You are on page 1of 6

Current Directions in Psychological

Science http://cdp.sagepub.com/

Regret in Decision Making


Terry Connolly and Marcel Zeelenberg
Current Directions in Psychological Science 2002 11: 212
DOI: 10.1111/1467-8721.00203

The online version of this article can be found at:


http://cdp.sagepub.com/content/11/6/212

Published by:

http://www.sagepublications.com

On behalf of:

Association for Psychological Science

Additional services and information for Current Directions in Psychological Science can be found at:

Email Alerts: http://cdp.sagepub.com/cgi/alerts

Subscriptions: http://cdp.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Dec 1, 2002

What is This?

Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014


212 VOLUME 11, NUMBER 6, DECEMBER 2002

Collins, A.M., & Loftus, E.F. (1975). A spreading- analytic integration. Basic and Applied Social
Paulus, P.B., & Nijstad, B.A. (Eds.). activation theory of semantic processing. Psy- Psychology, 12, 3–23.
(in press). Group creativity. New chological Review, 82, 407–428. Paulus, P.B. (2000). Groups, teams and creativity:
York: Oxford University Press. Dennis, A.R., & Williams, M.L. (in press). Elec- The creative potential of idea generating groups.
tronic brainstorming: Theory, research, and Applied Psychology: An International Review, 49,
future directions. In P.B. Paulus & B.A. Nijstad 237–262.
(Eds.), Group creativity. New York: Oxford Paulus, P.B., & Brown, V. (in press). Ideational creativity
University Press.
Note in groups: Lessons from research on brainstorming.
Diehl, M., & Stroebe, W. (1991). Productivity loss In P.B. Paulus & B.A. Nijstad (Eds.), Group creativity.
in idea-generating groups: Tracking down the New York: Oxford University Press.
1. Address correspondence to Paul B. blocking effect. Journal of Personality and Social Paulus, P.B., Dugosh, K.L., Dzindolet, M.T.,
Paulus, Department of Psychology, Uni- Psychology, 61, 392–403. Coskun, H., & Putman, V.L. (2002). Social and
versity of Texas at Arlington, Arlington, Dugosh, K.L., Paulus, P.B., Roland, E.J., & Yang, cognitive influences in group brainstorming:
TX 76019; e-mail: paulus@uta. edu. H.-C. (2000). Cognitive stimulation in brain- Predicting production gains and losses. Euro-
storming. Journal of Personality and Social Psy- pean Review of Social Psychology, 12, 299–325.
chology, 79, 722–735. Paulus, P.B., Larey, T.S., & Ortega, A.H. (1995).
Leggett, K.L. (1997). The effectiveness of categorical Performance and perceptions of brainstormers
References priming in brainstorming. Unpublished master’s in an organizational setting. Basic and Applied
thesis, University of Texas at Arlington. Social Psychology, 18, 3–14.
Bennis, W., & Biederman, P.W. (1997). Organizing Leggett, K.L., Putman, V.L., Roland, E.J., & Paulus, Paulus, P.B., & Yang, H.-C. (2000). Idea generation
genius: The secrets of creative collaboration. Read- P.B. (1996, April). The effects of training on per- in groups: A basis for creativity in organiza-
ing, MA: Addison Wesley. formance in group brainstorming. Paper presented tions. Organizational Behavior and Human Deci-
Brown, V., Tumeo, M., Larey, T.S., & Paulus, P.B. at the annual meeting of the Southwestern sion Processes, 82, 76–87.
(1998). Modeling cognitive interactions during Psychological Association, Houston, TX. Sutton, R.I., & Hargadon, A. (1996). Brainstorming
group brainstorming. Small Group Research, 29, Mullen, B., Johnson, C., & Salas, E. (1991). Produc- groups in context. Administrative Science Quar-
495–526. tivity loss in brainstorming groups: A meta- terly, 41, 685–718.

Regret in Decision Making an intensely emotional experience.


Of course, most people give such
Terry Connolly1 and Marcel Zeelenberg decisions careful thought: What are
the options? What consequences
Department of Management and Policy, University of Arizona, Tucson, Arizona
might each lead to? How likely are
(T.C.), and Department of Economic and Social Psychology, Tilburg University,
Tilburg, the Netherlands (M.Z.) they? How desirable are they? But,
in addition to these important cog-
nitive considerations, there are
strong emotional factors. At the time
Abstract blame for having made a poor of the decision, the person has feel-
Decision research has only choice. We reinterpret several ings about the decision itself (fear
recently started to take seri- existing studies in DJT terms. of surgery, anxiety about a career
ously the role of emotions in We then report some new choice) and expectations about
choices and decisions. Regret is studies that directly tested feelings he or she may experience
the emotion that has received (and support) DJT, and pro- later (relief at a good outcome, sad-
the most attention. In this arti- pose a number of research is- ness at a poor one). After the fact,
cle, we sample a number of the sues that follow from this new the individual experiences emotions,
initial regret studies from psy- approach to regret. which may or may not track with
chology and economics, and the earlier expectations.
trace some of the complexities Keywords The emotion that has received
and contradictions to which regret; decision making; emo- the most research attention from
they led. We then sketch a new tion; decision justification theory decision theorists is regret. Most
theory, decision justification people can readily recall or imag-
theory (DJT), which synthe- ine situations in which a poor deci-
sizes several apparently con- sion led to painful regret. In the
flicting findings. DJT postulates Decision researchers have only early 1980s, researchers in both lab-
two core components of deci- recently started to take seriously an oratory and field settings started to
sion-related regret, one associ- aspect of making choices that every produce elegant theories and vivid
ated with the (comparative) lay person already knows about: demonstrations of the antecedents
evaluation of the outcome, the Making a choice, whether of a va- and consequences of regret. Contin-
other with the feeling of self- cation destination, a spouse, a med- ued research through the 1990s
ical treatment, or a career, can be overturned some of these early

Published by Blackwell Publishing Inc.


Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014
CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE 213

findings, modified others, and led studied regret (though their use of on your decision. You knew at the
to a more complex and comprehen- the word differs from ordinary time that you had had too much to
sive understanding of regret in de- usage; Connolly & Butler, 2002). drink, and that options were avail-
cisions. These theorists predicted that people able. You have no excuse: Your de-
In this short review, we first would feel regret if a decision out- cision to drive was entirely unjusti-
sample a number of the initial come was worse than what they fied, and could easily have led to
studies, and trace some of the com- would have received by choosing a disastrous outcomes. The actual
plexities and contradictions to different option, and that they outcome was good, but you still
which they led. We then sketch a would try to avoid options that ex- feel regret.
new model that synthesizes these posed them to this possibility. Initial Compare this with another sce-
apparently conflicting findings and studies seemed to confirm these pre- nario: Imagine yourself as a parent
identify several research issues for dictions, but newer experiments trying to decide whether or not to
new work. (e.g., Starmer & Sugden, 1993) sug- vaccinate your 1-year-old against
gest that these studies were flawed. some serious disease. You consult
As a result, research in this tradition doctors, gather information, think
seems mainly to have stopped. long and hard, and decide to vacci-
EARLY STUDIES There is, however, good evidence nate, despite knowing the vaccine
that choices are influenced by antici- sometimes has bad side effects.
Perhaps the best-known of the pated regret, and that such anticipa- You are unlucky, and your child
early regret studies is by Kahneman tion is affected by whether or not one suffers the bad side effects. Of
and Tversky (1982). They asked stu- expects to learn the outcomes of un- course, you regret this misfortune:
dents to assess the regret that chosen options (Zeelenberg, 1999). The outcome is not what you had
would be felt by two investors, hoped (and would not have hap-
both of whom lose $1,200, one as a pened if you had decided against
result of buying a particular stock, vaccination). But, even in retro-
the other as a result of holding on DECISION JUSTIFICATION spect, you see no reason to blame
to the same stock. A stunning 92% THEORY yourself. You did everything a pru-
of the respondents guessed that the dent parent could to make the best
active buyer would feel more re- To accommodate these conflict- possible choice. Your decision was
gret than the passive holder. A bad ing results, we propose a model of fully justified (as comforting
outcome resulting from action regret called decision justification friends will remind you; being ab-
seemed to be more regrettable than theory (DJT). DJT postulates two solved from blame is a great way to
the same bad outcome when it was core components of decision-related reduce regret). Driving home
the result of inaction. regret, one associated with the drunk leaves you with self-blame
Intriguingly, just the opposite pat- (comparative) evaluation of the regret but (if you are lucky) no
tern emerged from a series of stud- outcome, the other with the feeling bad-outcome regret; an unfortu-
ies by Gilovich and Medvec (1995) of self-blame for having made a poor nate vaccination result leaves you
looking at retrospective regrets. Peo- choice. The overall feeling of regret feeling bad-outcome regret, but not
ple asked to recall real-life regrets at some decision is a combination the extra pain of self-blame.
tended to recall omissions more fre- of these two components: You Of course, most decisions that
quently than commissions, the ad- regret both that the outcome is result in bad outcomes generate
ventures they had passed up rather poorer than some standard (often some mixture of these two regret
than the leaps they had taken. the outcome of the option you re- components. Sometimes a decision
Gilovich and Medvec argued that jected) and that the decision you that felt well justified at the time
regret follows a characteristic tem- made was, in retrospect, unjustified. appears unjustified later. Crawford,
poral pattern: Regrettable actions The two components do not nec- McConnell, Lewis, and Sherman
hurt more than omissions in the essarily go together. Someone (2002), for example, induced exper-
short run, but when looking back, might experience high self-blame imental participants to follow the
people experience more regret over regret even when the outcome is (bad) advice of an unknown stranger
paths not taken. Research we de- good. Suppose that you leave a as to which of two football teams to
scribe later showed, however, that party somewhat inebriated and de- bet on, despite having detailed and
people may regret inactions more cide to drive home rather than take useful information on which to
than actions in the short run also. a cab. You arrive home safely, but base their bets. At the decision
Economic choice theorists (e.g., the following morning you are point, most participants followed
Loomes & Sugden, 1982) have also racked with regret as you think back the stranger’s advice. In retrospect,

Copyright © 2002 American Psychological Society


Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014
214 VOLUME 11, NUMBER 6, DECEMBER 2002

after they had bet and lost, they feel substantial regret for a bad but found the reverse for the
seemed to scold themselves: “How outcome, despite having no de- bolder investors, who were seen
could I have been so stupid? In ret- cision responsibility. Apparently as regretting losses from inac-
rospect, there was no reason for me (though neither of us interpreted tion more than losses from action.
to trust this anonymous tipster. My the results this way at the time), Apparently behavior can some-
decision was unjustified.” In this the computer-assigned group ex- times be justified by its consis-
case, the regret appears to have in- pected only outcome-evaluation tency with one’s personality.
volved both poor-outcome and regret, whereas the self-choice
self-blame components. group considered possible self- These are, of course, simply post
DJT provides a parsimonious ac- blame regret as well. (It is also hoc reinterpretations in DJT terms
count of many earlier findings on interesting that the key reference of findings from existing studies
regret. For example: point in all these studies was the (though we are impressed with the
status quo, the quality of the ease with which the theory re-
• Simonson (1992) offered stu-
course section initially assigned. solves apparently conflicting re-
dents two imaginary choices:
A variety of reference points sults). We have also undertaken
whether to take advantage of a
may be involved in regret- several direct tests of the theory. In
current, moderate sale or wait
related comparisons.) one (Connolly & Reb, 2002), we
for a later, possibly better sale,
took three scenarios (the vaccina-
and whether to buy a bargain- • In another study (Zeelenberg,
tion, two-investors, and soccer-coach
brand VCR or a more expensive, Van den Bos, Van Dijk, & Piet-
problems) and asked subjects to as-
well-known, high-quality brand ers, 2002), participants were
sess a range of emotions the pro-
(a SONY). Control subjects dis- asked how much regret a soccer
tagonists might feel if the decision
played no clear preference in ei- coach would feel if his team lost
turned out badly. We also asked
ther choice, but subjects asked to after he either changed or did
them to rate five short arguments
think about the regret they not change the team—a soccer
that one might use to guide deci-
might feel after making their de- parallel of the two-investors
sion making in each context (e.g.,
cision predominantly chose the problem discussed earlier. As in
“When it comes to [this topic], I
safer options (the current sale the case of the investors, partici-
just have to trust my gut instincts
and the SONY). Apparently pants expected more regret for
about what to do”; “It is better to
thinking about regret led them the active than for the inactive
sit still in situations like this than to
to look for justifications for their coach—but only if the team had
take action”). Half the participants
choices, and the safer brand and previously been enjoying a win-
evaluated these arguments before,
guaranteed sale offered the justi- ning record. If the team had
and half evaluated them after, as-
fication they were looking for. been losing, the inactive coach
sessing the protagonists’ likely
was seen as feeling more regret.
• In separate studies of the role of In DJT terms, the winning coach
emotions. Considering the argu-
decision responsibility in regret ments was intended to prime the
was unjustified in changing his
(Connolly, Ordóñez, & Coughlan, participants with a range of alter-
team, and can thus be blamed
1997; Zeelenberg, Van Dijk, & native rationales that could miti-
for the subsequent loss. In con-
Manstead, 2000), we explored gate the feelings of decision-related
trast, the losing coach was justi-
how students expected to feel if regret when things went wrong. In
fied in making changes, even if
they changed course sections two of the three scenarios, this is
they did not immediately work
and then found the section they exactly what we found: For both
out, and thus was expected to
moved into was better than, the vaccination and the soccer-
feel less blame, and less regret.
worse than, or the same as the coach scenarios, the active protago-
section they had left. Students • Seta, McElroy, and Seta (2001) ran nist was seen as significantly less
who made the decision for a between-subjects replication of regretful by participants who had
themselves expected to feel the two-investors problem, but just read the justificatory argu-
more regret if the outcome was attaching short personality ments than by those who had not.
bad (and more rejoicing if the sketches of the investors as ei- (Interestingly, the appropriateness
outcome was good) than stu- ther cautious risk avoiders or en- of the arguments varied signifi-
dents who had the same out- trepreneurial risk takers. They cantly across scenarios. A justifica-
come imposed on them by an ar- replicated Kahneman and Tver- tion that is perfectly appropriate for
bitrary computer reassignment. sky’s original result (more regret an investment decision may be
However, even the computer- expected for action than for inac- seen as inappropriate for a health-
assigned students expected to tion) for the cautious investors, care decision for one’s child.)

Published by Blackwell Publishing Inc.


Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014
CURRENT DIRECTIONS IN PSYCHOLOGICAL SCIENCE 215

A second test of DJT (Inman & people vary systematically in the influenced by, the choices people
Zeelenberg, 2002) studied con- justifications they find compelling, make. As researchers improve their
sumer regrets. In the scenarios used and whether people produce justi- understanding of regret and other
in this study, consumers either fications spontaneously or only un- decision-related emotions, people
made repeat purchases or switched der prompting. Understanding the should be able to improve their
products. In some experimental production of justifications might choices, and their feelings about the
conditions, subjects got information provide a basis for “regret ther- consequences of those choices. We
justifying the consumer’s decision apy,” for helping people find the feel good about the prospect.
(e.g., she had bought the product rationales that will work for them
only once before, found it unsatis- in reducing painful regret. It might Recommended Reading
factory, and switched); in other even help them make better deci-
conditions, the decision seemed un- sions. For example, when we asked Gilovich, T., & Medvec, V.H. (1995).
(See References)
justified (e.g., she had bought the subjects to rate justifications (Con-
Landman, J. (1993). Regret: The persis-
product several times before, liked nolly & Reb, 2002), one of the best tence of the possible. Oxford, En-
it, but switched anyway). The study general-purpose justifications was gland: Oxford University Press.
showed that regret over poor out- that one made a careful, competent Mellers, B.A., Schwartz, A., & Ritov,
comes was associated not with decision based on a wide range of I. (1999). Emotion-based choice.
Journal of Experimental Psychology:
whether the consumer switched or input information. If anticipated
General, 128, 332–345.
stayed, but with whether or not the regret led people to engage in Sugden, R. (1985). Regret, recrimina-
decision was justified. thoughtful decision making of this tion and rationality. Theory and
A third set of related studies kind, it could lead directly to im- Decision, 19, 77–99.
(Pieters & Zeelenberg, 2002) exam- proved decision making.
ined the regret people experience We are also eager to explore the
Acknowledgments—We thank Lisa Or-
when they behave in ways they did effects of decision-related rejoicing, dóñez and Jochen Reb for helpful com-
not originally intend. These studies if that is the positive emotion that ments on an earlier draft.
showed that such inconsistency, corresponds to regret. (Preliminary
which is often hard to justify, am- data in Connolly & Butler, 2002,
plifies regret independently of the suggest that regret and rejoicing Note
outcomes of the behavior. This in- may not be simple polar oppo-
consistency effect was obtained us- sites.) Certainly people expect posi- 1. Address correspondence to Terry
Connolly, Department of Management
ing a scenario approach (Study 1), tive decision outcomes to lead to
and Policy, 405Q McClelland Hall,
when people were asked about au- rejoicing, though as yet there is lit- University of Arizona, Tucson, AZ
tobiographical memories of regret- tle evidence that justification af- 85721; e-mail: connolly@u.arizona.edu.
table events (Study 2), and in a lon- fects such rejoicing. Good out-
gitudinal study of a significant comes seem not to provoke the
real-life decision, voting in national sorts of soul-searching that trigger References
elections (Study 3). These results regret and the search for justifica-
suggest that the notion of justifi- tion. Regret and rejoicing may thus Connolly, T., & Butler, D. (2002). Searching for the
“Regret” in “Regret Theory.” Manuscript sub-
ability as put forward in DJT may be driven by rather different mech- mitted for publication.
also explain other effects of deci- anisms, and are not necessarily Connolly, T., Ordóñez, L.D., & Coughlan, R.
(1997). Regret and responsibility in the evalua-
sion procedures on regret. mirror images of one another. tion of decision outcomes. Organizational Be-
The issues, then, are wide open. havior and Human Decision Processes, 70, 73–85.
The emotional side of decision mak- Connolly, T., & Reb, J. (2002). “No reason to blame
yourself”: Justification in decision-related regret.
ing is clearly important, but research- Manuscript submitted for publication.
FUTURE DIRECTIONS ers are only now starting to under- Crawford, M.T., McConnell, A.R., Lewis, A.C., &
Sherman, S.J. (2002). Reactance, compliance,
stand it. Compared with the huge and anticipated regret. Journal of Experimental
We have been surprised too research literature describing how Social Psychology 38, 56–63.
often in our studies of regret to people think, and should think, Gilovich, T., & Medvec, V.H. (1995). The experi-
ence of regret: What, when, and why. Psycho-
prophesy with any confidence about their decisions, the research on logical Review, 102, 379–395.
where the next steps will come, but how people feel in and about the de- Inman, J.J., & Zeelenberg, M. (2002). Regret in re-
peat versus switch decisions: The attenuating
there are plenty of excellent re- cision-making process is in its in- role of decision justifiability. Journal of Con-
search topics available. As yet, psy- fancy. The research we have sumer Research, 29, 116–128.
chologists know almost nothing sketched here considers one emotion, Kahneman, D., & Tversky, A. (1982). The psychol-
ogy of preferences. Scientific American, 246,
about the sorts of arguments that regret, and describes some of the sub- 160–173.
make for good justifications, whether tle ways in which it influences, and is Loomes, G., & Sugden, R. (1982). Regret Theory:

Copyright © 2002 American Psychological Society


Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014
216 VOLUME 11, NUMBER 6, DECEMBER 2002

An alternative theory of rational choice under Simonson, I. (1992). The influence of anticipating Zeelenberg, M., Van den Bos, K., Van Dijk, E., &
uncertainty. Economic Journal, 92, 805–824. regret and responsibility on purchase deci- Pieters, R. (2002). The inaction effect in the
Pieters, R., & Zeelenberg, M. (2002). On bad deci- sions. Journal of Consumer Research, 19, 105–118. psychology of regret. Journal of Personality and
sions and deciding badly: Intention-behavior in- Starmer, C., & Sugden, R. (1993). Testing for juxta- Social Psychology, 82, 314–327.
consistency as a source of regret. Manuscript position and event-splitting effects. Journal of
submitted for publication. Risk and Uncertainty, 6, 235–254. Zeelenberg, M., Van Dijk, W.W., & Manstead,
Seta, J.J., McElroy, T., & Seta, C.E. (2001). To do or Zeelenberg, M. (1999). Anticipated regret, ex- A.S.R. (2000). Regret and responsibility re-
not to do: Desirability and consistency medi- pected feedback and behavioral decision-mak- solved? Evaluating Ordóñez and Connolly’s
ate judgments of regret. Journal of Personality ing. Journal of Behavioral Decision Making, 12, (2000) conclusions. Organizational Behavior and
and Social Psychology, 80, 861–870. 93–106. Human Decision Processes, 81, 143–154.

Is Laughter the Best Medicine? Humor, gested, for example, that vigorous
laughter exercises and relaxes mus-
Laughter, and Physical Health cles, improves respiration, stimulates
circulation, increases the production
Rod A. Martin1 of pain-killing endorphins, decreases
Department of Psychology, University of Western Ontario, London, Ontario, Canada the production of stress-related hor-
mones, and enhances immunity. Ac-
cording to this theoretical model,
cate the use of “therapeutic hu- hearty laughter is crucial in the hu-
Abstract mor-health connection, whereas hu-
This article examines re- mor” in the treatment of illness and
maintenance of health, and clowns morous perceptions and amusement
search evidence for the popu- without laughter would not be ex-
lar idea that humor and and comedy carts have become fa-
miliar sights in many hospitals. pected to confer any health benefits.
laughter have beneficial ef- Second, humor and laughter
fects on physical health. Poten- The idea that laughter is good for
one’s health can be traced to bibli- might affect health by inducing pos-
tial theoretical mechanisms for itive emotional states, which may
such effects are discussed first. cal times, and was revived periodi-
cally by various physicians and in turn have beneficial effects on
Empirical evidence for benefi- health, such as increasing pain tol-
cial effects of humor and philosophers through the centuries.
In recent decades, interest in the erance, enhancing immunity, and
laughter on immunity, pain undoing the cardiovascular conse-
tolerance, blood pressure, lon- healing power of laughter was
given new impetus by the best-sell- quences of negative emotions (Fred-
gevity, and illness symptoms is rickson, 2000). Compared with the
then summarized. Overall, the ing account by Cousins (1979) of
his recovery from a progressive and first model, this model gives humor
evidence for health benefits of and laughter a less unique role in
humor and laughter is less con- painful rheumatoid disease after
treating himself with daily bouts of health enhancement, as they are only
clusive than commonly be- one means of increasing positive
lieved. Future research in this laughter, along with massive doses
of vitamin C. emotions, along with love, joy, op-
area needs to be more theoreti- timism, and so forth. Furthermore,
cally driven and methodologi- according to this model, overt
cally rigorous. laughter may not even be neces-
THEORETICAL MECHANISMS sary for health benefits to occur, be-
Keywords cause humor and amusement may
humor; laughter; health; im- How might humor and laugh- induce positive moods even with-
munity; pain ter influence physical health? out laughter.
There are at least four potential Third, humor might benefit
mechanisms, each involving a dif- health indirectly by moderating the
Belief in beneficial effects of hu- ferent aspect of humor, and each adverse effects of stress on health.
mor and laughter on physical suggesting different implications A considerable body of research in-
health has become increasingly for the application of humor to dicates that stressful life experi-
popular in recent years. The media well-being. First, laughter might ences can have adverse effects on
frequently report claims about sci- produce physiological changes in various aspects of health, including
entific evidence for health benefits various systems of the body, which suppression of the immune system
of humor and laughter. Some prac- may have beneficial effects on and increased risk of infectious dis-
titioners have even begun to advo- health. Various authors have sug- ease and heart disease (O’Leary,

Published by Blackwell Publishing Inc.


Downloaded from cdp.sagepub.com at JOHNS HOPKINS UNIVERSITY on April 13, 2014

You might also like