Content uploaded by Scott Rick
Author content
All content in this area was uploaded by Scott Rick on Apr 20, 2018
Content may be subject to copyright.
Research Report
The benefits of retail therapy: Making purchase decisions
reduces residual sadness
Scott I. Rick ⁎, Beatriz Pereira, Katherine A. Burson
Ross School of Business, University of Michigan, 701 Tappan Street, Ann Arbor, MI 48109, USA
Received 14 August 2012; received in revised form 10 December 2013; accepted 22 December 2013
Available online 27 December 2013
Abstract
People often shop when feeling sad, but whether and why shopping reduces residual (lingering) sadness remains an open question. Sadness is
strongly associated with a sense that situational forces control the outcomes in one's life, and thus we theorized that the choices inherent in shopping
may restore personal control over one's environment and reduce residual sadness. Three experiments provided support for our hypothesis. Making
shopping choices helped to alleviate sadness whether they were hypothetical (Experiment 1) or real (Experiment 2). In addition, all experiments found
support for the underlying mechanism of personal control restoration. Notably, the benefits of restored personal control over one's environment do not
generalize to anger (Experiments 2 and 3), because anger is associated with a sense that other people (rather than situational forces) are likely to cause
negative outcomes, and these appraisals are not ameliorated by restoring personal control over one's environment.
© 2013 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.
Keywords: Retail therapy; Shopping; Appraisal tendency theory; Sadness; Decision-making
Introduction
How do people regulate distress? Several common responses
to distress have been documented, such as rumination, overeat-
ing, and alcohol consumption. Distress can also encourage
unplanned purchases (e.g., Atalay & Meloy, 2011, Study 1).
Shopping that is motivated by distress—“retail therapy”—is
often lamented as ineffective, wasteful, and a “dark side”of
consumer behavior (Kasser & Sheldon, 2000). Popular press
accounts of retail therapy typically paint an equally dismal picture
(Tuttle, 2010).
We propose that retail therapy has been viewed too negatively.
Shopping may be an effective way to minimize sadness that
lingers (residual sadness) following a sadness-inducing event.
We focus on shopping's potential to reduce residual sadness in
particular, as previous research has demonstrated that sadness
increases comfort-seeking (Raghunathan, Pham, & Corfman,
2006) and willingness-to-pay (Cryder, Lerner, Gross, & Dahl,
2008; Lerner, Small, & Loewenstein, 2004).
Prior research has provided some suggestive evidence that
shopping can convey psychological benefits (Gardner & Rook,
1988). In a diary study, Atalay and Meloy (2011, Study 3)
found that most participants reported positive feelings when
reflecting on their most recent purchase that was motivated by a
desire to repair mood. Faber and Christenson (1996, Table 3)
found that people recalled that they were less likely to
experience sadness while shopping than immediately before
going shopping.
However, causal conclusions remain elusive, as no prior
research investigating the influence of shopping on emotion
or mood has utilized experimental designs. Without random
assignment to shopping or equally engaging “control”
activities, it is unclear whether shopping conveys benefits
beyond those produced merely by distraction or the passage
of time.
In addition, research in this area has only loosely conceptu-
alized both affect and shopping. Atalay and Meloy (2011) utilized
broad measures of mood (p. 642) and positive emotion and
negative emotion indices (p. 653), rather than investigating the
experience of specific emotions. Faber and Christenson (1996,
p. 809) asked participants to report how they generally feel
“while shopping,”without referencing any specific shopping
⁎Corresponding author.
E-mail addresses: srick@umich.edu (S.I. Rick), bpereira@umich.edu
(B. Pereira), kburson@umich.edu (K.A. Burson).
1057-7408/$ -see front matter © 2013 Society for Consumer Psychology. Published by Elsevier Inc. All rights reserved.
http://dx.doi.org/10.1016/j.jcps.2013.12.004
Available online at www.sciencedirect.com
ScienceDirect
Journal of Consumer Psychology 24, 3 (2014) 373 –380
episode. Because “shopping”can have many components, in-
cluding browsing, interacting with salespeople, choosing, paying,
acquiring, and consuming, retrospective reports about “shopping”
cannot shed light on which component(s) are necessary for
healing to occur.
This last point is not merely a descriptive shortcoming.
Differences in the effectiveness of specific components could
shed light on why shopping reduces residual sadness. To develop
hypotheses about why some components will be particularly
influential, we consider sadness from an appraisal tendency theory
perspective (Han, Lerner, & Keltner, 2007). Appraisal theory
suggests that the way people cognitively appraise their environ-
ment is both a cause and consequence of different emotions.
Smith and Ellsworth (1985) identified six appraisals that
differentiate emotions: the extent to which the current situation
is pleasant, predictable, demanding of attention, demanding of
effort, under human (versus situational) control, and under one's
own or other people's control. Thus, similarly valenced emotions
can differ on other important dimensions (e.g., anger and fear are
both aversive, but anger is associated with greater certainty;
Lerner & Keltner, 2001).
Sadness, more than any other emotion, is associated with a
perceived deficiency in personal control over one's environment
(Smith & Ellsworth, 1985). People who are sad are especially
likely to view outcomes as governed by situational forces and
chance, rather than their own actions. To the extent that these
appraisals create or maintain the experience of sadness (Han et al.,
2007), aspects of shopping that restore a sense of personal control
over one's environment may subsequently reduce residual
sadness. Indeed, Garg and Lerner (2013, p. 112) proposed that
researchers should investigate whether “feeling less helpless
correspond[s] with feeling less sad.”
Prior research suggests that the ability to choose tends to
enhance one's sense of personal control (Inesi, Botti, Dubois,
Rucker, & Galinsky, 2011; Langer, 1975). Because choices are
inherent to shopping (e.g., choosing whether to buy), shopping
may restore a sense of control and thus minimize residual
sadness.
Of course, aside from choice, other aspects of the shopping
experience could influence sadness. For example, shopping
may provide distraction (cf. Kim & Rucker, 2012) or social
interaction (O'Guinn & Faber, 1989). In what follows, we
experimentally isolate the influence of choice on the experience
of sadness by utilizing simplified paradigms that necessarily
strip away extraneous factors that can accompany naturalistic
shopping. For example, there is no consumption or social
interaction in our experiments. We control for the benefits of
distraction in Experiment 1 by including a “browsing”control
condition, in which participants must interact with products but
cannot buy any.
We focus on choice for two reasons. First, choice is the
component of shopping that is most theoretically linked to
personal control. Given that sadness is characterized by a lack
of personal control over one's environment, the control imbued
by making shopping choices may help reduce residual sadness.
Second, we focus on choice because it is arguably the most
fundamental component of shopping. While shopping may or
may not involve factors not present in our experiments (e.g.,
social interaction), shopping always involves choice.
We propose thatmaking shopping choices can help to restore a
sense of personal control over one's environment, but many
people may have difficulty quantifying and articulating the extent
to which they feel control over their environment. (In their classic
demonstration, SmithandEllsworth(1985,p.820)utilized a
group of participants pre-screened to be highly emotionally
expressive, and asked them to recall their experiences of control
during a specific emotional event, rather than their current,
ambient feelings of control over their environment.) Thus, inwhat
follows we shed light on our proposed process by experimentally
manipulating personal control (cf. Spencer, Zanna, & Fong,
2005). We do so by manipulating whether participants can freely
choose among a broad product assortment (Experiment 1),
whether participants believe they can ensure that they obtain
their preferred product (Experiment 2), and whether participants
recall an instance of high or low control over their environment
(Experiment 3).
Experiment 1
Experiment 1 tested the hypothesis that making shopping
choices helps to restore personal control over one's environ-
ment, which can in turn help to alleviate residual sadness. We
randomly assigned participants to choose which of several
products they would hypothetically buy, or to judge which of
those products would be most useful when traveling. Concep-
tually, our intention was to manipulate the extent to which
participants could exercise personal autonomy during the task
(since only a handful of the products are appropriate for travel,
but any could be selected by hypothetical buyers), while
holding constant distraction and (lack of) product acquisition
across conditions.
Method
One hundred adults (52% female, mean age: 36) participated
in an online study for a small payment. We recruited participants
via Amazon Mechanical Turk (MTurk), a platform validated by
Paolacci, Chandler, and Ipeirotis (2010). We initially collected a
baseline measure of emotions. Specifically, participants indicated
the extent to which they were currently experiencing seven
different emotions (amused,sad,indifferent,angry,depressed,
happy,rage), by moving a slider along a 12 mm line anchored by
the labels “not at all”and “very much.”Responses were scored on
a0–100 scale based on where participants rested the slider.
Participants then viewed a three-minute clip from The Champ
portraying the death of a boy's father, which reliably induces
sadness (Rottenberg,Ray,&Gross,2007). We then took a
second measure of emotions, identical to our baseline measure.
We then randomly assigned participants to a Choosing or
Browsing condition, adapting a design by Mazar and Zhong
(2010). Choosers were told to “imagine buying $100 worth of
products, by placing them in a shopping cart.”Choosers then
viewed 12 products (e.g., slippers, headphones; see Fig. 1),
each priced at $25. Choosers were asked to select four products
374 S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
they would hypothetically like to buy, by clicking on four
products and dragging them into a box labeled “Your Shopping
Cart.”Choosers were informed that the shopping was
hypothetical, and they had no expectation of obtaining these
items.
Browsers viewed the same 12 products and were asked to
judge which four products would be most useful when traveling,
by clicking on four products and dragging them into a box labeled
“Travel Items.”There was no significant difference in the amount
of time spent on the Choosing and Browsing tasks (60.7 s vs.
58.2 s; t(98) b1).
We then administered a third (and final) measure of
emotions, identical to our baseline measure.
We conducted a pre-test on MTurk (N = 42, 43% female,
mean age: 33) to confirm that the Choosing task was more likely
to generate feelings of control than the Browsing task.
Participants completed both the Choosing and Browsing tasks
(which were labeled the Shopping Cart task and the Travel Items
task, respectively, for participants). The order of tasks was
counterbalanced across participants. We then asked, “In which
task did you feel you had more control over the items you
selected?”Participants selected one of five options: Definitely the
Travel Items task, Probably the Travel Items task, No difference
between the Shopping Cart and Travel Items tasks, Probably the
Shopping Cart task, and Definitely the Shopping Cart task.
Seventy-nine percent of participants reported that they (probably
or definitely) felt more control while Choosing, whereas only 2%
reported that they (probably or definitely) felt more control while
Browsing (χ
2
(1) = 50.6, pb.0001). Thus, while both tasks
likely generate, to some extent, a sense of personal control over
one's environment, the Choosing task is a more effective way to
increase a sense of personal control.
Results
To verify that our manipulation worked as intended, we
examined whether the selection of products reflected greater
autonomy among Choosers. Free choice among options imbues
a sense of control and autonomy by allowing people to
implement or reveal their individual preferences. Browsers
Fig. 1. Products utilized in Experiment 1.
375S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
were asked to select the products most appropriate for travel,
but only a handful of the products were appropriate for travel.
By contrast, Choosers were free to select any product.
Choosers' greater ability to control which products they select
should be reflected by greater variance in their selections.
Fig. 2 displays the proportion of participants selecting each
product by condition. There was significantly greater variance
in product selection among Choosers than Browsers (F(1,22) =
4.47, pb.05, Levene's test). If Browsers could freely choose
to the same extent as Choosers, there would be a similar degree
of variance in selected products across conditions. The unequal
variance in product selections across conditions is consistent
with the notion that Choosers experienced more control during the
task.
We next tested our central hypothesis that choosing would be
more likely than browsing to alleviate residual sadness. Sad and
depressed ratings correlated highly (r(98) N.65, pb.0001 at
each measurement), and were averaged to form a sadness index at
each measurement. We created a residual sadness score by
subtracting participants' baseline index scores from their final
index scores. This commonly-used method controls for broad
individual differences in the tendency to experience and express
emotions (cf. Kermer, Driver-Linn, Wilson, & Gilbert, 2006;
Oveis et al., 2009; Rogosa & Willett, 1983; Wilson, Wheatley,
Meyers, Gilbert, & Axsom, 2000). Three outliers, with residual
sadness scores more than three standard deviations from the
mean, were excluded.
As expected, residual sadness scores were significantly lower
among Choosers than Browsers (M=2.9,SD = 8.6 vs. M=8.1,
SD =14.5; t(95) = 2.13, pb.05). In other words, making
hypothetical buying choices was more likely to return partici-
pants to their baseline level of sadness than was browsing.
Thus, Experiment 1 provides initial support for the hypothesis
that making shopping choices helps to alleviate residual sadness.
The results of the pre-test and the greater variance in product
selections among Choosers suggest that the mechanism under-
lying this effect is a restoration of personal control.
Experiment 2
Experiment 2 extends our investigation in three ways. First,
all participants made real (consequential) shopping decisions.
Second, to isolate the role of restored control in retail therapy, all
participants actually obtained their preferred option, but we
manipulated the extent to which they had apparent control over
the process of obtaining that option. Third, to shed additional
light on the process of personal control restoration, we
examined whether the benefits of making shopping choices
are specific to sadness or generalize to other negative emotions.
In particular, we examined whether making shopping choices
also helps to alleviate anger. Anger is generally as aversive as
sadness, but is associated with a sense that other people (rather
than situational forces) cause negative outcomes (Smith &
Ellsworth, 1985). Whereas making shopping choices may help
restore a sense of personal control over one's environment,
those choices are unlikely to reduce the extent to which other
people are viewed as unfairly or unduly influential. Thus,
enhancing control over one's environment should have more
impact on sadness than on anger.
Method
One hundred forty-seven undergraduates from a Midwestern
university participated for course credit. Six participants who
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
Head-
phones
Gel
Mask
Slippers Keyboard Speakers Massager Puzzle Apples
Game
Pet
Lounge
Salt &
Pepper
Decanter Cheese
Board
Proportion of
participants
who selected
the product
Browsers
Choosers
Fig. 2. Proportion of participants selecting each product by condition (Experiment 1).
376 S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
failed to follow instructions (e.g., texting during the experi-
ment) were excluded. We initially collected a baseline measure
of eight emotions (sad,indifferent,angry,happy,depressed,
enraged,amused,neutral)on0–100 scales.
We then induced either sadness or anger. In the Sadness
condition, participants viewed the same clip from The Champ
used in Experiment 1. In the Anger condition, participants viewed
a2.5-minuteclipfromCry Freedom that portrays young,
unarmed protesters being gunned down by opposition forces.
This clip reliably induces anger (Rottenberg et al., 2007).
Following the emotion induction, participants were told that
they would be given a $5 spending budget as part of a real
shopping experience. They were told that they could buy one of
three products (a set of post-it notes, a set of highlighters, or a set
of ball point pens), all offered in the lab for $5 (approximately
equal to their actual retail prices). They were also told that they
could choose to buy none of these products, and trade in their
spending budget for $1 in real cash. We disincentivized not
buying because walking away with the full $5 budget would
have been an easy choice for most participants (as they did not
come to the lab intending to stock up on office supplies). This
ensured that there was engagement in the shopping task and that
choices were actively considered. Our sadness versus anger
induction did not significantly influence the proportion of
participants choosing to buy one of the products (62% vs. 74%,
χ
2
(1) = 2.46, pN.10). We pooled across the buyer versus
non-buyer distinction in our analyses, since both actively made
a choice.
We also embedded a manipulation of personal agency in the
shopping task by adapting a procedure validated by Berman and
Small (2012). Specifically, we told participants that, after they
made their choice, the computer would randomly draw a number
from 1 to 10. If the randomly generated number was even, they
would simply obtain whatever they chose. However, if the
randomly generated number was odd, the computer would ignore
their choice and randomly make a selection on their behalf. Note
that there were four possible choices in our paradigm —buying
the post-it notes, buying the highlighters, buying the pens, or
trading in the spending budget for $1 in real cash. Regardless of
participants' own choice, if the randomly generated number was
odd, participants believed that the computer could randomly
select any one of these four options on their behalf.
In the Personal Control condition, the randomly selected
number was even, and participants were told that they would
obtain what they chose. In the Situational Control condition,
the randomly selected number was odd, and then the computer
ostensibly made a random selection on their behalf (in fact,
always selecting the option participants had selected for
themselves). Thus, at the end of this task and before collecting
our final measure of emotions, all participants knew that they
would obtain the option they preferred. All that varied across
the Personal Control versus Situational Control conditions was
the amount of control participants believed they had over the
process.
We then administered a second (and final) measure of
emotions, identical to our baseline measure. Finally, partici-
pants either obtained their selected product or $1 in cash.
Results
Sad and depressed ratings correlated highly (r(139) N.42,
pb.0001 at each measurement), and were averaged to form a
sadness index at each measurement. Angry and enraged ratings
correlated highly (r(139) N.56, pb.0001 at each measure-
ment), and were averaged to form an anger index at each
measurement. We created residual sadness and anger scores by
subtracting participants' baseline index scores from their final
index scores.
Fig. 3 displays the focal residual emotion in each condition (i.e.,
residual sadness in the Sadness conditions and residual anger in the
Anger conditions). As predicted, residual sadness scores in the
Sadness conditions were significantly higher in the Situational
Control condition than in the Personal Control condition (M=3.73,
SD = 10.78 vs. M=−2.28, SD = 11.79; t(69) = 2.24, pb.05).
However, residual anger scores in the Anger conditions
did not differ significantly between the Situational Control
and Personal Control conditions (M= 14.75, SD = 25.69 vs.
M=14.17, SD = 30.91; t(68) = .09, p=.93).
1
Experiment 2 suggests that real shopping can help to
alleviate residual sadness, unless that shopping experience
further reduces personal control over situational forces. By
contrast, restoring personal control is as ineffective as further
jeopardizing personal control at alleviating residual anger.
Anger is naturally associated with a sense that other people are
likely to cause negative outcomes, and thus restoring control
over ambient environmental forces does not appear to address
the key control deficit associated with anger.
Experiment 3
Experiment 3 extends our investigation by focusing more
closely on the underlying mechanism of personal control
restoration. Specifically, we examine whether a direct manipu-
lation of control over one's environment (outside of a shopping
context) is more likely to influence sadness than anger.
Method
Three hundred and one undergraduates from a Southern
university (51% female) participated for course credit. We
initially collected a baseline measure of nine emotions (happy,
angry,sad,indifferent,enraged,depressed,amused,anxious,
and neutral)on0–100 scales.
1
It is worth noting that the Situational Control versus Personal Control
manipulation did not operate by influencing happiness, despite the fact that
residual happiness (final happy rating minus baseline happy rating) was
significantly correlated with both residual sadness (r(139) = −.48, pb.0001)
and residual anger (r(139) = −.28, pb.001), pooling across conditions.
Residual happiness in the Sadness conditions did not differ significantly
between the Situational Control and Personal Control conditions (M= 7.02,
SD = 28.22 vs. M= 4.61, SD = 25.49; t(69) = .38, p= .71). Likewise,
residual happiness in the Anger conditions did not differ significantly between
the Situational Control and Personal Control conditions (M= 7.83, SD = 33.38
vs. M= 17.15, SD = 29.96; t(68) = 1.23, p= .22).
377S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
We then induced sadness, anger, or neutral emotion. We
included a Neutral condition to examine whether our subse-
quent control manipulation (discussed below) generated any
emotional costs or benefits even in the absence of a negative
emotion induction. In the Sadness condition, participants
viewed the clip from The Champ used in previous experiments.
In the Anger condition, participants viewed a four-minute clip
from My Bodyguard that portrays a bullying incident. This clip
reliably induces anger (Rottenberg et al., 2007). In the Neutral
condition, participants viewed a 2.5-minute clip from National
Geographic about coral reefs (cf. Lerner et al., 2004).
Following the emotion induction, we asked participants to
either recall an instance in which they experienced control over
an important situation (Personal Control condition) or experi-
enced no control over an important situation (Situational
Control condition). We carefully worded our recall prompts to
prevent participants from considering instances in which they
had control over other people, or were controlled by other
people (see Appendix).
We then administered a second (and final) measure of
emotions, identical to our baseline measure.
Results
Sad and depressed ratings correlated highly (r(299) N.67,
pb.0001 at each measurement), and were averaged to form a
sadness index at each measurement. Angry and enraged ratings
correlated highly (r(299) N.73, pb.0001 at each measure-
ment), and were averaged to form an anger index at each
measurement. We created residual sadness and anger scores by
subtracting participants' baseline index scores from their final
index scores.
Fig. 4 displays the focal residual emotion in the Sadness and
Anger conditions. As predicted, residual sadness scores in the
Sadness conditions were significantly higher in the Situational
Control condition than in the Personal Control condition (M=
12.1, SD =14.8vs.M=2.89,SD = 16.3; t(99) = 2.97, pb.01).
However, residual anger scores in the Anger conditions did not
differ significantly between the Situational Control and Personal
Control conditions (M=4.6,SD =14.8vs.M=4.7,SD =17.7;
t(97) = .03, p=.97).
In the Neutral conditions, both residual sadness and residual
anger were slightly higher in the Situational Control condition
than in the Personal Control condition, though neither difference
was statistically significant (residual sadness: M=3.7,SD =15.5
vs. M=−0.3, SD =7.3; t(99) = 1.74, p= .085; residual
anger: M=3.6,SD =11.0vs.M=0.2,SD =6.1;t(99) = 1.97,
p= .052). Residual scores for all the other individual emotions
measured (happy,indifferent,amused,anxious,neutral)alsodid
not differ significantly across the Situational Control and Personal
Control conditions (all psN.05). Thus, the Neutral condition
offers marginal evidence suggesting that personal control can help
to mitigate mild levels of naturally occurring sadness and anger.
However, Fig. 4 suggests that when experiences of sadness and
anger are more focal and acute, personal control uniquely helps to
alleviate sadness.
Thus, consistent with our theoretical framework (and
Experiment 2), Experiment 3 suggests that the acute experience
of sadness (which is naturally associated with a sense that
situational forces control outcomes) is smaller when personal
Note: Error bars represent ± 1 standard error.
Fig. 4. Focal residual emotion by condition (Experiment 3).
Note: Error bars represent ± 1 standard error.
20
Situational Control
Personal Control
15
10
Intensity of
Residual Emotion
5
0
-5
Residual Sadness
in Sadness Conditions
Residual Anger
in Anger Conditions
Fig. 3. Focal residual emotion by condition (Experiment 2).
378 S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
control over one's environment is restored than when it is further
jeopardized. By contrast, the acute experience of anger (which is
naturally associated with a sense that other people cause negative
outcomes) is unaffected by differences in control over one's
environment. To the extent that making shopping choices
enhances feelings of personal control over one's environment,
these results suggest that shopping is likely to alleviate sadness
but not necessarily anger.
General discussion
Previous research suggests that distress can increase willingness
to spend (e.g., Atalay & Meloy, 2011; Lerner et al., 2004), but the
question of whether “retail therapy”actually helps to reduce
distress has only been addressed in correlational designs, utilizing
surveys and interviews of people who chose to shop when feeling
bad. We addressed this gap by experimentally examining whether
making shopping choices could help to reduce residual sadness.
Three experiments provided support for the notion that
making shopping choices helps to restore a sense of personal
control over one's environment, and thus helps to alleviate
residual sadness. We observed these benefits regardless of
whether the shopping was hypothetical or real. We also
documented support for the underlying mechanism of personal
control restoration. We found that the effects of manipulating
personal control over one's environment did not generalize to
anger. Anger is associatedwith a sense that other people are likely
to cause negative outcomes, and changes in personal control over
situational forces cannot necessarily reduce the extent to which
other people are viewed as unfairly or unduly influential.
Our work contributes to research on emotion and decision-
making. Most work in this area has focused on how specific
emotions influence decision-making and consumption (e.g.,
Cryder et al., 2008; Garg & Lerner, 2013; Garg, Wansink, &
Inman, 2007; Lerner et al., 2004). By contrast, our paper joins a
growing stream of research examining how decision-making
influences the experience of specific emotions (Berman &
Small, 2012; Gal & Liu, 2011).
Our work also contributes to appraisal theories of emotion.
Prior work had demonstrated that the cognitive appraisals that
accompany an emotion can be deactivated by addressing the
source of emotion (e.g., the cognitive effects of anger are extin-
guished when the perpetrator that caused the anger is punished;
Goldberg, Lerner, & Tetlock, 1999). Our work suggests that
counteracting a particular cognitive appraisal (here, restoring
personal control after it has been lost) may help to extinguish the
emotion that elicited the appraisal, consistent with the “recursive”
nature of emotions and their associated appraisals (Han et al.,
2007).
Limitations and future directions
An ideal test of the impact of restored personal control on the
experience of sadness and anger would require inductions that
initially produced similarly intense experiences of sadness and
anger (before the focal control manipulation). Otherwise,
differences in initial emotional intensity could potentially explain
differences in responsiveness to the focal control manipulation.
Unfortunately, pooling across the Personal and Situational
Control conditions in Experiment 2, residual anger in the Anger
conditions was significantly greater than residual sadness in the
Sadness conditions (14.45 vs. 0.68, pb.01). By contrast, residual
anger in the Anger conditions was slightly lower than residual
sadness in the Sadness conditions in Experiment 3 (4.67 vs. 7.53,
p= .21). Despite these differences (within and across experi-
ments), we always found that residual sadness was sensitive to the
Personal versus Situational control manipulations, but residual
anger never was. Thus, our results do not appear to be an artifact
of intensity differences across induced emotions. (An intensity
account would suggest that only the most intense or least intense
emotion would be influenced by our control manipulations.)
Future research on retail therapy could examine shopping
aspects other than choice that might address the key symptom
(sadness) but not its cause (loss of personal control). While we
carefully controlled for many features associated with shopping
so that we could isolate restoration of control as a key feature of
retail therapy, we believe that residual sadness may also be
reduced in ways that do not directly address the control deficit.
For example, social interaction may increase positive emotions
and perhaps mitigate sadness (O'Guinn & Faber, 1989). The
distraction provided by shopping is another possibility. Neither
of these shopping aspects replace the lost control intrinsic to
sadness, but both may impact a consumer's overall emotional
state.
Future research could also further explore the foundational
link between control and sadness. We found that restoring
personal control following exposure to sad stimuli helped to
reduce residual sadness. Future work could examine whether
imbuing people with high personal control helps immunize
them from the effects of sad stimuli encountered later.
Conclusion
Retail therapy is often considered wasteful and ineffective. For
example, when we asked 100 adults (52% female, mean age: 35)
for the first word that came to mind when hearing “retail therapy,”
they were more than twice as likely to provide a clear negative
response (e.g., nonsense,debt) than to provide a clear positive
response (e.g., fun,enjoyment) (19% vs. 8%; χ
2
(1) = 5.18,
pb.05; other responses were neutral, such as shopping). But no
prior research had experimentally examined whether and why
shopping when sad can actually help to reduce residual sadness.
Our work suggests that making shopping choices can help to
restore a sense of personal control over one's environment and
reduce residual sadness. Whether the increased control afforded
by shopping results in a loss of control later (due to increased
debt), and thus counteracts the temporary benefits of retail
therapy, remains an important open question.
Authors' note
We thank the Editor, the Associate Editor, three anonymous
reviewers, Simona Botti, Cynthia Cryder, and Gabriele Paolacci
for valuable feedback.
379S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
Appendix. Control inductions used in Experiment 3
Situational control
Please recall a particular incident in which you did not have
any control over a situation important to you. By control, we
mean a circumstance in which you could not control your
environment in a way that allowed you to achieve something
you wanted. For instance, this could be a time when you failed
to exert the control necessary to overcome an obstacle, or when
your actions could not change an important situation to meet
your needs. Note that this does not involve lack of control—or
lack of power—over other people, just lack of control over
your environment. Please describe this situation in which you
did not have any control—what happened, how you felt, etc.
Personal control
Please recall a particular incident in which you had complete
control over a situation important to you. By control, we mean a
circumstance in which you controlled your environment in a way
that allowed you to achieve something you wanted. For instance,
this could be a time when you succeeded in exerting control to
overcome an obstacle, or when your actions effectively changed
an important situation to meet your needs. Note that this does not
involve control—or power—over other people, just control over
your environment. Please describe this situation in which you had
complete, effective control—what happened, how you felt, etc.
References
Atalay, A. S., & Meloy, M. G. (2011). Retail therapy: A strategic effort to
improve mood. Psychology and Marketing,28(6), 638–659.
Berman, J. Z., & Small, D. A. (2012). Self-Interest without selfishness: The
hedonic benefit of imposed self-interest. Psychological Science,23(10),
1193–1199.
Cryder, C. E., Lerner, J. S., Gross, J. J., & Dahl, R. E. (2008). Misery is not
miserly: Sad and self-focused individuals spend more. Psychological
Science,19(6), 525–530.
Faber, R. J., & Christenson, G. A. (1996). In the mood to buy: Differences in
the mood states experienced by compulsive buyers and other consumers.
Psychology and Marketing,13(8), 803–819.
Gal, D., & Liu, W. (2011). Grapes of wrath: The angry effects of self-control.
Journal of Consumer Research,38(3), 445–458.
Gardner, M. P., & Rook, D. W. (1988). Effects of impulse purchases on
consumers' affective states. In M. J. Houston (Ed.), Advances in Consumer
Research. (pp. 127–130) Provo, UT: Association for Consumer Research.
Garg, N., & Lerner, J. S. (2013). Sadness and consumption. Journal of
Consumer Psychology,23(1), 106–113.
Garg, N., Wansink, B., & Inman, J. J. (2007). The influence of incidental affect
on consumers' food intake. Journal of Marketing,71(1), 194–206.
Goldberg, J. H., Lerner, J. S., & Tetlock, P. E. (1999). Rage and reason: The
psychology of the intuitive prosecutor. European Journal of Social
Psychology,29, 781–795.
Han, S., Lerner, J. S., & Keltner, D. (2007). Feelings and consumer decision
making: The appraisal-tendency framework. Journal of Consumer Psychology,
17(3), 158–168.
Inesi, M. E., Botti, S., Dubois, D., Rucker, D. D., & Galinsky, A. D. (2011).
Power and choice: Their dynamic interplay in quenching the thirst for
personal control. Psychological Science,22(8), 1042–1048.
Kasser, T., & Sheldon, K. M. (2000). Of wealth and death: Materialism, mortality
salience, and consumption behavior. Psychological Science,11(4), 348–351.
Kermer, D. A., Driver-Linn, E., Wilson, T. D., & Gilbert, D. T. (2006). Loss
aversion is an affective forecasting error. Psychological Science,17(8),
649–653.
Kim, S., & Rucker, D. D. (2012). Bracing for the psychological storm:
Proactive versus reactive compensatory consumption. Journal of Consumer
Research,39(4), 815–830.
Langer, E. J. (1975). The illusion of control. Journal of Personality and Social
Psychology,32(2), 311–328.
Lerner, J. S., & Keltner, D. (2001). Fear, anger, and risk. Journal of Personality
and Social Psychology,81(1), 146–159.
Lerner, J. S., Small, D. A., & Loewenstein, G. (2004). Heart strings and purse
strings: Carryover effects of emotions on economic decisions. Psychological
Science,15(5), 337–341.
Mazar, N., & Zhong, C. -B. (2010). Do green products make us better people?
Psychological Science,21(4), 494–498.
O'Guinn, T. C., & Faber, R. J. (1989). Compulsive buying: A phenomenolog-
ical exploration. Journal of Consumer Research,16, 147–157.
Oveis, C., Cohen, A. B., Gruber, J., Shiota, M. N., Haidt, J., & Keltner, D.
(2009). Resting respiratory sinus arrhythmia is associated with tonic
positive emotionality. Emotion,9(2), 265–270.
Paolacci, G., Chandler, J., & Ipeirotis, P. G. (2010). Running experiments on
Amazon Mechanical Turk. Judgment and Decision Making,5(5), 411–419.
Raghunathan, R., Pham, M. T., & Corfman, K. P. (2006). Informational
properties of anxiety and sadness, and displaced coping. Journal of
Consumer Research,32(4), 596–601.
Rogosa, D. R., & Willett, J. B. (1983). Demonstrating the reliability of the
difference score in the measurement of change. Journal of Educational
Management,20(4), 335–343.
Rottenberg, J., Ray, R. D., & Gross, J. J. (2007). Emotion elicitation using
films. In J. A. Coan, & J. J. B. Allen (Eds.), The handbook of emotion
elicitation and assessment (pp. 9–28). New York: Oxford University Press.
Smith, C. A., & Ellsworth, P. C. (1985). Patterns of cognitive appraisal in
emotion. Journal of Personality and Social Psychology,48(4), 813–838.
Spencer, S. J., Zanna, M. P., & Fong, G. T. (2005). Establishing a causal chain:
Why experiments are often more effective than mediational analyses in
examining psychological processes. Journal of Personality and Social
Psychology,89(6), 845–851.
Tuttle, B. (2010, July 19). Seeking ‘retail therapy’when what you need is real
therapy. Time Magazine, July 19, 2010 (Retrieved from http://moneyland.
time.com/2010/07/19/too-much-retail-therapy-and-youll-need-real-therapy)
Wilson, T. D., Wheatley, T., Meyers, J. M., Gilbert, D. T., & Axsom, D. (2000).
Focalism: A source of durability bias in affective forecasting. Journal of
Personality and Social Psychology,78(5), 821–836.
380 S.I. Rick et al. / Journal of Consumer Psychology 24, 3 (2014) 373–380
