This article was downloaded by: [University of Nebraska, Lincoln]

On: 04 January 2015, At: 09:54

Publisher: Routledge
Informa Ltd Registered in England and Wales Registered Number: 1072954
Registered office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH,
UK

European Review of Social

Psychology
Publication details, including instructions for
authors and subscription information:
http://www.tandfonline.com/loi/pers20

Affect, Attitudes and Decisions:

Let's Be More Specific
a a
Joop van der Pligt , Marcel Zeelenberg , Wilco W.
a a a
van Dijk , Nanné K. de Vries & René Richard
a
University of Amsterdam
Published online: 15 Apr 2011.

To cite this article: Joop van der Pligt , Marcel Zeelenberg , Wilco W. van
Dijk , Nanné K. de Vries & René Richard (1997) Affect, Attitudes and Decisions:
Let's Be More Specific, European Review of Social Psychology, 8:1, 33-66, DOI:
10.1080/14792779643000074

To link to this article: http://dx.doi.org/10.1080/14792779643000074

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the
information (the “Content”) contained in the publications on our platform.
However, Taylor & Francis, our agents, and our licensors make no
representations or warranties whatsoever as to the accuracy, completeness,
or suitability for any purpose of the Content. Any opinions and views
expressed in this publication are the opinions and views of the authors, and
are not the views of or endorsed by Taylor & Francis. The accuracy of the
Content should not be relied upon and should be independently verified with
primary sources of information. Taylor and Francis shall not be liable for any
losses, actions, claims, proceedings, demands, costs, expenses, damages,
and other liabilities whatsoever or howsoever caused arising directly or
indirectly in connection with, in relation to or arising out of the use of the
Content.
 This article may be used for research, teaching, and private study purposes.
Any substantial or systematic reproduction, redistribution, reselling, loan,
sub-licensing, systematic supply, or distribution in any form to anyone is
expressly forbidden. Terms & Conditions of access and use can be found at
http://www.tandfonline.com/page/terms-and-conditions
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015
 Chapter 2
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Affect, Attitudes and Decisions: Let’s

Be More Specific
Joop van der Pligt, Marcel Zeelenberg, Wilco W. van Dijk,
Nanne K. de Vries and Rend Richard
University of Amsterdam

ABSTRACT
This chapter focuses on the role of affect in attitudes and decision-making. First
we will briefly discuss the role of affect in attitude-formation and -change pro-
cesses. Two issues have played an important role in this research area: first, the
distinction between affect-based and cognition-based attitudes; second, the
effects of mood on persuasion. Generally these traditions rely on a crude dicho-
tomy between positive and negative affect and rather general, holistic measures
of affect. Moreover, these traditions tend to emphasize automatic information
processing. We focus on controlled information processing and continue with a
discussion of the role of affect in expectancy-value models of behaviour such as
Ajzen’s theory of planned behaviour. Affect received only limited attention in
these models. It will be argued that people anticipate post-behuviourul affective
consequences of their actions, and take these into account when deciding about
their behavioural preferences. We will argue that the inclusion of anticipated
postbehavioural affective outcomes could improve the predictive validity of
expectancy-value models. Next, we will contrast research on affect and attitudes
with research on behavioural decision-making. The latter area tends to focus on
more specific affective determinants of behaviour. One of these is anticipated
regret. Antecedents of anticipated regret will be discussed and the predictive
validity of anticipated regret will be tested in the context of Ajzen’s theory of
planned behaviour. Finally we will show that it is relatively easy to increase the
salience of postbehavioural affective reactions such as regret and worry and

European Review of Social Psychology, Volume 8. Edited by Wolfgang Stroebe and Miles Hewstone.
0 1998 John Wiley & Sons Ltd.
 34 JOOP VAN DER PLlGT ET AL.

that this increased salience has an impact on both behavioural intentions and
self-reported behaviour. Implications for the study of affect in expectancy-value
models of behaviour will be briefly discussed.
Preferences are a central issue in social psychology and are examined in
research on attitudes, impression formation, interpersonal relationships,
decision-making, and many other fields. For the past few decades the prevail-
ing paradigms in these fields of research treated preferences as the outcome of
a cognitive process focusing on utilities and values. Moreover, decomposing
these utilities and values into more elementary components was a general
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

solution to improve our understanding of these processes. The one factor

often missing in these approaches was affect. This emphasis on cognition was
corrected in the 1980s, resulting in a marked increase in research on the role
of affect in social judgment, attitude-formation and -change processes, and
decision-making.
In this chapter we will discuss research on the role of affect in attitudinal
processes. First, we will briefly summarize research on affective information
processing and the effects of mood on attitudes. Next we turn to the main
issue of this chapter, the role of affect in expectancy-value models of attitudes.
We will focus on the conscious processing of affect and contrast research on
the role of affect in expectancy-value models of attitudes with research on
behavioural decision-making. We will show that affect can play an indepen-
dent role in attitudes and behavioral decisions. It will be argued that incor-
porating affective factors in expectancy-value models such as Ajzen’s theory
of planned behaviour requires less holistic measures of affect, should focus on
specific emotional reactions, and should also take account of anticipated,
postbehavioural affective reactions.
The renewed interest in the role of affect can be traced to Zajonc’s (1980)
influential paper on affective vs. cognitive information processing. Zajonc,
on the basis of his mere exposure studies, argued for separate systems and
suggested that affective and cognitive processes may proceed independently
from one another and that affective reactions could even precede and influ-
ence cognition. Basically, he argued that the affective quality of a stimulus
can be processed outside of conscious awareness. Researchers of the affect-
cognition interface often used subliminal presentation of affective primes to
demonstrate evaluative reactions t o stimuli that cannot be traced to some
conscious computation of liking. For instance, Edwards (1990) and Murphy
and Zajonc (1993) used subliminal presentation of faces showing a positive
or negative emotion to prime affective reactions to a subsequent (ambig-
uous) stimulus. Generally, this research relies on a “crude” dichotomy of
affect by valence (good or bad), and it can be placed in the context of a
larger tradition in social psychology concerning the role of awareness and
control on a person’s judgements, and research on automatic phenomena in
social judgement (see Bargh, 1994). More recent findings (Bargh et al., 1996)
 AFFECT, ATTITUDES AND DECISIONS 35

in this line of research confirm the so-called automatic evaluation effect, and
show that attitudes can be activated without conscious processing and facil-
itate or interfere with the conscious and intended evaluation of a target
stimulus.
Based on Zajonc’s (1980) distinction between feeling and thinking,
Abelson et al. (1982) explored comparisons between conventional semantic
judgements focusing on a cognitive appraisal of attitude objects, and affec-
tive responses to these objects. In two large-scale surveys, respondents were
asked to ascribe personality traits to prominent national politicians as well
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

as to report the feelings that these politicians elicited. Affect elicited by the
politicians was highly predictive of political preference, and this effect was
independent of, and more powerful than, the effect of trait judgements. On
the basis of these findings, Abelson et al. (1982) concluded that affective
responses add to the predictive validity of attitude measures beyond that
available from standard semantic judgements. Moreover, both Zajonc
(1980) and Abelson et al. (1982) proposed that individual preferences can be
based on affect per se.
Other researchers did not deal directly with this question but obtained
results suggesting that affect does influence attitudes. A variety of research
findings including behaviours such as energy conservation (Seligman et al.,
1979), health behaviour (Ajzen & Timko, 1986). responses to victimization
(Tyler & Rasinsky, 1984), and contraceptive behaviour (Fisher, 1984) sug-
gests that affect can have a strong and independent impact on attitudes.
Zanna and Rempel (1988) also propose to distinguish affective and cogni-
tive aspects of attitudes. They regard an attitude as the categorization of a
stimulus object along an evaluative dimension, and argue that this evalua-
tion can be based upon three different sources of information: (a) cognitive
information, (b) affective/emotional information, and (c) information con-
cerning behaviour in the past. Their view goes back to Rosenberg and
Hovland’s (1960) “three-component’’ view of attitudes. Zanna and Rempel
build upon this early work, and argue that it is possible for an attitudinal
judgement to be dependent strictly on cognitive beliefs, but also on affect or
on past behaviour vis-u-vis the attitude object. These beliefs, feelings and
behaviours are not mutually exclusive and constitute different ways in which
the attitude is formed and experienced. Moreover, the three components
need not be consistent and can have different valences. It needs to be added
that Zanna and Rempel’s approach seems to have the same shortcoming as
Rosenberg and Hovland’s view; i.e., it is not specified when and under what
conditions attitudes will be based on one source of information rather than
another. Most important for our discussion is that Zanna and Rempel’s
conceptualization implies that evaluation and affect are different compo-
nents of attitudes. Further support for this view is provided by Breckler
(1984) and Breckler and Wiggins (1989).
 36 JOOP V A N D E R PLIGT E 7 AL.

Breckler and Wiggins (1989) proposed to use the term “evaluation” (as
opposed to “cognition”) to refer to attitudes based on judgements abouf the
attitude object because cognition can include the appraisal of emotions and
emotion-related functioning (see also Lazarus, 1981, 1982. 1984). They re-
served the term “affect” for emotional responses and feelings engendered by
an attitude object. In their study they collected multiple measures of evalua-
tion and affect in six attitude domains (blood donation, legalized abortion,
computers, nuclear weapons, standardized admission tests, and college com-
prehensive examinations). Although they relied on self-report measures, cor-
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

relations between affective and evaluative responses showed considerable

variation. Results of a second study on attitudes towards blood donation
indicated a modest correlation between affect and evaluation and a stronger
relationship between affect and (self-reported) behaviour than between eval-
uation and behaviour.
Other research focused on affect in the context of attitude change processes
and persuasion. For instance, Edwards (1990) distinguished between affect-
and cognition-based attitudes, and investigated the effectiveness of affective
and cognitive means of persuasion. Her findings showed that affect-based
attitudes (created by a subliminal or supraliminal affective prime) exhibit
more change under affective means of persuasion than under cognitive means
of persuasion. Cognition-based attitudes (created by the provision of informa-
tion about the attitude object) exhibited equal change under both forms of
persuasion. It needs to be added, however, that research on this topic has
yielded conflicting findings. Millar & Millar (1990) found that cognition-based
attitudes were more easily changed by affective means than cognitive means,
and vice versa. Contrary to Edwards (1990) who used Chinese ideographs as
attitude objects, they relied on attitude object with which people were already
familiarized.
Mood (as a general affective state without a specific focus) has also been
related to attitudinal change processes, especially in the context of Petty and
Cacioppo’s Elaboration Likelihood Model (e.g. Bless ef al., 1990; Petty et al.,
1994). Generally, individuals in an elated mood are less likely to engage in
systematic message elaboration than individuals in a neutral or depressed
mood. According to the Elaboration Likelihood Model (Petty & Cacioppo,
1986), affective (peripheral) cues are particularly potent determinants of atti-
tude change when the ability or motivation to process issue-relevant informa-
tion is low.
Both the literature on mood and persuasion and the literature on affect vs.
cognition-based attitudes focus on the role of a universal and ubiquitous
characteristic of affect-valence. This seems partly a consequence of the tend-
ency to contrast affect with cognition (especially in research on the primacy of
affect and research on cognition vs. affect-based attitudes). One unfortunate
consequence of this contrast is that much attention has been paid to this
 AFFECT, ATTITUDES AND DECISIONS 37

(sometimes rather artificial) dichotomy at the expense of adequate and dif-

ferentiated measures of cognitive and affective determinants of attitudes and
behaviour.’
Some of the studies described earlier (e.g. Breckler & Wiggins, 1989) relied
on scales that present generally applicable evaluative word pairs and instruct
respondents to indicate how (e.g. good or bad) the attitude object makes them
feel (to assess affect) or how good or bad the attitude object is (to assess
evaluation). Using the same scales to assess the affective and cognitive com-
ponent of attitudes has some advantages. For instance, Eagly, Mladinic, and
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Otto (1994) argue that the use of different sets of items to assess affective and
cognitive components of attitudes might result in one set being more satu-
rated with evaluative meaning than the other set. The disadvantage of using
the same set of items is, however, that one has to rely heavily on general
evaluative terms that can be used meaningfully when accompanied by ques-
tions referring to feelings triggered by the attitude object as well as when
accompanied by questions referring to characteristics of the attitude object.
As argued before, one inevitable consequence of such an approach is that the
measurement of affect is limited to a crude positive-negative classification.
Another disadvantage is that, because of the rather general and widely appli-
cable items, one could miss (cognitive or affective) attributes that are
extremely relevant for judging a specific attitude object. Moreover, if the
under-representation of relevant attributes is not evenly distributed over the
affective and cognitive components of attitudes, this could have serious
consequences for the assessment of their relative impact on attitudes.
Crites, Fabrigar, and Petty (1994) argued that the assessment of affective
and cognitive properties of attitudes is plagued by a number of problems.
They mention a lack of consistency in procedures across studies, the possible
impact of structural characteristics of the measures of affect and cognition on
responses, and the limited attention paid to the reliability and validity of
scales to assess these properties of attitudes. Crites, Fabrigar, and Petty (1994)
developed scales for assessing the affective and cognitive properties of atti-
tudes and examined their reliability and validity. Their aim was to develop
more general scales that could be applied to a variety of attitude objects.
Their affective scale was based on eight affective word pairs; their cognitive
scale focused on more utilitarian dimensions and consisted of seven “cogni-
tive” pairs. Their analyses showed that the measures were applicable to mul-
tiple attitude objects and had good and stable psychometric properties.
Contrary to most of the literature discussed earlier in this chapter, Crites,

There is one notable exception: research on the effects of fear-arousing messages on (preven-
tive) behaviour. This older tradition (Janis, 1967; Leventhal, 1970 Rogers, 1975) differs from the
research described in the previous paragraphs because it focuses on one specific emotion (fear) as
opposed to general affect: and it also tends to pay more attention to behavioural consequences.
 38 JOOP V A N DER PLlCT ET A L .

Fabrigar, and Petty (1994) stress that affect is not an undifferentiated compo-
nent of attitudes but consists of discrete, qualitatively different emotions.
Although Crites, Fabrigar, and Petty propose a more differentiated approach,
they still focus on a more general index score, combining the various items
into an overall score that primarily reflects the valence of the affective deter-
minants of attitudes.
Most of the research described in this section focused on attitudes without
explicitly addressing behaviour. For instance, Abelson et al. (1982) studied
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

attitudes towards political candidates, and Edwards (1990) used Chinese ideo-
graphs as attitude objects. In the remainder of this chapter we will focus on
attitudes towards behavioural options. This brings us to research on attitude-
behaviour models.

AFFECT IN ATTITUDE-BEHAVIOUR MODELS

Fishbein and Ajzen’s (1975) theory ofreasoned action and Ajzen’s (1985, 1991)
theory of planned behaviour assume that attitudes are an important determi-
nant of intentions. In their approaches attitudes are based on the summed
products of the likelihood of consequences associated with behavioural actions
and the evaluation of these consequences. Thus, the more positive con-
sequences are associated with a specific behaviour and the more likely their
occurrence, the more attractive the behaviour is. Fishbein and Ajzen (1975)
also incorporated perceived social norms in their model, and Ajzen added
perceived behavioural control as a third determinant of behavioural intentions.
Both theories have been applied to a wide variety of behaviours. such as shop-
ping (Madden. Ellen, & Ajzen, 1992), food intake (Sparks, Hedderly, &
Shepherd, 1992), violations of traffic regulations (Parker et al., 1992) weight
reduction (Schifter & Ajzen, 1985), dental hygiene (McCaul et al., 1993) and
contraceptive use (Middlestadt & Fishbein, 1995; Chan & Fishbein. 1993).
Both the theory of reasoned action and the theory of planned behaviour
thus assume an informational foundation of human conduct, focusing on the
expected outcomes of behavioural action. Individuals are expected to respond
to the analytic features of stimuli. This approach is also central to normative
theories of decision-making to which we will turn later. Initially, Fishbein and
Ajzen were rather unspecific about the role of affect, as illustrated by the
following quote:

The term “affect“ and “evaluation” are used synonymously throughout this
book. Although it might be argued that there is a difference between a person’s
judgement that an object makes him feel good and his evaluation that the object
is good, there is little evidence to suggest that a reliable empirical distinction
between the two variables can be made (Fishbein & Ajzen, 1975, p. 11).
 AFFECT, ATTITUDES AND DECISIONS 39

This view is different from Zanna and Rempel’s (1988) proposal discussed
earlier. In the theory of reasoned action affect is represented through evalua-
tions of each of the possible outcomes (outcome evaluations) and through an
overall evaluation of the attitude object (attitude). Fishbein and Ajzen (1975)
argue that learning (both classical conditioning and other forms of learning)
may determine affective reactions towards particular behaviours, objects or
specific outcomes. Fishbein and Ajzen thus also rely on a simple positive-
negative dichotomy of affect and, more importantly, “evaluation” and “af-
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

fect” are used synonymously and both are assumed to be determined by

outcome related beliefs and their evaluations.
Ajzen’s (1989) theory of planned behaviour also pays limited attention to
the role of affective processes. Ajzen admits that the role of affect is not
spelled out very clearly in the Ajzen-Fishbein framework. In his view, affec-
tive reactions may depend at least in part on cognitions and, like cognitions,
they may feed into the overall evaluative response to an attitude object.
Moreover, “Affect may also be associated with the perceived attributes of the
attitude object and thus also determine the evaluative direction and intensity
of a person’s beliefs” (Ajzen, 1989, p. 248).
Manstead and Parker (1995) argue that the evidence showing that evalua-
tive responses based on affective reactions (i.e., specific emotions) are em-
pirically distinguishable from measures of evaluative responses based on
beliefs about the attitude object is sufficiently strong to also incorporate this
distinction in attitude-behaviour models. Their view is supported by the find-
ings of Breckler and Wiggins (1989) and Edwards (1990), discussed earlier in
this chapter. More recently, Eagly, Mladinic, and Otto (1994) and Pfister and
Bohm (1992) provided further support for this distinction. Indirect evidence
for this view is provided by research focusing on the relative contribution of
direct, holistic measures of attitudes and indirect (belief-based) measures of
attitudes. Quite often the two measures do not correspond as well as they
should (see, e.g., Terry, Gallois, & McCamish (1993)). This could be related to
the fact that some relevant beliefs are not included in the set of beliefs
presented to the respondents. Another reason for this modest correspondence
could be that the direct measure of attitudes is more likely to also tap affective
reactions that generally are not included in the indirect belief-based measure.
The latter tends to be dominated by (instrumental) outcome beliefs, in which
affective or emotional outcomes play only a minor role.
Manstead and Parker (1995) define affective evaluations of behaviour as
referring to an individual’s positive or negative feelings about performing the
behaviour in question. This is in contrast to the behavioural beliefs typically
tapped in the context of attitude-behaviour models such as the theory of
planned behaviour. As indicated in the previous paragraphs, these beliefs
usually focus on the utilitarian aspects of the outcomes of the behaviour (i.e.,
costs and benefits). This is in accordance with subjective expected utility
 40 JOOP V A N DER PLIGT € T AL.

(SEU) approaches on which Fishbein and Ajzen’s theory is based. Later in

this chapter we will discuss SEU theory in more detail.
Ajzen and Driver (1991) were probably the first to investigate directly the
usefulness of the distinction between affective evaluations and behavioural
beliefs in relation to the theory of planned behaviour. Their study focused on
leisure activities. Results provided some support for the discriminatory valid-
ity of the measures of these two variables, but treating the two variables as
separate factors did not result in a significant improvement in the predictive
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

ability of the model. Manstead and Parker (1995) also report some prelimin-
ary findings showing that measures based on affective evaluations (feelings
associated with the behaviour) correlate only moderately with measures
based on behavioural beliefs (beliefs about the outcomes of a behavioural
action). In their research on driving behaviour they report correlations as low
as 0.14 and 0.29. These modest correlations between affect-based and
cognition-based measures were obtained in one specific domain (driving vio-
lations). Manstead and Parker rightly argue that it would be interesting to see
whether similar findings are found in other domains. Richard, van der Pligt,
and de Vries (1996a) attempted to do this; and it is to that research we turn
next.
Research on the impact of affect on attitudes usually relies on a general
measure of positivehegative affect associated with the behaviour. Attitude-
behaviour models, however, deal with future behaviour, and belief-based
measures of attitudes focus on the possible consequences of behavioural ac-
tions. Generally, research in this tradition presents respondents with a set of
possible consequences of a behavioural action and asks them to assess their
likelihood and their evaluation. Richard van der Pligt and de Vries (1996a)
argued that it would be appropriate to also incorporate anticipated
postbehavioural affective reactions as possible consequences of behavioural
actions.
In their study, Richard, van der Pligt, and de Vries (1996a) assessed the
evaluative response (attitude) towards a number of behaviours, affective reac-
tions towards these behaviours, and anticipated postbehavioural affective re-
actions. These three concepts differ in terms of time perspective and the
affect-evaluation distinction. The aim of their study was to investigate the
discriminant validity of the three measures. The inclusion of both general
affective reactions and anticipated affective reactions allowed them to investi-
gate the predictive utility of anticipated affective reactions and general affect
over and above attitudes and other components of the theory of planned
behaviour. Richard, van der Pligt, and de Vries selected four behaviours:
eating “junk food”, using soft drugs (marijuana, hashish), drinking alcohol,
and studying hard. They relied on a direct attitude measure; respondents were
asked to evaluate each of the four target behaviours on three semantic dif-
ferential scales: pleasant-unpleasant, nice-awful, and good-bad. Next,
 AFFECT, AITITUDES AND DECISIONS 41

respondents were asked to indicate their general feelings and their anticipated
postbehavioural feelings towards each of the target behaviours on the same
set of scales. Thus the same scales were used to assess the three evaluative
scores (attitudes, general feelings and anticipated postbehavioural feelings).
Richard, van der Pligt, and de Vries (1996a) did this to prevent one set of
items being more saturated with evaluative meaning than the other set (see
our earlier discussion of the arguments presented by Eagly, Mladinic, & Otto,
1994).
As expected, eating junk food, using soft drugs, and drinking alcohol were
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

associated with negative anticipated affective reactions, and these were more
negative than both the evaluations of, and general affective reactions towards,
the target behaviours. Similarly, anticipated affective reactions after “having
studied hard” were more positive than both attitudes and general affective
reactions towards this behaviour. Overall, evaluations (attitudes) towards
each target behaviour did not differ significantly from the general affective
reactions associated with the behaviour. Anticipated affective reactions,
however, differed substantially from both evaluations and general affective
reactions.
Richard, van der Pligt, and de Vries (1996a) compared two models, one
with two separate factors (evaluations and anticipated affect) and one which
combined these two factors into one overall factor. The two-factor models
fitted the data well, whereas all one-factor models were statistically rejected.
More importantly, further testing revealed that for all four behaviours the
two-factor model (separating anticipated affect and evaluation) fitted the data
significantly better than the one-factor model. Overall, these findings sup-
ported the discriminant validity of their measure of anticipated affect; antici-
pated affect proved different from both more cognitive evaluations and
general affect associated with the behaviour.
For three of the four investigated behaviours, anticipated postbehavioural
affective reactions predicted a significant proportion of variance in be-
havioural expectations, over and above the components of the theory of
planned behaviour. The only exception concerned the behaviour with rela-
tively positive anticipated postbehavioural affective reactions (“studying
hard”). Thus, the results of this study show that the predictive power of the
theory of planned behaviour may improve if anticipated postbehavioural
affective reactions are incorporated in the model. Figure 2.1 illustrates the
findings for “using soft drugs”. The regression parameters in this Figure
should be interpreted as P-weights; zero-order correlation coefficients are
given in brackets. Although the estimated correlation between “anticipated
affect” and “attitudes” was 0.76, the two-factor model fitted the data better
than the one-factor model combining anticipated affect with attitudes.
Two shortcomings of the research by Richard, van der Pligt, and de Vries
(1996a) need to be noted. First, they opted for the solution to use the same
 42 JOOP V A N DER PLlGT ET AL.

?I attitudes

I !- .76 \
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

h .I3 (.!
\

Y ::

Ll
.80 self-repotted
I II -.31 behaviour

I .42

Figure 2.1 The role of anticipated affect as a determinant of “using soft drugs”. Note:
parameter estimates are standardized; all parameters are significant at p < 0.01 except
ap < 0.10. Adapted from Richard, van der Pligt, and de Vries (1996a, p. 122)

scales to assess evaluations (attitudes), general affect and anticipated affective

reactions. They did not include a belief-based indirect measure of attitudes,
but relied on a general, direct measure of attitudes. Obviously, anticipated
postbehavioural affective reactions can also be included in the set of possible
outcomes that constitutes the indirect attitude measure. Opting for such a
solution would have allowed another test of the independent contribution of
anticipated affect. Moreover, it would have allowed for the assessment of the
subjective probability of the anticipated affective outcomes. People are likely
 AFFECT, ATTITUDES AND DECISIONS 43

to have different views on the subjective probability of particular affective

outcomes just as they tend to vary with regard to the subjective probabilities
they attach to other kinds of outcomes.
A second potential shortcoming of the study is that Richard, van der Pligt,
and de Vries (1996a) as many other studies in this field of research, relied on a
simple dichotomy, positive vs. negative affect, and assessed (anticipated) af-
fective reactions at a rather general level (e.g., “expecting to feel good vs.
expecting to feel bad”) without paying attention to more specific affective
reactions that could determine attitudes and/or behaviours. Overall, the main
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

contribution of Richard, van der Pligt, and de Vries’s (1996a) paper is that it
points at the possible role of anticipated postbehavioural affective reactions in
attitude-behavioural models. However, more research is needed to assess the
independent role of anticipated affect in expectancy-value approaches to atti-
tudes. This research should not only rely on general measures of affect but
also include specific anticipated affective reactions in the set of outcome
beliefs that constitutes the indirect measure of attitude.
Interestingly, research in the related area of behavioural decision-making
emphasizes the role of specific affective reactions and does not rely on the
crude positive-negative dichotomy of affect we tend to see in research on
attitudes. Several researchers in the area of behavioural decision-making
stressed the need to look at more specific emotions and studied the impact of
a variety of anticipated emotions on human decision-making. These include
guilt, sadness and anger (Baron, 1992), regret and disappointment (e.g., Bell,
1982, 1985; Loomes & Sugden, 1987b), and envy and gloating (Loewenstein,
Thompson, & Bazerman, 1989).
Research on emotions has also spent considerable effort to distinguish
between the various emotions. There is some dispute as to whether there are
such things as “basic emotions” (see, e.g., Ortony & Turner, 1990), but
present-day emotion theorists agree that there are important differences be-
tween emotions. Research findings show that emotions can be differentiated
by their accompanying thoughts and feelings, by their appraisal, physiological
activity, expression, action tendencies, and behavioural actions (Frijda,
Kuipers, & ter Schure, 1989; Roseman, Wiest & Swartz, 1994). This research
also indicates that different emotions with the same valence have different
implications for behaviour. For example, fear evokes a tendency to flight,
whereas anger evokes a tendency to fight. Similarly, the experience of regret
tends to make people active and problem-oriented, while disappointment,
albeit its strong relation to regret, tends to be related to a more passive
reaction focusing our attention away from the problem (Zeelenberg er al.,
1997b).
Distinguishing between various affective or emotional states could help our
understanding of anticipated, postbehavioural affective reactions and their
role in attitudinal and decision processes. As mentioned earlier, researchers in
 44 JOOP VAN DER PLlGT ET AL.

the area of behavioural decision-making have focused on more specific emo-

tions. In the next section we turn to this field of research.

AFFECT AND DECISION-MAKING

The increased attention for affect since the early 1980s was not restricted to
research on emotions and social cognition but is also apparent in research on
behavioural decision-making, While the renewed interest in affect in social
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

psychology can be traced to Zajonc’s 1980 article on preferences and in-

ferences, the renewed interest in the area of decision-making was mainly
caused by the inability of normative models of decision-making to explain and
describe preferences. Interestingly, research on decision-making did not rely
on the rather holistic approach to affect that dominated social psychological
research on attitudes, but focused on more specific emotions. This could well
be caused by methodological differences between the two areas. One import-
ant difference between the two fields is that behavioural decision-making
research almost always uses choice of some sort as a dependent variable,
whereas attitude research focuses on evaluative judgements and behavioural
intentions. The fact that behavioural choice is a prime dependent variable in
research on decision-making probably resulted in a preference to study con-
crete emotions with clear consequences for (choice) behaviour. To illustrate
how this came about we will first briefly describe the normative model of
decision-making that is central to this field of research.
Edwards’ (1954) Subjective Expected Utility (SEU) Theory is a normative
theory of decision-making and has had a profound impact on research on
behavioural decision-making. His expectancy-value approach also inspired
Fishbein and Ajzen’s (1975) theory of reasoned action and Ajzen’s (1991)
theory of planned behaviour. Edwards assumed that people generally aim to
maximize utility and prefer behavioural options that are associated with the
highest expected utility. The overall utility or desirability of a behavioural
alternative is assumed to be based on the summed products of the probability
and utility of specific outcomes or consequences. Thus:
SEUj = Ii:Pii . Uij
where SEUj is the SEU of action or behavioural alternative j ; Pii is the per-
ceived probability of outcome i of action j ; ,/!t is the subjective utility or
desirability of outcome i of action j. Each action or behavioural alternative
may have a different subjective expected utility because of the outcomes
associated with that action and/or the probabilities of these outcomes. Quite
often, however, people d o not behave in accordance with SEU-theory, and
this behaviour is frequently called “irrational”. There is considerable
 AFFECT, ATTITUDES AND DECISIONS 45

disagreement about the use of the term “irrational” in this context (see, e.g.,
Cohen, 198l), and it has been argued that discrepancies between normative
models such as SEU-theory and actual decision-making does not warrant the
use of this term. Some researchers attempted to improve the descriptive
validity of SEU-theory and suggested further refinements (e.g., Kahneman &
Tversky, 1979). Others proposed to incorporate additional aspects that should
be taken into account when investigating decision-making processes.
Kahneman and Snell (1992) argued that most research and theorizing in
decision-making rely on a severely impoverished conception of utility. In their
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

view, rigid operationalism has led to a situation in which little is left of the
original broader sense of utility in the writings of Bentham and Bernoulli, who
related utility to the hedonic quality of experience. Kahneman and Snell
(1990) propose to distinguish two concepts of utility: first, the decision utility
of an outcome which is defined as the sign and weight of that outcome in the
context of choice; second, the experienced utility, which is defined by the
quality and intensity of the hedonic experience with that outcome. Decision
utility focuses on the decision rather than on experience and is linked to the
basic assumption of rationality: i.e., rational individuals can be trusted to
know what will be good for them and are entitled to their sovereignty (cf.
Kahneman & Snell, 1990, p. 296). This basic assumption is also central to
expectancy-value models such as those of Fishbein and Ajzen. Kahneman and
Snell (1992) investigated people’s ability to predict experienced utility over
time for stimuli such as ice creams, yoghurt and short musical pieces. This
ability turned out to be limited. The major contribution of their approach is
that they tried to broaden the concept of utility, partly in order to improve the
descriptive validity of decision theory. Other researchers opted for more spe-
cific solutions, and to these we turn next.

Regret and Decision-making

Bell (1982) and Loomes and Sugden (1982) argued that incorporating one
specific emotion (anticipated regret) in SEU theory could help to improve its
descriptive validity. Regret theory differs from classical theories of decision-
making, such as SEU theory, which are based on the same premises as
attitude-behaviour models such as those of Ajzen and Fishbein. In regret
theory the utility of a choice option additionally depends on the feelings
evoked by the outcomes of rejected options. Regret theory rests on two basic
assumptions (Loomes & Sugden, 1982). The first holds that people compare
the actual outcome with what the outcome would have been, had a different
choice been made, and that they experience emotions as a consequence.
People experience regret when the foregone outcome would have been bet-
ter, and rejoicing when the foregone outcome would have been worse. The
second assumption of regret theory is that the emotional consequences of
 46 JOOP V A N DER PLIGT E T AL.

decisions are anticipated and taken into account when making decisions. Thus,
the tendency to avoid negative postdecisional emotions such as regret, disap-
pointment and self-recrimination, and to strive for positive feelings and emo-
tions such as rejoicing, elation and pride, are assumed to be important
determinants of individual decision-making.
According to Bell (1982) and Loomes and Sugden (1982), the expected
utility of a specific behavioural option x should be modified by incorporating
the amount of regret for not choosing y . More formally:
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Modified SEU, = SEU, f Regret,,t-y

Regret is thus defined as the difference between the value of the obtained
outcome and that of the alternative(s) not chosen.
Summarizing, research on regret attempts to explain violations of the
principles of rationality assumed by SEU-theory by postulating that people
sometimes sacrifice utility in order to prevent the experience of regret. This
approach thus assumes that the anticipation of future regret affects current
choices (Bell, 1981,1982; Loomes & Sugden, l982,1987a), and this assump-
tion has received considerable empirical support (Loomes, 1987; Loomes
and Sugden, 1987a; Ritov, 1996; Simonson, 1992; Zeelenberg et al., 1996),
although there are some exceptions (Harless, 1992; Starmer & Sugden,
1993).
The first work in social psychology paying attention to post-decisional re-
gret is Festinger’s dissonance theory (e.g. Festinger & Walster, 1964). In their
work, regret is almost always conceptualized as the reversal of the initial
decision. Janis and Mann (1977) also pay attention to the role of anticipatory
regret in decision-making. They conceive of anticipatory regret as a “hot”
cognition and argue that:

Before undertaking any enterprise “of great pith and moment”, we usually delay
action and think about what might happen that could cause regret . . . Anricipa-
tory regret is a convenient generic term to refer to the main psychological effects
of the various worries that beset a decision-maker before any losses actually
materialize . . . Such worries, which include anticipatory guilt and shame, pro-
voke hesitation and doubt, making salient the realization that even the most
attractive of the available choices might turn out badly (Janis & Mann, 1977,
p. 222).

Regret is generally regarded as a cognitive-laden or cognitively-determined

emotion. Hampshire (1960) argued that the question of whether one regrets a
decision induces people to think about the decision and not merely to inspect
their feelings. Landman (1993) argues that:

Regret is a more or less painful cognitive and emotional state of feeling sorry for
misfortunes, limitations, losses, transgressions, shortcomings, or mistakes. It is an
experience of felt-reason or reasoned-emotion (p. 36).
 AFFECT, ATTITUDES AND DECISIONS 47

Thus, cognitive evaluations of the outcomes received and the outcomes fore-
gone underlie the comparison of different states of the world that can lead to
the experience of regret.
Why focus on this specific emotion? Research findings show that regret is a
powerful predictor of behaviour and it thus seems a prime candidate to be
incorporated in attitude-behaviour models. For instance, Josephs et al. (1992)
found that the threat of regret reduced the tendency to take risky decisions.
Larrick and Boles (1995) found that the tendency to avoid regret affects
negotiation decisions; Beattie et af. (1994) found that anticipated regret can
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

make people reluctant to make decisions. Bar-Hillel and Neter (1996) argued
that people’s reluctance to exchange lottery tickets could well be a function of
the perceived possibility of postbehavioural regret. Regret can also be related
to Kahneman and Snell’s distinction between decision utility and experienced
utility. Regret is unpleasant and thus affects the experienced utility of the
decision. Next, we address the possible antecedents of anticipated, post-
behavioural regret.

Antecedents of regret
Knowledge about the outcomes of both the chosen and the unchosen op-
tion(s) is central to regret theory: if you cannot compare “what is” with “what
would have been”, there should be no reason for regret. In economic ap-
proaches to regret it is assumed that regret only arises if the outcomes of the
rejected alternatives are revealed. In other words regret will only occur if
people receive feedback about chosen and unchosen alternatives. Zeelenberg
et al. (1996) studied the role of feedback more closely. In accordance with
regret theory they assumed that people are regret-averse and are therefore
motivated to make regref-minimizing choices. In three experiments respond-
ents were given a choice between a risky and a safe gamble. Possible feedback
on one of the options was manipulated orthogonally to the riskiness of the
gambles. Respondents always expected to learn the outcome of the chosen
option; sometimes they could also receive feedback on the foregone outcome.
Those who expected to receive feedback on the safe option, regardless of
their choice, had a preference for this option, thereby protecting themselves
from threatening feedback for the foregone outcome. Likewise, respondents
who expected to receive feedback on the risky option tended to choose the
risky option. When asked to explain their preferences, respondents quite
often referred to the anticipation of regret. Thus, findings of Zeelenberg er al.
(1996) confirm the role of feedback as one of the determinants of regret.
There is evidence showing that the extent to which people feel regret when
confronted with the outcome of a decision does not depend solely upon the
comparison of that decision with the possible outcomes of other courses of
action, but also on how the outcome is achieved. Outcomes achieved through
 48 JOOP V A N DER PLIGT ET AL.

action tend to lead to more extreme regret than the same outcomes achieved
through inaction (Kahneman & Tversky, 1982a; Gleicher et af., 1990; Land-
man, 1987). This can be illustrated by Kahneman and Tversky’s (1982a) study
in which they presented the following scenario:

Mr. Paul owns shares in company A. During the past year he considered switch-
ing to stock in company B, but he decided against it. He now finds out that he
would have been better off by $1200 if he had switched to the stock of company
B. Mr George owned shares in company B. During the past year he switched to
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

stock in company A. He now finds out that he would have been better off by
$1200 if he had kept his stock in company B. Who feels greater regret? (Kahne-
man & Tversky, 1982a).

In Kahneman and Tversky’s study more than 90% of the respondents thought
that Mr. George, whose misfortune stems from an action taken, would experi-
ence more regret. Kahneman and Tversky argue that Mr. George seems more
likely to be plagued by thoughts of what “might have” or “should have” been,
partly because it tends to be easier to imagine oneself abstaining from actions
that one has carried out, than carrying out actions that were not performed.
Landman (1987b) termed this “the actor effect”. Thus, if feedback is not
actually present people are assumed to simulate mentally what could have
been different. This process has been termed “counterfactual thinking”. Un-
fortunately none of the studies dealing with this explanation (e.g. Landman,
1987b; Kahneman & Tversky, 1982a; Gleicher et af., 1990) provide direct
evidence for these processes. The only direct test of the effects of counterfac-
tual thoughts resulted in mixed findings (N’gbala & Branscombe, 1997).
The explanation focusing on the role of counterfactuals differs from the
economic approach to the study of regret. As argued before, economic theor-
ists generally assume that regret does not arise if the outcomes of the rejected
alternatives are not revealed (Bell, 1982, 1983; Kelsey & Schepanski, 1991;
Sage & White, 1983). Thus, as argued by Gilovich and Medvec (1995), no
allowance is made for the fact that individuals might consider (and even be
tormented by) what they imagine to be the outcomes of alternatives not
chosen. Psychological research on counterfactual thinking (Kahneman &
Tversky, 1982b; Kahneman & Miller, 1986) stresses that events are not evalu-
ated in isolation, but are compared to alternative events that “could have”
happened. Research on counterfactual thinking thus has focused on two
issues: (a) the rules by which counterfactual alternatives are generated (i.e.,
some alternatives are more easily imagined than others); and (b) the con-
sequences of comparing actual events with imagined events that might have
happened.
Another possible explanation for the actor-effect focuses on differences in
the perceived responsibility for the outcome, and assumes that outcomes for
which a person is responsible give rise to more extreme affective reactions.
 AFFECT, ATTITUDES AND DECISIONS 49

According to this explanation actions imply greater personal responsibility

than inaction. This can be related to research on attributions indicating that
affective reactions following success and failure are to a large extent deter-
mined by attributions (McFarland & Ross, 1982). Weiner (1986) argues that
affective reactions to outcomes are different and more extreme when the
outcome is attributed to the actor as opposed to situational factors. According
to this reasoning, actions will lead to more extreme affective reactions be-
cause outcomes following action tend to be attributed to the actor. Outcomes
following inaction, however, can be attributed to any external event that
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

preceded the outcome. This line of reasoning is also supported by the fact that
actions, compared to inactions, are more salient, more often used to infer
one’s own attitude, and are perceived to be more informative (see for instance
Fazio, Sherman, & Herr, 1982). Further support is provided by research show-
ing that people who cause harm by acting are judged to be more immoral and
more personally responsible than people who cause the same harm not by
acting (Spranca, Minsk, & Baron, 1991; Ritov & Baron, 1990). Zeelenberg,
van der Pligt, and de Vries (1997a) investigated the role of attributions and
counterfactuals as determinants of the experience of regret. Their results
showed that attributions and affective reactions to outcomes are highly corre-
lated; outcomes after action were associated with both more internal attribu-
tions and more extreme affective reactions. Zeelenberg et al. (1997~)argued
that attributions are likely to mediate the counterfactual-emotion relation.
Research discussed in this section points at a limited number of factors that
can influence the experience of regret. Both counterfactual thinking and at-
tributions seem important determinants of the amount of regret experienced
after misfortune. In the next section we will focus on the consequences of
anticipated regret.

ANTICIPATED REGRET AND THE THEORY OF

PLANNED BEHAVIOUR
As discussed earlier in this chapter, general anticipated affect assessed with
the same scales as evaluations and general affective reactions can improve the
predictive power of the theory of planned behaviour. In this section we will
focus on the role of anticipated regret in the context of other behavioural
determinants incorporated in Ajzen’s theory of planned behaviour. We inves-
tigated this for behaviour characterized by the possibility of a clear discre-
pancy between evaluative reactions towards the behavioural act itself and
postbehavioural feelings; i.e. unsafe vs. safe sex.
Richard, van der Pligt, and de Vries (1995) tested the effects of anticipated
regret on intentions to engage in (un)safe sex. In a first study a total of 822
adolescents participated. Anticipated regret was assessed in accordance with
 50 JOOP V A N DER PLIGT ET AL.

Janis and Mann’s (1977) definition of regret. They use the concept “anticipa-
tory regret” as a generic term for the various worries that beset a decision-
maker before any negative outcomes materialize. Their definition is more
general than the one proposed in regret theory. Attitudes towards condom
use with new and/or casual partners were assessed with a coherent 12-item
scale. Perceived behavioural control was assessed with eight items, and sub-
jective norms were also assessed in accordance with the theory of planned
behaviour. Anticipated regret was assessed with three semantic-differential
scales (regret-no regret; worried-not worried; and tense-relaxed).
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Richard, van der Pligt, and de Vries (1995) tested their model for two be-
havioural actions: “refraining from sexual intercourse” and “condom use”. For
each behaviour respondents were asked to indicate their expectations for three
different situations (e.g. “Suppose you meet a boy/girl you like and both of you
want to make love”; “Suppose you have a date with a person from your
school-after a great evening both of you want to make love”) and these scores
formed a reliable index score. Figure 2.2 summarizes the results of their
LISREL analysis for the behaviour “condom use”. The regression parameters
should be interpreted as P-weights; zero-order correlation coefficients are given
in brackets. Anticipated regret was assessed with the three scales mentioned
earlier and we calculated the difference score between anticipated affect associ-
ated with not having used a condom and anticipated affect associated with
having used a condom. A higher score indicated that higher levels of antici-
pated regret were associated with unsafe sex (not using a condom). Results
show that anticipated regret has an independent and significant impact on
behavioural expectations. For “condom use” the correlation between antici-
pated regret and behavioural expectation was 0.37 and slightly higher than the
correlation between attitudes and behavioural expectations (0.33). While for
“refraining from sexual intercourse” these correlations were 0.48 and 0.51 respec-
tively. The four independent factors explained nearly 40% of the variance in
expectations to refrain from casual sexual intercourse, and nearly 30% of the
variance in expectations to use a condom when having sex with a casual partner.
Richard, van der Pligt, and de Vries (1995) also compared the fit of the
models in which anticipatory regret and attitudes were reflected by a single
latent construct (with five indicators) with the model shown in Figure 2.2. If
anticipated regret and attitudes are essentially equivalent, the overall fit of the
proposed alternative model should not differ significantly from the overall fit
of the two-factor model. This was not so; the possibility that anticipated regret
was similar to attitudes was statistically rejected for both refraining from
sexual intercourse and condom use.
The same model was also tested in a slightly older group of respondents
(Richard, de Vries, & van der Pligt, 1998 in press). A total of 451 students
from the University of Amsterdam participated in this study, in which the
model combining attitudes, subjective norms, perceived behavioural control
 AFFECT, ATTITUDES AND DECISIONS 51

I attitudes

I -,.
I
.50 \
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

I II -.35 1-..pectations
I,,

I I? z: Y,,,
be-havioural

II -.20 /

subjective
I
Figure 2.2 The role of anticipated regret as a determinant of safe sex (condom use).
Note: parameter estimates are standardized; all parameters are significant at p < 0.01
except ap < 0.10. Adapted from Richard, de Vries and van der Pligt (1998, in press)
and anticipated regret explained 65% of the variance in behavioural expecta-
tions (in this case condom use when having sex with a new and/or casual
partner). The correlation between expectation and self-reported behaviour
was 0.58, thus expectations explained 35% of the variance in contraceptive
behaviour of the respondents who had (casual) sex in the 4 weeks following
the first session of the study. The number of respondents who had engaged in
casual sex in this period was considerably lower than the total number of
respondents, hence self-reported behaviour could not be included in the
LISREL analysis summarized in Figure 2.3.
 52 JOOP V A N DER PLIGT ET AL.

yA attitudes I
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Figure 2.3 The role of anticipated regret as a determinant of safe sex. Note: para-
meter estimates are standardized; *n.s, all other parameters are significant at p < 0.01.
Adapted from van der Pligt, de Vries and Richard (1997)

Thus, results of this study show that anticipated regret (in Janis and Mann’s
broader sense) can add to the prediction of behavioural expectations and
(self-reported) behaviour, at least in the context of sexual risk-taking be-
haviour. Parker, Manstead, and Stradling (1995) also tested the role of antici-
pated regret in the context of the theory of planned behaviour. Their study
focused on intentions to commit each of three driving violations (cutting
across traffic to leave a motorway; weaving in and out of two lanes of slow
moving traffic: and overtaking on the inside). Results of their study showed
 AFFECT, ATTITUDES AND DECISIONS 53

substantially improved predictive ability of the model when anticipated regret

was included. In their approach anticipated regret was conceptualized as part
of a moral norm which was measured by items such as, “It would be quite
wrong for me to . . .” (followed by the violation in question) and, “Having
committed a violation would make me feel really sorry for doing it”. On
average their model explained approximately 50% of the variance in inten-
tions to commit driving violations.
To provide further evidence for the causal role of anticipated affective
reactions such as regret and worry, Richard, van der Pligt, and de Vries
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

(1996b) carried out two studies in which they attempted to induce anticipa-
tory, postbehavioural negative affect and test its impact on behavioural expec-
tations and actual behaviour. To this we turn next.

INCREASING THE SALIENCE OF ANTICIPATORY

POSTBEHAVIOURAL AFFECT
Wilson and Hodges (1992) argued that people often have contradictory beliefs
about a topic, and that their attitude towards a topic (which could be an object, a
person or a specific action) depends on the subset of beliefs to which they attend.
This also applies to affective reactions associated with certain behaviours. Quite
often people have mixed emotions about behaviours. For example, one could
have positive feelings about drinking alcohol and smoking cigarettes, but also
negative feelings. The latter most likely stem from the possible more long-term
effects of these behavioural practices. Often, as in this case, there is a temporal
pattern to these different feelings. The feelings associated with performing the
behaviour are positive, but the feelings towards the possible consequences are
negative. This can be related to research on intertemporal choice showing that
people tend to discount more remote future outcomes, and that their decisions
are mostly based on more proximate outcomes (see, e.g., Loewenstein, 1992;
Roelofsma, 19%). Consequently, we expect that feelings about performing a
behaviour will generally receive more weight than the anticipated post-
behavioural feelings when deciding whether to perform the behaviour or not. As
argued by Wilson and Hodges, the beliefs (and feelings) people attend to are
likely to be influenced by both contextual factors and thought processes, and they
present evidence that attitudes and behaviour are easily changed if people are led
to attend to a particular subset of beliefs. Thus, when people think about their
feelings about a specific behavioural action, different beliefs may be salient than
what they think about the feelings they would experience afrer carrying out the
action. The earlier presented evidence that anticipated regret adds to the predic-
tive power of attitude-behaviour models indicates that people differ in the extent
to which they consider postbehavioural affective outcomes. The distinction be-
tween people’s feelings about an action and their anticipated feelings after an
 54 JOOP V A N DER PLIGT ET AL.

action seems most relevant for domains in which there is a clear evaluative and/or
affective discrepancy between the behavioural action itself and the (possible)
postbehavioural outcomes.
Richard, van der Pligt, and de Vries (1996b) investigated whether students’
unsafe sexual practices (i.e., not using condoms with casual partners) would
be reduced by stimulating them to extend their time perspective and think
about their postbehavioural feelings. Condom use is often associated with
reduced sexual pleasure, and for this reason people’s feelings about not using
a condom may be relatively positive. However, an important negative
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

postbehavioural consequence of not using a condom is a possible infection

with a sexually transmitted disease (STD). Thus, if a person failed to use a
condom in a casual sexual interaction, he/she is likely to worry about a poss-
ible STD (including HIV infection) and experience regret and other negative
feelings. Since worry and regret are negative feelings that are more likely to
be experienced afrer risky sexual behaviour, these feelings are likely to be-
come more salient when people think about how they would feel afterwards.
Thus, by stimulating respondents to focus on their feelings after unsafe sex, we
expected them to become increasingly aware of the negative affective con-
sequences of unsafe sexual behaviour. In line with van der Pligt and Richard
(1994), Richard, van der Pligt, and de Vries (1996b) predicted that this in-
creased awareness would make people more risk-averse and would reduce the
likelihood of engaging in risky sexual practices.
In the first study, respondents were asked to imagine the following hypo-
thetical situation: “Suppose you are on a holiday and you meet a very attrac-
tive boy (girl) and after spending some time with him (her) the two of you
have sex”. Respondents were randomly allocated to two groups. Respondents
in the “feelings about” condition were asked to describe the feelings they
would have about having sex (in the above situation) when not using a con-
dom. On the next page they were asked to describe the feelings they would
have about sexual intercourse (in the same situation) when using a condom.
Respondents in the “feelings after” condition were also asked about the
feelings they would have, but now the questions referred to anricipared feel-
ings after having had safe or unsafe sex.
Feelings were assessed with a list of 40 affect-terms, of which 18 were
positive and 22 negative. Examples of these are “elated”, “active” and “ex-
cited” for positive affect terms, and “regretful”, “fearful”, “guilty” and “wor-
ried” for negative affect terms. In each case, respondents described their
feelings by selecting 10 affect terms. Overall, affective reactions to unsafe sex
were much more negative than were affective reactions to safe sex (use of
condoms). More importantly, results also showed a significant main effect of
time perspective on the number of negative feelings mentioned with respect
to the risky behaviour (unsafe sex), with significantly more negative affect
terms being selected in the “feelings after’’ condition.
 AFFECT, ATTITUDES AND DECISIONS 55

Further analyses showed that the increased tendency to endorse negative

affect terms in the “feelings after” condition was largely due to a greater
endorsement of the terms “regret” and “guilt”. For unsafe sex, the “feelings
after” condition resulted in nearly 60% of the respondents selecting the term
“regret”, which was nearly twice as many as in the “feelings about” condition.
The observed greater endorsement of negative terms in the “feelings after”
condition was largely at the expense of positive affect terms (such as “ex-
cited”, “pleasurable” and “energetic”) associated with the act itself. The
increased salience of negative (postbehavioural) feelings with respect to un-
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

safe sex was related to increased expectations to engage in safe, preventive

behaviour; in the “feelings after” condition respondents indicated stronger
expectations to use condoms in future casual sexual interactions than in the
“feelings about” condition.
Richard, van der Pligt, and de Vries (1996b) also tested the effects of the
time-perspective manipulation on actual behaviour in a second study which
included two follow-up measures of condom use in casual encounters over a
period of 5 months. In this study respondents were asked to report their actual
condom use in new or casual sexual encounters after 1 and 5 months. Re-
spondents indicated their feelings on 10 nine-point scales referring to the
following affective states: regretful, enjoyable, contented, worried, tense, pleas-
ant, anxious, satisfied, good, and ill-at-ease. For each of these terms respond-
ents were asked to indicate whether the affect term was more strongly
associated with safe sex (condom use) or with unsafe sex (not using a con-
dom). As expected, respondents in the “feelings after” condition associated
unsafe sex more strongly with negative feelings, and the use of condoms more
strongly with positive feelings. This difference was obtained for all positive
and negative affect terms. Thus, people associated negative feelings (e.g.,
regret, anxiety, worry) more with unsafe sex than with safe sex, but only when
they were asked to describe their postbehavioural feelings. This difference
was not obtained when respondents indicated their feelings about the be-
haviour itself. Again, respondents in the “feelings after” condition had
stronger expectations that they would use condoms because they associated
the use of other contraceptives more with negative affect and the use of
condoms more with positive affect. Moreover, relative to respondents who
focused on the feelings they have about the behavioural act itself, those who
focused on their anticipated feelings after the behavioural alternatives also
reported a more frequent use of condoms in sexual encounters in the
5 months following the experimental manipulation. Thus, a simple mani-
pulation intended to focus respondents on postbehavioural affective con-
sequences of risky behaviour had a clear impact on behavioural intentions
and also an actual behaviour as measured 5 months after the manipulation.
These findings suggest that interventions stressing postbehavioural affect
could be useful in applied settings such as health education.
 56 JOOP V A N DER PLIGT ET AL.

Although additional research seems to be required to illuminate the precise

mechanisms underlying our findings, they suggest that increasing the aware-
ness that an action can have negative postbehavioural affective consequences
is an important factor in producing behavioural change. Simonson (1992) used
an experimental manipulation similar to ours in an attempt to influence con-
sumer behaviour (e.g., a cheap consumer good of an unknown brand and
quality vs. a more expensive alternative from a respectable producer). In his
study respondents were asked to anticipate how they would feel if, after
choosing between two alternatives, the wrong decision were made. The re-
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

maining respondents were not asked how they would feel. Results showed
that respondents in the experimental condition were more risk-averse (i.e.,
made more conventional choices) than were those in the control condition.
Simonson also asked respondents to think aloud as they made their decisions.
Protocol analysis revealed that the main difference between the conditions
was that in the experimental condition respondents tended to base their deci-
sions more on anticipated regret. Thus, Simonson’s findings also show that
asking people to indicate postbehavioural feelings makes them aware of the
affective consequences which are subsequently taken into account when they
make their decision. Our own findings revealed modest but significant and
stable effects due to a simple rating task requiring little effort. It would be
interesting to study ways to enhance the effectiveness of the time perspective
manipulation and test whether more demanding tasks (e.g., writing a short
essay about postbehavioural feelings) increase the size of the effect.

SUMMARY AND CONCLUSIONS

Since the early 1980s there has been a dramatic increase in research on the
impact of affect on attitudinal judgement, attitude-change processes and
decision-making. Unfortunately, the term “affect” is used to denote many
different things. For instance, Zajonc (1980) refers to affect as the outcome of
valenced holistic processes that may proceed automatically without aware-
ness, and are independent from cognitive processes. Others (e.g., Petty et al.,
1993) studied affect by focusing on the impact of mood on information pro-
cessing and attitude change. In the context of expectancy-value approaches to
attitudes and behaviour, the term “affect” is used to refer to rather holistic,
undifferentiated affective reactions to the attitude object. Still others study
affect by looking at specific emotions. In this chapter the emphasis was on
affect as the outcome of a cognitive appraisal. We discussed recent theoretical
and empirical developments concerning the role of affect in attitudinal judge-
ment and decision-making. Several studies indicate that it is possible to dis-
tinguish between affective and more cognitive, analytical determinants of
attitudes and behavioural decisions. We argued that this distinction should be
 AFFECT, AITITUDES AND DECISIONS 57

extended to attitude-behaviour models such as the theory of planned be-

haviour. Most research on the role of affect in attitudes relies on holistic
measures of affect and a rather crude dichotomy of positive vs. negative
affect. Early attempts to incorporate affective determinants of behavioural
intentions in the theory of planned behaviour resulted in only marginal im-
provements of its predictive validity (see, e.g., Ajzen & Driver, 1991). Later
research (see, e.g., Manstead & Parker, 1995) supported the view that more
specific measures of affect-based beliefs can make an independent contribu-
tion to the prediction of behavioural intentions. Their research focused on
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

driving behaviour. The findings of Richard, van der Pligt, and de Vries
(1996a), discussed earlier, point to the independent role of anticipated affect
as a determinant of intention in other behavioural domains.
This evidence thus provides further support for the usefulness of the affect-
cognition distinction in the context of attitude-behaviour models. Generally,
research based on these models emphasizes utilitarian beliefs about the out-
comes of behavioural alternatives. Research using the models of Fishbein and
Ajzen tends to rely on elicitation techniques to assess the modal set of beliefs
that require respondents to list advantages and disadvantages of the target
behaviour. This wording may induce respondents to focus more on the utili-
tarian consequences and pay less attention to the possible affective or emo-
tional consequences of the behaviour. Paying explicit attention to both
components could provide more information about the structure of attitudes
and the relative importance of affect- and belief-based considerations.
Most research on the role of affect in attitudinal judgement, attitude forma-
tion and attitude change relies on a simple dichotomy between positive and
negative affect. We argued that it is essential to be more specific about affec-
tive determinants of attitudes, behavioural intentions and behaviour,
especially when we are dealing with controlled as opposed to automatic as-
pects of information processing. Even if we rely on a simply dichotomy be-
tween positive and negative affect, it seems essential to assess whether we are
dealing with general affective reactions towards the behavioural act itself or
with anticipated, postbehavioural affective reactions. Especially in the context
of attitude-behaviour models, it is not only crucial to distinguish between the
evaluation of the behaviour and affect associated with the behaviour, but also
between the latter and anticipated postbehavioural affective consequences.
Interestingly, most belief-based measures of attitudes focus on specific pos-
sible consequences of the behaviour in question, while studies incorporating
affect tend to rely on rather general affective associations with the be-
havioural activity itself. We would like to argue that an approach paying
attention to both utilitarian and affective consequences of behavioural actions
is to be preferred. Thus, it seems essential to distinguish between the various
possible affective reactions towards behavioural actions, and go beyond the
rather crude distinction between positive and negative affect. Crites, Fabrigar,
 58 JOOP V A N DER PLIGT ET A L .

and Petty (1994) also argue in favour of a differentiated measure of affect and
recommend that affect should be conceptualized in terms of discrete emo-
tional reactions. In this chapter we have paid attention to one specific emo-
tion, i.e., anticipatory regret. Research in the area of decision-making has
shown that this emotion can have profound effects on behavioural choice. The
discriminant validity of measures of anticipatory regret seems adequate, and
the anticipation of this emotion is especially relevant in domains that are
characterized by a discrepancy between the affective responses to the be-
havioural act itself and postbehavioural affective reactions.
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Some issues remain unresolved, however. The studies presented in this chap-
ter do not provide an answer to the question whether affective and cognitive
determinants of attitudes can be really separated and related to different
information-processing systems. The view of separate systems was advanced by
Zajonc (1980, 1984), but his claims for the primacy of affect and the indepen-
dence of cognitive appraisal and affective judgement have proved controversial
(see, e.g., Birnbaum, 1981; Lazarus, 1982, 1984). This controversy partly re-
volves around definitional issues concerning the nature of cognition. Moreover,
Zajonc’s view of affect makes it difficult to distinguish it from cognitive mecha-
nisms that can proceed automatically, without awareness. Buck (1985) sug-
gested that Zajonc’s definition of affect resembles what Tucker (1981) called
“syncretic cognition” (holistic and vague), while his definition of cognition rese-
mbles what Tucker called “analytic cognition”. Although Zajonc acknowledges
that affect and cognition influence each other, he tends to emphasize their
autonomy. The main contribution of Zajonc’s work is that it showed that auto-
matic processes can also influence attitudes. His proposal to distinguish two
separate systems has also influenced attitude research but has generated a
sometimes mysterious discussion and has often relied on rather contrived as-
sessment techniques (e.g., presenting the same rating-scales preceded by ques-
tions about one’s thoughts (cognition) or feelings (affect)).
Assessment techniques requiring respondents to think about their affective
responses tap cognitions about these responses and seem far removed from
the automatic processes Zajonc referred to. We argue that if we focus on
cognitions about affect it seems unwise to treat affect as an undifferentiated,
valenced response and recommended decomposing affect into concrete emo-
tions. In the context of attitude-behaviour models this emphasis should also
include anticipated postbehavioural emotions. One of these is regret, which
seems a powerful predictor of behaviour. This is especially the case when
there is the possibility of alternative actions with different outcomes and when
some of these outcomes are irrevocable. This suggests that it could be useful
for expectancy-value models to also consider alternative courses of action.
This might capture the decision process better (see also Ronis, 1992).
Interestingly, the counterpart of regret (rejoicing) seems to have less
predictive power. Much theorizing and research has focused on (antici-
 AFFECT, ATTITUDES A N D DECISIONS 59

pated) regret and related negative emotions, while positive emotions such as
rejoicing and elation have received far less attention. One reason for this
could be that the impact of (anticipated) positive emotions is far less pro-
nounced than that of negative emotions. It seems that Tversky and Kahne-
man’s view that “losses loom larger than gains” not only applies to
utilitarian beliefs but also to anticipated emotions. In the one study where
we attempted to test for the effects of anticipated elation (Richard, van der
Pligt, & de Vries, 1996a) we found very modest effects: moreover, more
recent attempts (van der Pligt & de Vries, in preparation) show similar
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

(non-)effects of the positive counterpart of regret. This seems in accordance

with emotion research. Attempts to develop a taxonomy of emotions tend to
differentiate between many more negative emotions (each with their own
characteristics and action tendencies) than positive emotions (see for in-
stance Ortony & Turner, 1990).
Are there other negative emotions that should receive attention in the
context of attitude-behaviour models apart from regret? One obvious candi-
date is disappointment. Disappointment is studied in research on behavioural
decision-making (e.g. Bell, 1985; Kelsey & Schepanski, 1991; Loomes, 1987;
van Dijk & van der Pligt, 1997) and the anticipation of disappointment can
affect expectations (Sheppard, Oulette, & Fernandez, 1996) and attitudes
towards different goals and/or tasks. Empirical evidence for the impact of
anticipated disappointment on attitudes and behaviour is limited, and this also
applies to other anticipated emotions that could have an impact on attitudes
such as fear, guilt, embarrassment and shame. On the positive side one could
think of (anticipated) emotions such as pride, relief and joy. As mentioned
before, empirical evidence concerning the possible impact of these anticipated
emotions is extremely limited. Our main argument is that the various possible
anticipated emotions should receive more attention in belief-based measures
of attitudes. If we want to learn more about the role of affective vs. more
cognitive determinants of attitudes and behaviour, both should be assessed
adequately and be based on a similar time perspective. Moreover, both should
be assessed at the same level of specificity. Too often, utilitarian beliefs are
assessed in great detail, while affect is assessed at a more general level, often
with only one or two items.
An added benefit of focusing on specific emotions may be that it helps us to
get rid of the sometimes artificial contrast between affect and cognition. Clore
(1 994) argued that all emotions require cognitions. The anticipation of
postbehavioural affective consequences such as regret requires considerable
cognitive processing. It requires thinking through various courses of action
and estimating the likelihood of regret by comparing a chosen action with
non-chosen actions. The anticipation and experience of regret thus obviously
requires cognition and can hardly be seen as affect in terms of the crude
dichotomy discussed earlier. The same applies to related emotions such as
 60 JOOP VAN DER PLIGT ET A L .

disappointment. which is a function of the difference between expectations

and the obtained outcomes of one’s behaviour.
The findings of the last two studies presented in this chapter show that it is
relatively easy to increase the salience of anticipated regret and related (nega-
tive) affective reactions. A simple procedure requiring respondents to take a
slightly longer time perspective made these emotions more salient, and this
also influenced behavioural intentions and self-reported behaviour as shown
by the results of Richard, van der Pligt, and de Vries (1998, in press). Their
findings were limited to sexual risk-behaviour but there is other evidence
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

suggesting that the awareness that an action can have negative

postbehavioural affective consequences is an important factor in producing
behavioural change. For instance, Simonson (1992) used a similar experimen-
tal manipulation to influence consumer behaviour.
The applied value of this research is that these outcomes point to the
potential usefulness of focusing on more immediate affective consequences of
risky behaviour, such as regret and worry, to foster behavioural change. This
seems especially relevant in the context of health behaviours, because many
negative consequences can be seen as a long way off and are more likely to be
discounted (van der Pligt & de Vries, 1998, in press). Health education cam-
paigns that stress rather immediate postbehavioural affective consequences
could help to increase the willingness to engage in preventive behaviour.
Moreover, stressing postbehavioural affective consequences seems a less ex-
treme manipulation than fear appeals. Research has shown that the amount of
fear aroused by a communication is predictive of intentions to adopt the
recommended action independently of cognitive, more utilitarian consider-
ations. However, it has also been shown that excessive levels of fear can
trigger maladaptive coping-styles and reduce the effectiveness of recom-
mendations to change one’s behaviour (e.g. Liberman & Chaiken, 1992;
Joseph et al., 1987). It could be that stressing postbehavioural affective con-
sequences results in lower anxiety levels and still stimulates preventive
behaviour.
Summarizing, affect seems to have regained its place in research on atti-
tudes and decision-making. Expectancy-value models of attitudes and be-
haviour can benefit from this renewed interest in affect. This would be
enhanced if research in this tradition also focuses on specific affective reac-
tions or emotions and also takes account of their time-perspective. Holistic
assessments of affective associations with behavioural acts are less likely to
help improve our understanding of the role of affect in controlled information
processing underlying attitudes and decisions. More correspondence of mea-
surement between outcome-related beliefs and expected affects, as well as a
differentiated approach to the measurement of affect, should help to increase
the expected utility of expectancy-value approaches to attitudes and
behaviour.
 AFFECT, AlTITUDES AND DECISIONS 61

ACKNOWLEDGEMENTS
We would like to thank Tony Manstead, Dave Messick, Miles Hewstone,
Wolfgang Stroebe, Herbert Bless and Stephen Sutton for their comments on
an earlier draft of this paper. Marcel Zeelenberg is currently at Tilburg
University.

REFERENCES
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Abelson, R. P., Kinder, D. R., Peters, M. D., & Fiske, S. T. (1982). Affective and
semantic components in political person perception. Journal of Personality and
Social Psychology, 42,619-30.
Ajzen, I. (1985). From intentions to actions: A theory of planned behavior. In
J. Kuhl & J. Beckmann (Eds), Action-control: From Cognition to Behavior (pp. 11-
39). Heidelberg: Springer.
Ajzen, I. (1989). Attitude structure and behavior. In A. R. Pratkanis, S. J. Breckler &
A. G. Greenwald (Eds), Attitude Structure and Function. Hillsdale, NJ: Erlbaum.
Ajzen, I. (1991). The theory of planned behavior. Organizational Behavior and HU-
man Decision Processes, 50, 179-21 1.
Ajzen, I., & Driver, B. E. (1991). Prediction of leisure participation of leisure parti-
cipation from behavioral, normative, and control beliefs: An implication of theory of
planned behavior. Leisure Sciences, 13, 185-204.
Ajzen, I., & Timko, C. (1986). Correspondence between health attitudes and be-
havior. Journal of Basic and Applied Social Psychology, 7,259-76.
Bargh, J. A. (1994). The four horsemen of atomaticity: Awareness, intention, effici-
ency, and control in social cognition. In R. S. Wyer, J r & T. K. Srull (Eds), Hand-
book of Social Cognition (Vol. 1): Basic Processes (pp. 3-40). Hillsdale, NJ:
Erlbaum.
Bargh, J. A., Chaiken, S., Raymond, P., & Hymes, C. (1996). The automatic evalua-
tion effect: Unconditional automatic attitude activation with a pronunciation task.
Journal of Experimental Social Psychology, 32,104-28.
Bar-Hillel, M., & Neter, E. (1996). Why are people reluctant to exchange lottery
tickets? Journal of Personality and Social Psychology, 70,17-27.
Baron, J. (1992). The effect of normative beliefs on anticipated emotions. Journal of
Personality and Social Psychology, 63, 320-30.
Beattie, J., Baron, J., Hershey, J. C., & Spranca, M. D. (1994). Psychological determi-
nants of decision attitude. Journal of Behavioral Decision Making, 7,129-44.
Bell, D. E. (1981). Explaining utility theory paradoxes by decision regret. In J. Morse
(Ed.), Proceedings of the Fourth International Conference on Multiple Criteria Deci-
sion Making (pp. 28-39). New York: Springer.
Bell, D. E. (1982). Regret in decision making under uncertainty. Operations Re-
search, 21,961-81.
Bell, D. E. (1983). Risk premiums for decision regret. Management Science, 29,1156
66.
Bell, D. E. (1985). Disappointment in decision making under uncertainty. Operations
Research, 33,l-27.
Birnbaum, M. H. (1981). Thinking and feeling: A skeptical review. American Psycho-
logist, 36, 99-101.
 62 JOOP VAN DER PLIGT E T A L .

Bless, H., Bohner, G., Schwarz, N.. & Strack, F. (1990). Mood and persuasion: A
cognitive response analysis. Personalitji and Social Psychology Bulletin. 16. 33145.
Breckler, S. J. (1984). Empirical validation of affect, behavior and cognition as dis-
tinct components of attitude. Journal of Personality and Social Psychology, 47.1191-
1205.
Breckler, S. J., & Wiggins, E. C. (1989). Affect versus evaluation in the structure of
attitudes. Journal of Experimental Social Psychology, 25, 253-71.
Buck, R. (1985). Prime theory: an integrated view of motivation and emotion. Psy-
chological Review, 92,389-413.
Chan, D. K-S., & Fishbein, M. (1993). Determinants of college women’s intentions to
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

tell their partners to use condoms. Journal of Applied Social Psychology, 23, 1455-
70.
Clore, G . L. (1994). Why emotions require cognition. In P. Ekman & R. J. Davidson
(Eds), The Nature of Emotions: Fundamental Questions (pp. 181-91). New York:
Oxford University Press.
Cohen, L. J. (1981). Can human irrationality be experimentally demonstrated? Be-
havioral and Brain Sciences, 4, 317-31.
Crites, S. L. Jr, Fabrigar, L. R.. & Petty, R. E. (1994). Measuring the affective and
cognitive properties of attitudes: Conceptual and methodological issues. Personality
and Social Psychology Bulletin, 20, 619-34.
Eagly, A. H., Mladinic. A., & Otto, S. (1994). Cognitive and affective bases of atti-
tudes towards social groups and social policies. Journal of Experimental Social Psy-
chology, 30.113-37.
Edwards. W. (1954). The theory of decision making. Psychological Bulletin, 51, 380-
417.
Edwards, K. (1990). The interplay of affect and cognition in attitude formation and
change. Journal of Personality and Social Psychology, 59, 202-16.
Fazio, R. H., Sherman, S. J., & Herr. P. M. (1982). The feature-positive effect in the
self-perception process: does not doing matter as much as doing? Journal of Person-
ality and Social Psychology, 42,404-1 1.
Festinger, L.. & Walster, E. (1964). Postdecision regret and decision reversal. In
L. Festinger (Ed.). Conflict, Decision, and Dissonance (pp. 100-1 12). Stanford. CA:
Stanford University Press.
Fishbein, M.. & Ajzen. I. (1975). Beliefs, Attitudes, Intention, and Behavior: A n Intro-
duction to Theory and Research. Reading, MA: Addison-Wesley.
Fisher, W. A. (1984). Predicting contraceptive behaviour among university men: The
role of emotions and behavioral intentions. Journal of Applied Social Psychology,
14,10423.
Frijda, N. H., Kuipers. P.. & ter Schure, E. (1989). Relations among emotion,
appraisal and emotional action readiness. Journal of Personality and Social Psycho-
logy, 57,212-28.
Gilovich, T., & Medvec, V. (1995). The experience of regret: What, when and why.
Psychological Review, 2,379-95.
Gleicher, F., Kost, K. A., Baker, S. M., Strathman, A. J., Richman, S. A.. & Sherman,
S. J. (1990). The role of counterfactual thinking in judgments of affect. Personnlity
and Social Psychology Bulletin, 16,284-95.
Hampshire, S. (1960). Thought and Action. London: Chatto and Windus.
Harless, D. W. (1992). Actions versus prospects: The effect of problem presentation
on regret. American Economic Review, 82,634-49.
 AFFECT, AITITUDES AND DECISIONS 63
Janis, I. L. (1967). Effects of fear arousal on attitude change: Recent developments in
theory and experimental research. In L. Berkowitz (Ed.), Advances in Experimental
Social Psychology (Vol. 3, pp. 166-224). San Diego, CA: Academic Press.
Janis, I. L., & Mann, L. (1977). Decision Making: A Psychological Analysis of Con-
flict, Choice, and Commitment. New York: Free Press.
Joseph, J. G., Montgomery, S. B., Emmons, C. A., Kirscht, J. P., & Kessler, R. C., et al.
(1987). Perceived risk of AIDS: Assessing the behavioral and psychological con-
sequences in a cohort of gay men. Journal of Applied Social Psychology, 17,231-50.
Josephs, R. A., Larrick, R. P., Steele, C. M, & Nisbett, R. E. (1992). Protecting the
self from the negative consequences of risky decisions. Journal of Personality and
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Social Psychology, 62,2637.

Kahneman, D., & Miller, D. T. (1986). Norm theory: Comparing reality to its alterna-
tives. Psychological Review, 93, 136-53.
Kahneman, D., & Snell, J. (1990). Predicting utility. In Hogarth, R. M. (Ed.), Insights
in Decision Making (pp. 295-310). Chicago: The University of Chicago Press.
Kahneman, D., & Snell, J. (1992). Predicting a changing taste: Do people know what
they will like? Journal of Behavioral Decision Making, 5, 187-200.
Kahneman, D., & Tversky, A. (1979). Prospect theory: Analysis of decision under
risk. Econometrica, 47, 263-91.
Kahneman, D., & Tversky, A. (1982a). The psychology of preferences. Scienf$c
American, 246(1), 160-73.
Kahneman, D., & Tversky, A. (1982b). The simulation heuristic. In D. Kahneman,
P. Slovic & A. Tversky (Eds), Judgement Under Uncertainty: Heuristics and Biases
(pp. 201-8). Cambridge: Cambridge University Press.
Kelsey, D., & Schepanski, A. (1991). Regret and disappointment in taxpayer report-
ing decisions: an experimental study. Journal of Behavioral Decision Making, 4 3 3 -
53.
Landman, J. (1987). Regret and elation following action and inaction: Affective re-
sponses to positive versus negative outcomes. Personality and Social Psychology
Bulletin, 13,524-36.
Landman, J. (1993). Regret: The Persistence of the Possible. New York: Oxford Uni-
versity Press.
Larrick, R. P., & Boles, T. L. (1995). Avoiding regret in decisions with feedback: a
negotiation example. Organizarional Behavior and Human Decision Processes, 63,
87-97.
Lazarus, R. S. (1981). A cognitive reply to Zajonc on emotion and cognition. Ameri-
can Psychologist, 36,222-3.
Lazarus, R. S . (1982). Thoughts on the relations between emotion and cognition.
American Psychologist, 37, 1019-24.
Lazarus, R. S. (1984). On the primacy of cognition. American Psychologist, 39,124-9.
Leventhal, H. (1970). Findings and theory in the study of fear communications. In
L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 5 , pp. 119-
86). San Diego, CA: Academic Press.
Liberman, A., & Chaiken, S. (1992). Defensive processing of personally relevant
health messages. Personality and Social Psychology Bulletin, 18,669-79.
Loewenstein, G. (1992). The fall and rise of psychological explanations in the econ-
omics of intertemporal choice. In G. Loewenstein & J. Elster (Eds), Choice Over
Time (pp. 3-34). New York: Russel Sage Foundation.
Loewenstein, G. F., Thompson, L, & Bazerman, M. H. (1989). Social utility and
decision making in interpersonal contexts. Journal of Personality and Social Psy-
chology, 57,42641.
 64 JOOP VAN DER PLIGT ET A L.

h o m e s , G. (1987). Further evidence of the impact of regret and disappointment in

choice and uncertainty. Economica, 55.47-62.
Loomes, G., & Sugden, R. (1982). Regret theory: an alternative theory of rational
choice under uncertainty. Economic Journal, 92,805-25.
h o m e s , G., & Sugden, R. (1986). Disappointment and dynamic consistency in
choice under certainty. Review of Economic Studies, 53,271-82.
Loomes, G., & Sugden, R. (1987a). Some implications of a more general form of
regret theory. Journal of Economic Theory, 41,270-87.
Loomes, G., & Sugden, R. (1987b). Testing for regret and disappointment in choice
under uncertainty. Economic Journal, 97,118-29.
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Madden, T. J., Ellen, P. S.. & Ajzen, I. (1992). A comparison of the theory of planned
behavior and the theory of reasoned action. Personality and Social Psychology
Bulletin, 18,3-9.
Manstead, A. S. R., & Parker, D. (1995). Evaluating and extending the theory of
planned behaviour. In W. Stroebe & M. Hewstone (Eds), European Review of Social
Psychology (Vol. 6, pp. 69-95). Chichester: Wiley.
McCaul, K. D., Sandgren, A. K., O’Neill, H. K., & Hinsz, V. B. (1993). The value of
the theory- of planned
- behavior, perceived control, and self-efficacy expectations for
predicting health-protective behaviors. Basic and Applied Social Psychology, 14,
231-52.
McFarland, C.. & Ross, M. (1982). Impact of causal attribution on affective reactions
to success and failure. Journal of Personality and Social Psychology, 43.93746.
Middlestadt, S. E., & Fishbein, M. (1995). Evaluating the impact of a national AIDS
prevention radio campaign in St. Vincent and the Grenadines. Journal of Applied
Social Psychology, 25,21-34.
Millar, M. G., & Millar, K. U. (1990). Attitude change as a function of attitude type
and argument type. Journal of Personality and Social Psychology, 59,217-28.
Murphy, S. T., & Zajonc, R. B. (1993). Affect, cognition, and awareness: Affective
priming with optimal and suboptimal stimulus exposures. Journal of Personality and
Social Psychology, 64,723-39.
N’gbala, A., & Branscombe, N. B. (1997). When does action elicit more regret than
inaction and is counterfactual mutation the mediator of this effect? Journal of
Experimental Social Psychology, 33,324-43.
Ortony, A., & Turner, T. J. (1990). What’s basic about basic emotions. Psychological
Review, 97,313-31.
Parker, D., Manstead, A. S. R., & Stradling, S. G . (1995). Extending the theory of
planned behaviour: The role of personal norm. British Journal of Social Psychology,
34,127-37.
Parker, D., Manstead, A. S. R., Stradling, S. G., Reason, J. T., & Baxter, J. S. (1992).
Intention to commit driving violations: An application of the theory of planned
behavior. Journal of Applied Psychology, 77,94101.
Petty, R. E., & Cacioppo, J. T. (1986). The elaboration likelihood model of persua-
sion. In L. Berkowitz (Ed.), Advances in Experimental Social Psychology (Vol. 19).
Orlando. FL: Academic Press.
Petty, R. E., Schumann, D. W., Richman, S. A., & Strathman, A. J. (1993). Positive
mood and persuasion: Different roles for affect under high- and low-elaboration
conditions. Journal of Personality and Social Psychology, 6 3 , 5 2 0 .
Pfister. H.. & Bohm, G. (1992). The function of concrete emotions in rational deci-
sion making. Acta Psychologica, 80, 105-16.
Richard, R., van der Pligt, J., & de Vries, N. K. (1995). The impact of anticipated
affect on (risky) sexual behavior. British Journal of Social Psychology, 34, 9-21.
 AFFECT, ATTITUDES AND DECISIONS 65

Richard, R., van der Pligt, J., & de Vries, N. K. (1996a). Anticipated affect and
behavioral choice. Basic and Applied Social Psychology, 18, 11 1-29.
Richard, R., van der Pligt, J., & de Vries, N. K. (1996b). Anticipated regret and time
perspective: Changing sexual risk-taking behavior. Journal of Behavioral Decision
Making, 9, 185-99.
Richard, R., de Vries, N.K., & van der Pligt, J. (1998). Anticipated regret and precau-
tionary sexual behaviour. Journal of Applied Social Psychology, in press.
Ritov, I. (1996). Probability of regret: anticipation of uncertainty resolution in choice.
Organizational Behavior and Human Decision Making Processes, 66,228-36.
Ritov, I., & Baron, J. (1990). Reluctance to vaccinate: Omission bias and ambiguity.
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Journal of Behavioral Decision Making, 3,263-77.

Roelofsma, P. H. M. P. (1996). Modelling intertemporal choice: An anomaly ap-
proach. Acta Psychbologica (in press).
Rogers, R. W. (1975). A protection motivation theory of fear appeals and attitude
change. Journal of Psychology, 91,93-114.
Ronis, D. L. (1992). Conditional health threats: Health beliefs, decisions, and be-
haviors among adults. Health Psychology, ll, 127-34.
Roseman, I. J., Wiest, C., & Swartz, T. S. (1994). Phenomenology, behaviors, and
goals differentiate discrete emotions. Journal of Personality and Social Psychology,
67,206-1 1.
Rosenberg, M. J., & Hovland, C. I. (1960). Cognitive, affective and behavioral com-
ponents of attitude. In M. J. Rosenberg, C. I. Hovland, W. J. McGuire, R. P.
Abelson, & J. W. Brehm (Eds), Attitude Organization and Change. New Haven, CT:
Yale University Press.
Sage, A. P., & White, F. B. (1983). Decision and information structures in regret:
Models of judgment and choice. IEEE: Transitions on Systems, Man, and Cyber-
netics, 13, 136-43.
Schifter, D. B., & Ajzen, I. (1985). Intention, perceived control, and weight loss: An
application of the theory of planned behavior. Journal of Personality and Social
Psychology, 49. 843-51.
Seligman, C., Kriss, M., Darley, J. M., Fazio, R. H., Becker, L. J., & Pryor, J. B.
(1979). Predicting summer energy consumption from homeowners’ attitudes. Jour-
nal of Applied Social Psychology, 9,7&90.
Sheppard, J. A., Ouelette, J. A., & Fernandez, J. K. (1996). Abandoning unrealistic
optimism: Performance estimates and temporal proximity of self-relevant feedback.
Journal of Personality and Social Psychology, 70, 844-55.
Simonson, I. (1992). The influence of anticipating regret and responsibility on pur-
chase decisions. Journal of Consumer Research, 19,105-18.
Sparks, P., Hedderley, D., & Shepherd, R. (1992). An investigation into the relation-
ship between perceived control, attitude variability and the consumption of two
common foods. European Journal of Social Psychology, 22,55-71.
Spranca, M., Minsk, E., & Baron, J. (1991). Omission and commission in judgment
and choice. Journal of Experimental Social Psychology, 27,76105.
Starmer, C., & Sugden, R. (1993). Testing for juxtaposition and event splitting
effects. Journal of Risk and Uncertainty, 6,235-54.
Terry, D. J., Gallois. C., & McCamish, M. (Eds), (1993). The Theory of Reasoned
Action: Its Application to AIDS-preventive Behaviour. Oxford: Pergamon.
Tucker, D. M. (1981). Lateral train function, emotion, and conceptualizaiton. Psy-
chological Bulletin, 89, 1 9 4 6 .
Tyler, T. R., & Rasinsky, K. (1984). Comparing psychological images of the social
perceiver: Role of perceived informativeness, memorability, and affect in mediating
the impact of crime victimization. Journal of Personality and Social Psychology, 46,
30S29.
 66 JOOP V A N DER PLIGT ET A L.

van der Pligt, J., & de Vries, N. K. (1998). Expectancy-value models of health be-
haviour: The role of salience and effect. Psychology and Health, in press.
van der Pligt, J. & de Vries, N. K (1997). Anticipated regret and elation as determi-
nants of behavioural intentions (manuscript in preparation).
van der Pligt, J., & Richard, R. (1994). Changing adolescents’ sexual behaviour:
Perceived risk, self-efficacy, and anticipated regret. Patient Education and Counsel-
ing, 23, 187-96.
van Dijk, W. W., & van der Pligt, J. (1997). The impact of probability and magnitude
of outcome on disappointment and elation. Organizational Behaviour and Human
Decision Processes, 69,277-84.
Downloaded by [University of Nebraska, Lincoln] at 09:54 04 January 2015

Weiner, B. (1986). Attribution, emotion and action. In R. M. Sorrentino & E. T.

Higgins (Eds), Handbook of Motivation and Cognition (pp. 281-312). Chichester:
Wiley.
Wilson, T. D., & Hodges, S. D. (1992). Attitudes as temporary constructions. In L. L.
Martin & A. Tesser (Eds), The Construction of Social Judgments (pp. 37-65). Hills-
dale, NJ: Erlbaum.
Zajonc, R. B. (1980). Feeling and thinking: preferences need no inferences. Ameri-
can Psychologist, 35,151-75.
Zajonc, R. B. (1984). On the primacy of affect. American Psychologist, 39,117-23.
Zanna, M.,& Rernpel, J. K. (1988). Attitudes: A new look at an old concept. In
D. Bar-Tal & A. Kruglanski (Eds), The Social Psychology of Knowledge (pp. 315-
34). Cambridge: Cambridge University Press.
Zeelenberg, M., Beattie, J., van der Pligt, J., & de Vries, N. K. (1996). Consequences
of regret aversion: Effects of expected feedback on risky decision making. Organiza-
tional Behavior and Human Decision Processes (in press).
Zeelenberg, M., van der Pligt, J., & de Vries, N. K. (1997a). The role of attributions
in post-decisional affect (manuscript under review).
Zeelenberg, M., van Dijk, W. W., Manstead, A. S. R., & van der Pligt, J. (1997b). The
experience of regret and disappointment. Cognition and Emotion (in press).
Zeelenberg, M., van Dijk, W. W., van der Pligt, J., Manstead, A. S. R., van Empelen,
P., & Reinderman, D. (1997~). The role of behavior-focused and situation-focused
counterfactuals in the experience of regret and disappointment (manuscript under
review).