How-
ever, the gap between the mean levels of risk aversion
exhibited by social (M=?1.63) and problem gamblers
(M=?0.94) in the safe feedback condition was statisti-
cally signi?cant. Similarly, the magnitude of risk seeking
in the risky feedback condition differed signi?cantly be-
tween social (M=0.80) and problem (M=0.49) gamblers.
Safe feedback
Risky feedback
In the high and low variance models, social gam-
blers continued with their pattern of risk aversion in the
safe feedback and risk seeking in the risky feedback
Figure 1: Mean levels of risk preference for social and
condition, however the magnitude of risk aversion was
problem gamblers in the two feedback conditions.
much stronger in the low variance model (M=?2.02 vs.
M=?0.48) and the risk seeking was much more intense

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
when there was a large variation in outcome probabilities
seeking in the safe feedback condition and risk aversion
(M=2.56 vs. M=0.10). By comparison, problem gam-
in the risky feedback condition. This indicates that the
blers exhibited risk aversion in the low variance model
failure to identify the regret-minimizing option has an
and risk seeking in the high variance model regardless of
adverse effect on risk taking behavior. Moreover, when
the feedback condition.
the difference in probabilities of gambling outcomes in-
creased, magnifying the chances of experiencing losses
and thus feeling regret, stronger gambling preference
4 Discussion
tended to result in risk seeking regardless of the feedback
Previous research has suggested that anticipated regret is
One of the limitations of the study is that it used
important in decision making. This study investigated the
students rather than community gamblers which weak-
role of anticipated regret in problem gambling, arguing
ens the applicability of the results to the population of
that it might be a contributing factor to excess gambling
pathological gamblers. A second limitation was that the
and risk taking. It found a negative correlation between
study was not able to determine whether problem gam-
the strength of gambling preferences and regret anticipa-
blers were less able to anticipate regret than social gam-
tion. Weaker gambling preference was associated with
blers or whether they simply experienced less regret. Al-
the choice of the safe gamble in the safe feedback con-
though previous studies have shown that it is the inac-
dition and the risky gamble in risky feedback condition,
curate anticipation of regret (Tochkov, 2009), more re-
suggesting that those with less gambling problems antic-
search is needed to examine this issue in more detail.
ipated regret as they selected the regret-minimizing alter-
Despite these limitations, the ?ndings of the study could
native. Such behavior is consistent with the ?ndings of
have important implications for the clinical practice as it
Zeelenberg et al. (1996) and Ritov (1996).
is likely that weaker anticipation of regret affects gam-
In contrast, stronger gambling preference led to risky
bling behavior in combination with other cognitive dis-
gambles being chosen more often regardless of the feed-
tortions. Cognitive treatments for pathological gamblers
back condition. Furthermore, despite the higher share of
have focused on challenging the irrational beliefs and
safe gambles in the safe feedback condition, the differ-
faulty cognitions involved in gambling fallacies and pro-
ence between the two types of gambles was not statis-
viding educational advice on the nature of randomness
tically signi?cant for those exhibiting stronger gambling
and probabilities (Ferland et al., 2002; Ladouceur et al.
preference. These ?ndings suggest that complete feed-
2001, 2003). A better anticipation of negative emotions
back and the resulting threat of experiencing regret did
could be promoted in a similar framework by identifying,
not lead to the adoption of a regret-minimizing strategy
challenging, and correcting faulty perceptions about neg-
across the two experimental conditions for individuals
ative emotions associated with the outcomes of gambles.
with more gambling problems, and did not deter them
In particular, the salience of regret needs to be increased
from seeking out the risky gamble more often than indi-
whereby the therapist would for instance depict the pos-
viduals with a weaker gambling preference.
sible consequences of excessive gambling by describing
Given that stronger gambling preference was associ-
the ?nancial and social issues involved. The goal of this
ated with less responsiveness to possible regret, it seems
technique would be to teach patients to focus on a more
that weak anticipation of negative emotions might fuel
realistic assessment of the consequences of their behav-
excessive gambling. Although future research needs to
ior and the concomitant negative feelings including regret
shed more light on this issue, it is possible that the expec-
before making the decision to continue gambling.
tation of experiencing a negative emotion such as regret
This approach needs to be studied in more detail in
over losing the next round of betting can serve as a nat-
the future given the ambiguities in the literature. While
ural inhibitor to prolonged gambling. Hills et al. (2001)
some studies have shown that a better grasp of probabili-
showed for instance that current negative mood did in fact
ties and common fallacies reduces risky behavior in gam-
result in shorter gambling sessions for non-pathological
bling (Floyd et al., 2006), others have found that it does
gamblers. In contrast, the failure to take into account the
not translate into decreases in actual gambling behavior
possibility of dealing with an unpleasant emotion after
(Williams & Connolly, 2006). Similarly, increasing the
the outcome of the bet has been revealed could amplify
salience of the negative emotions associated with risky
the lack of self control and result in excessive gambling.
behavior was reported to be effective in reducing risk tak-
The second hypothesis of this study, which addressed
ing in sexual behavior (Richard et al., 1996), whereas the
the link between the strength of gambling preference and
same technique improved the anticipation of regret over
risk taking was also supported by the ?ndings. Higher
heavy drinking but did not result in less drinking or less
levels of gambling preference were associated with risk
risk taking (Murgraff et al., 1999).

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
of pathological gambling. Journal of Nervous & Men-
tal Disease, 189, 773–80.
Baboushkin, H., Hardoon, K., Derevensky, J., & Gupta,
Ladouceur R, Sylvain C, Boutin C, & Doucet C. (2002).
R. (2001). Underlying cognitions in gambling behav-
Understanding and treating pathological gamblers.
ior amongst university students. Journal of Applied So-
London: Wiley.
cial Psychology, 31, 1–23.
Ladouceur, R., Sylvain, C., Boutin S., Lachance, C.,
Breen, R. & Zuckerman, M. (1999). Chasing in gambling
Doucet, C., & Leblond, J. (2003). Group therapy for
behavior: personality and cognitive determinants. Per-
pathological gamblers: a cognitive approach. Behavior
sonality and Individual Differences, 27, 1097–111.
Research and Therapy, 41, 587–96.
Delfabbro, P. H., & Wine?eld, A. H. (2000). Predictors
Lesieur, H. R., & Blume, S. B. (1987).
The South
of irrational thinking in regular slot machine gamblers.
Oaks Gambling Screen (SOGS): A new instrument for
Journal of Psychology, 134, 117–28.
the identi?cation of pathological gamblers. American
Eysenck, S., Pearson, P., Easting, G., & Allsopp, J.
Journal of Psychiatry, 144, 1184–1188.
(1985). Age norms for impulsiveness, venturesome-
Mellers, B. Ritov, I., & Schwartz, A. (1999). Emotion-
ness and empathy in adults. Personality and Individual
based choice. Journal of Experimental Psychology:
Differences, 6, 613–619.
General, 128, 332–345.
Ferland, F., Francine, Ladouceur, R. & Vitaro, F. (2002).
Murgraff, V., McDermott, M. R., White, D., & Phillips,
Prevention of problem gambling: Modifying miscon-
K. (1999). Regret is what you get? The effects of
ceptions and increasing knowledge. Journal of Gam-
manipulating anticipated affect and time perspective
bling Studies, 18, 19–30.
on riskier single-occasion drinking. Alcohol & Alco-
Fernandez-Duque, D. & Landers, J. (2008). “Feeling
holism, 34, 590–600.
more regret than I would have imagined”: Self-report
Parke, A., Grif?ths, M., & Irwing, P. (2004). Personal-
and behavioral evidence. Judgment and Decision Mak-
ity traits in pathological gambling: Sensation-seeking,
ing, 3, 449–456.
deferment of grati?cation and competitiveness as risk
Floyd, K., Whelan, J. P., & Meyers, A. W. (2006). Use
factors. Addiction Research and Theory, 12, 201–212.
of warning messages to modify gambling beliefs and
Parker, P., Manstead, A. S. R., & Stradling, S. G. (1995)
behavior in a laboratory investigation. Psychology of
Extending the theory of planned behaviour: the role of
Addictive Behaviors, 20, 69–74.
personal norm. British Journal of Social Psychology,
Gaboury, A. & Ladouceur, R. (1989). Erroneous percep-
34, 127–137.
tions and gambling. Journal of Social Behavior and
Rae, D. & Haw, J. (2005). The effect of depression and
Personality, 4, 411–20.
perceived skill on anticipated emotions and persistence
Goodie, A. S. (2005). The role of perceived control and
in off-course betting. International Gambling Studies,
overcon?dence in pathological gambling. Journal of
5, 199–208.
Gambling Studies, 21, 481–502.
Richard, R., van der Pligt, J. and de Varies, N. (1996) An-
Hills, A. M., Hill, S., Mamone, N., & Dickerson, M.
ticipated regret and time perspective: changing sexual
(2001). Induced mood and persistence at gaming. Ad-
risk taking behavior. Journal of Behavioural Decision
diction, 96, 1629–1628.
Making, 9, 185–199.
Houston, J. M., Harris, P., McIntire, S., & Francis, D.
Ritov, I. (1996). Probability of regret: Anticipation of
(2002). Revising the Competitiveness Index using fac-
uncertainty resolution in choice. Organizational Be-
tor analysis, Psychological Report, 90, 31–34.
havior and Human Decision Processes, 66, 228–236.
Josephs, R. A., Larrick, R. P., Steele, C. M. & Nisbett, R.
Sharpe, L. (2002). A reformulated cognitive-behavioral
E. (1992) Protecting the self from the negative conse-
model of problem gambling: A biopsychosocial per-
quences of risky decisions. Journal of Personality and
spective. Clinical Psychology Review, 22, 1–25.
Social Psychology, 62, 26–37.
Sheeran, P, & Orbell, S. (1999). Augmenting the theory
Joukhador, J., Maccallum, F., & Blaszczynski, A. (2003).
of planned behavior: Roles for anticipated regret and
Differences in cognitive distortions between problem
descriptive norms. Journal of Applied Social Psychol-
and social gamblers. Psychological Reports, 92, 1203–
ogy. 29, 2107–2142.
Sundali, J., & Croson, R. (2006). Biases in casino bet-
Ladouceur, R. (2004). Perceptions among pathological
ting: The hot hand and the gambler’s fallacy. Judgment
and nonpathological gamblers. Addictive Behaviors,
and Decision Making, 1, 1-12.
29, 555–565.
Tochkov, K. (2009). The impact of emotions on prob-
Ladouceur, R., Sylvain, C., Boutin, C., Lachance, S.,
lem gambling. American Journal of Psychological Re-
Doucet, C., & Leblond, J. (2001). Cognitive treatment
search, 5, 7–19.

Judgment and Decision Making, Vol. 4, No. 3, April 2009
Anticipated regret in gamblers
Toneatto, T. (1999). Cognitive psychopathology of prob-
Zeelenberg, M., Beattie, J., van der Pligt, J., & de Vries,
lem gambling. Substance Use and Misuse, 34, 1593–
N. K. (1996). Consequences of regret aversion: Effects
of expected feedback on risky decision making. Orga-
Toneatto, T., Blitz-Miller, T., Calderwook, K., Drag-
nizational Behavior and Human Decision Processes,
onetti, R., Tsanos, A. (1997) Cognitive distortions in
65, 148–158.
heavy gambling. Journal of Gambling Studies, 13,
Zeelenberg, M. & Pieters, R. (2004). Consequences of
regret aversion: The case of the Dutch postcode lot-
Van Empelen, P., Kok, G. J., Jansen, M. W. J. & Hoebe C.
tery. Organizational Behavior and Human Decision
J. P. A. (2001). The additional value of anticipated re-
Processes, 93, 155–168.
gret and psychopathology in explaining intended con-
Zeelenberg, M., van Dijk, W. W., Manstead, A. S. R.,
dom use among drug users. AIDS Care, 13, 309–318.
& van der Pligt, J. (2000). On bad decisions and dis-
Williams, R. J. & Connolly, D. (2006). Does learning
con?rmed expectancies: The psychology of regret and
about the mathematics of gambling change gambling
disappointment. Cognition and Emotion, 14, 521–541.
behavior? Psychology of Addictive Behaviors, 20, 62–
Wolfson, S. & Briggs, P. (2002). Locked Into gambling:
Anticipatory regret as a motivator for playing the na-
tional lottery. Journal of Gambling Studies, 18, 1–17.