Content uploaded by Stacey Finkelstein
Author content
All content in this area was uploaded by Stacey Finkelstein on May 11, 2015
Content may be subject to copyright.
Journal of Consumer Research Inc.
Tell Me What I Did Wrong: Experts Seek and Respond to Negative Feedback
Author(s): Stacey R. Finkelstein and Ayelet Fishbach
Source:
Journal of Consumer Research,
(-Not available-), p. 000
Published by: The University of Chicago Press
Stable URL: http://www.jstor.org/stable/10.1086/661934 .
Accessed: 22/08/2011 14:18
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact support@jstor.org.
The University of Chicago Press and Journal of Consumer Research Inc. are collaborating with JSTOR to
digitize, preserve and extend access to Journal of Consumer Research.
http://www.jstor.org
000
䉷2011 by JOURNAL OF CONSUMER RESEARCH, Inc. ●Vol. 39 ●June 2012
All rights reserved. 0093-5301/2012/3901-0003$10.00. DOI: 10.1086/661934
Tell Me What I Did Wrong: Experts Seek
and Respond to Negative Feedback
STACEY R. FINKELSTEIN
AYELET FISHBACH
A large proportion of marketing communication concerns feedback to consumers.
This article explores what feedback people seek and respond to. We predict and
find a shift from positive to negative feedback as people gain expertise. We doc-
ument this shift in a variety of domains, including feedback on languageacquisition,
pursuit of environmental causes, and use of consumer products. Across these
domains, novices sought and responded to positive feedback, and experts sought
and responded to negative feedback. We examine a motivational account for the
shift in feedback: positive feedback increased novices’ commitment, and negative
feedback increased experts’ sense that they were making insufficient progress.
Feedback is essential for individuals pursuing their goals.
Without it, individuals would not know whether, what,
and how much to invest in their goals (Ashford, Blatt, and
Van de Walle 2003; Frey and Ruble 1987; Kruglanski 1990;
Miller and Ross 1975; Swann and Read 1981; Wood 1989).
Accordingly, a large proportion of marketing communica-
tion involves collecting information on consumers and pro-
viding tailored feedback. For example, language programs
provide feedback to consumers on their mastery of a foreign
language, skin-product salespeople advise customers on how
to improve their skin-care regimen, and media campaigns
provide feedback to the public on the effectiveness of en-
vironmental actions. Given the pervasiveness of feedback
in marketing communication, understanding what feedback
consumers seek and how they respond to it as they gain
experience is important. In particular, in this article we ex-
amine whether, as consumers accumulate knowledge or gain
experience, their interest and response to feedback changes.
Generally speaking, we distinguish between positive feed-
back on strengths, correct responses, and accomplishments and
negative feedback on weaknesses, incorrect responses, and lack
Stacey Finkelstein (sf2559@columbia.edu) is assistant professor of
health policy and management at the Mailman School of Public Health at
Columbia University, 600 West 168th Street, New York, NY 10032. Ayelet
Fishbach (ayelet.fishbach@chicagobooth.edu) is professor of behavioral
science and marketing at the Booth School of Business, University of
Chicago, 5807 South Woodlawn Avenue, Chicago, IL 60637. Correspon-
dence may be addressed to either author.
Baba Shiv served as editor and Joel Huber served as associate editor for
this article.
Electronically published August 15, 2011
of accomplishments. For these two types of information to
constitute “feedback,” they need to be constructive: positive
information should not be needlessly flattering, and negative
information should not be unnecessarily detrimental. Instead,
both types of feedback should be beneficial by suggesting cor-
rective actions (see, e.g., Dweck and Leggett 1988). For ex-
ample, positive feedback will emphasize a consumer’s correct
use of cosmetic products, and negative feedback will emphasize
her incorrect use of these products and how she can improve.
In this article, we explore whether expertise (perceived
or actual) influences the type of feedback individuals seek
and respond to. In what follows, we present our theory and
findings in support of a shift toward seeking and responding
to negative feedback with increased expertise.
THEORETICAL BACKGROUND
Whether people acquire a new skill, learn to use a new prod-
uct, or seek to improve their behavior, both positive and neg-
ative feedback can allow for realistic self-assessment and ad-
justment of their efforts (Carver and Scheier 1998; Higgins
1987; Maheswaran and Meyers-Levy 1990). Clearly, addi-
tional reasons exist for why people might seek feedback,
including enhancing and maintaining their positive view of
themselves (Russo, Meloy, and Medvec 1998; Tormala and
Petty 2004). For example, people seek positive information
about products they are already using because such infor-
mation provides positive feedback that confirms their choices
(Ahluwalia, Burnkrant, and Unnava 2000; see also Wood,
Rhodes, and Biek 1995). However, when people wish to
change or improve their actions, the motivation to enhance
a positive view is often secondary to the motivation to re-
000 JOURNAL OF CONSUMER RESEARCH
alistically assess their skills and gain a sense of which direc-
tion they should pursue (Trope 1986).
With the objective of accurate self-assessment in mind, both
(constructive) positive and negative feedback on one’s perfor-
mance are potentially useful, and people might differentially
attend to positive and negative feedback over the course of
gaining experience or expertise on a goal. For example, to
maintain the motivation to improve, a person who is looking
to master a new language might desire different types of feed-
back at different points over the course of learning the language.
Our main proposition is that as people gain expertise in pursuing
a goal, they seek and respond more to negative than to positive
feedback. In what follows, we explore the possible reasons the
shift occurs and identify our leading reason—a motivational
explanation—which we explore in our studies.
A SHIFT TOWARD NEGATIVE
FEEDBACK
One potential reason we could predict an increase in seeking
and responding to negative feedback is that the informa-
tional value of the feedback could differ for novices and
experts. People can learn more from feedback on unusual
performance than on usual performance. According to this
potential account, positive feedback might be rarer and
therefore more informative for novices—those who are less
likely to perform a task well—whereas negative feedback
might be rarer and therefore more informative for experts
—those who are unlikely to perform poorly (Ashford and
Tsui 1991; Tesser 1988). For instance, a beginning piano
player is less likely to play a piece of music perfectly; she
is likely to make many mistakes. For this player, who rarely
plays the right note at the right time, hearing that she played
a series of notes correctly is more informative than com-
ments on a series of correct notes would be for a professional
piano player who already knows he plays most of the notes
correctly. On the other hand, a professional piano player is
unlikely to miss notes. Hearing that he missed some notes
is rare and carries more value than hearing of missed notes
would carry for a novice.
Whereas the informational account could potentially cre-
ate a shift toward seeking and responding to negative feed-
back as people gain expertise, it holds only to the extent
that novices and experts are evaluated on a similar scale.
On that scale, novices would indeed perform poorly more
frequently than experts. However, if the evaluation scales
are different (Brown and Hanlon 1971), novices do not per-
form poorly more frequently than experts and negative feed-
back is not more frequent for them. For example, a profes-
sional piano player expects to be evaluated based on his
ability to express his emotions, and his likelihood of suc-
ceeding should not be higher than that of the amateur pianist,
who expects to be evaluated based on her ability to play
the right notes. Because different scales are applied, negative
feedback is not less frequent for experts, and if a shift toward
negative feedback with expertise exists, an informational
account cannot explain it.
We propose instead a motivational account for the shift.
The motivational account suggests that the meaning people
derive from feedback on their goals changes over time and
that people seek either positive or negative feedback de-
pending on its meaning and its ability to serve as a moti-
vational tool that allows them to focus on tasks at hand.
Specifically, feedback can inform individuals of either their
level of commitment to or their rate of progress toward the
goal (Fishbach and Dhar 2005; Fishbach, Zhang, and Koo
2009). When feedback informs people of their commitment,
it provides information on the value of a goal and one’s
likelihood of success (Bandura 1991; Feather 1982; Fishbein
and Ajzen 1975; Fo¨rster, Liberman and Higgins 2005;
Vroom 1964). In this case, positive feedback on one’s ac-
complishments (e.g., that a person answered some answers
correctly) is more motivating because it signals that the goal
is valuable or that one’s likelihood of attaining the goal is
high. In contrast, when feedback informs people about their
rate of progress, it provides information about the rate of
progress relative to expectations (Carver and Scheier 1998;
Higgins 1987; Locke and Latham 1990; Miller, Galanter,
and Pribram 1960). In this instance, negative feedback (e.g.,
that a person responded incorrectly) increases motivation
because it signals insufficient progress. For example, a stu-
dent who wishes to motivate herself to study for an exam
would seek positive feedback if she wants to increase her
commitment but negative feedback if she wants to encourage
herself to progress at a more sufficient pace.
Earlier research by Koo and Fishbach (2008) demon-
strated that the meaning of feedback indeed determines its
motivational impact. These researchers compared feedback
on completed versus missing actions toward a goal, which
could represent positive versus negative feedback. They
found that uncommitted individuals, who infer their level
of commitment from feedback, worked harder after getting
feedback on completed actions. In contrast, committed in-
dividuals, who infer their progress from feedback, worked
harder after getting feedback on missing actions. For in-
stance, uncommitted individuals (those on the “cold list”)
donated more to a charitable organization when they read
that the charity had raised $5,000 thus far, whereas com-
mitted individuals (those on the “cold list”) donated more
when they read that the charity still needed to raise $5,000.
Building on this previous research, we predict that, as
people gain expertise, their way of motivating themselves
through feedback seeking changes. Thus, people move
from evaluating commitment to monitoring progress as
they gain expertise and seek more negative feedback. Com-
pared with experts, novices feel uncertain about their levels
of commitment. Positive feedback on novices’ goal pursuit
instills a sense of confidence that they can perform the
goal and encourages novices to internalize or integrate new
goals into their self-concept, thus increasing their com-
mitment to pursue the goal on subsequent occasions (Ryan
and Deci 2000). However, experts’ commitment is more
secure than novices’, and their focus is on monitoring their
TELL ME WHAT I DID WRONG 000
TABLE 1
OPERATIONAL DEFINITIONS OF EXPERTISE BY STUDY
Dimension of expertise Study
Formal training Study 1: Enrollment in beginner (novices) versus advanced (experts) French class
Study 4: Training in typing in an unfamiliar language
Frequency of performing goal-related actions Study 2: Affiliation with environmentalist groups (experts) versus not (novices)
Study 3: Perceived frequency of using nail-care services (high: experts; low: novices)
Knowledge Study 5: Perceived knowledge of environmental issues (high: experts; low: novices)
progress. Negative feedback signals to experts that they
should increase their efforts.
By this analysis, the same feedback (e.g., “you have a
good skin care regimen”) can either convey the extent of
commitment to pursuing the goal or how much progress one
has made toward the goal, and the meaning and its moti-
vational consequences depend on a person’s expertise level.
When novices hear they do well, they infer that their goals
are valuable and their expectancy of attainment is high; that
is, they interpret that feedback to mean they are committed
to the goal. In contrast, when experts hear the same feed-
back, they interpret that feedback as signaling that they have
invested enough effort toward pursuing their goals and thus
have made sufficient progress. For example, Louro, Pieters,
and Zeelenberg (2007) found that positive feedback moti-
vated dieters who were far from their weight-loss goal, and
we assume that such feedback increased their goal com-
mitment. In addition, in the authors’ study, negative feed-
back motivated those who were close to their weight-loss
goal, and we assume that such feedback increased dieters’
sense of insufficient progress.
Assuming that positive feedback has a greater impact on
novices than experts and that negative feedback has a greater
impact on experts than novices, we predict that experts will
actively seek more negative feedback than novices, whereas
novices will actively seek more positive feedback than ex-
perts. We further predict that experts will increase their ef-
forts more than novices in response to negative feedback,
whereas novices will increase their efforts more than experts
in response to positive feedback.
These hypotheses extend existing theory on feedback on
self-regulation (Koo and Fishbach 2008) to the domain of
consumer expertise in a few notable ways. First, we predict
that expertise (rather than prior commitment) creates a shift
toward negative feedback. Second, we predict effects on
feedback seeking, thus moving beyond the impact of feed-
back on peoples’ responses. Third, we predict a dynamic
process where, as people gain expertise, they seek more
negative feedback. Finally, whereas earlier work studied
only one aspect of feedback (missing vs. completed actions),
we focus on other, more explicit aspects of feedback (e.g.,
correct vs. incorrect actions).
PRESENT RESEARCH
We report five studies that test the hypotheses that expertise
is associated with seeking and responding to negative feed-
back. Prior research used various paradigms to assess ex-
pertise, including the frequency of performing goal-related
actions (Bettman and Park 1980; Kiel and Layton 1981),
prior knowledge (Hong and Sternthal 2010), and formal
training (Hutchinson 1983; for a summary, see Alba and
Hutchinson 1987). In our studies, we rely on these various
definitions to operationalize expertise. We both measure ex-
pertise (studies 1 and 2) and manipulate expertise (studies
3–5). In studies 1 and 2, we compare experts who frequently
engage in a goal with novices who do not. Because indi-
vidual differences associated with self-selection might drive
the preference for feedback, in studies 3–5, we manipulate
expertise. Altogether, expertise in our studies is defined in
a way that generalizes across various lines of earlier work
(see table 1).
Specifically, in study 1, we explore the impact ofexpertise
on the feedback-seeking behavior of students in beginning-
level versus advanced-level French classes. In study 2, we
explore the impact of expertise in pursuing environmentally
friendly actions on feedback seeking, as well as the tendency
to respond to feedback by donating to an environmental
organization. In study 3, we examine how perceived ex-
pertise affects feedback seeking on women’s use of beauty
products and how feedback influences their willingness to
pay for such products. In study 4, we examine changes in
feedback-seeking behavior and the meaning that feedback
conveys (progress vs. commitment) over time as participants
gain expertise with an unfamiliar language task. Finally, in
study 5, we examine how perceived expertise influences
endorsement of persuasive messages on the collective per-
formance of a shared environmental goal.
STUDY 1: LANGUAGE CLASSES
Consumers invest resources, including effort, time, and
money, acquiring new skills such as learning a new language
and, in the process, seek positive and negative feedback on
their performances (Ward 1974). To explore the impact of
expertise, we investigated feedback seeking among Amer-
ican students enrolled in beginning- and advanced-level
French classes. We predicted that, compared with those in
advanced-level classes (experts), beginners would express
greater interest in learning from an instructor who teaches
using a style that emphasizes what they do well. In ad-
dition, compared with beginners, advanced students would
express greater interest in learning from an instructor who
000 JOURNAL OF CONSUMER RESEARCH
teaches using a style that emphasizes their mistakes and
how they can improve.
Method
Eighty-seven undergraduate students volunteered to par-
ticipate in the study immediately after their French class.
This study employed a 2 (expertise: beginner- vs. advanced-
level French class students) #2 (feedback: positive vs.
negative) mixed design in which expertise was a between-
subjects factor and feedback was a within-subjects factor.
The experimenter surveyed American students in begin-
ning-level conversational French classes and advanced-level
French literature classes. We assumed that students enrolled
in a class titled “beginning level” saw themselves (and were
referred to by others) as relative novices, whereas those
enrolled in a class titled “advanced” saw themselves as rel-
ative experts. Beginners primarily take classes focused on
conversational and grammatical skills and learn material de-
signed to help them communicate at a basic level. Ad-
vanced-level students primarily take classes designed for
reading classic French literature in French and writing pa-
pers in French that offer insightful analyses of the text.
Participants completed a questionnaire about choosing an
instructor, which the French department presumably created
to improve instructors’ training to better meet student needs.
Participants read that two basic styles of teaching exist: one
style is for an instructor to “emphasize what students do
well in class by providing the student with feedback on their
strengths, like when they pronounce new words well or write
well in French” (positive feedback), and the other style is
for the instructor to mostly provide negative feedback on
“what mistakes they make when, for instance, pronouncing
new words, conjugating new verbs, or writing and how they
can fix those mistakes” (negative feedback).
As a measure of feedback seeking, participants rated their
interest in taking a class with an instructor who teachesusing
each particular style (for each instructor: 1 pnot at all, 7
pvery much). They then listed, among other demographic
information, how long they had been taking French classes
(in months).
Results and Discussion
In support of the manipulation, students in the advanced-
level (literature) classes indicated that they had studied
French for a longer time (Mp78.64 months, SD p43.38)
than students enrolled in the beginning-level (conversation)
classes (Mp25.29 months, SD p27.05; t(72) p5.52,
p!.001).
To test the hypothesis, we compared participants’ interest
in taking a class with an instructor who uses a style em-
phasizing what they do well versus one who uses a style
emphasizing how they can improve, as a function of their
expertise. These measures were not correlated (r(85) p.05),
suggesting that participants’ interest in positive feedback
and their interest in negative feedback were largely inde-
pendent of each other. An expertise #feedback repeated
measures ANOVA yielded a main effect for feedback, in-
dicating that participants preferred an instructor who uses a
style emphasizing negative feedback on what mistakes they
make and how they can improve (F(1, 79) p6.43, p!.02).
We found no main effect for expertise. The analysis also
yielded the predicted expertise #feedback interaction (F(1,
79) p7.31, p!.01). Contrast analysis revealed that be-
ginners expressed greater interest than advanced students in
an instructor who uses a style that emphasizes what they do
well (Mp4.96, SD p1.15, vs. Mp4.25, SD p1.47;
t(79) p2.35, p!.05). Additionally, advanced students were
marginally more likely than novices to express an interest
in an instructor who uses a style that emphasizes negative
feedback on how they can improve (Mp5.45, SD p1.22,
vs. Mp4.92, SD p1.29; t(79) p1.76, pp.08; see fig.
1).These results are consistent with the hypothesis that nov-
ices seek positive feedback more than experts, presumably
because they more likely infer greater commitment, whereas
experts seek negative feedback more than novices, presum-
ably because they more likely infer insufficient progress.
Experts also seek more negative feedback than novices be-
cause they can tolerate negative feedback more easily—
negative feedback does not undermine their commitment
(e.g., expertise acts as a buffer; Linville 1987; Raghunathan
and Trope 2002; Trope and Neter 1994). We argue that in
addition to tolerating negative feedback, experts actively
seek negative feedback to motivate themselves to invest
effort in a goal. To demonstrate the latter point, we compared
experts’ interest in positive and negative feedback. We found
that students in the advanced course expressed greater in-
terest in an instructor who uses a style that emphasizes how
they can improve (Mp5.45, SD p1.22) than an instructor
who uses a style that emphasizes what they do well (Mp
4.25, SD p1.47; t(54) p4.44, p!.001); hence, in this
study, experts not only tolerated constructive negative feed-
back but preferred it over constructive positive feedback.
Interestingly, even novices were not averse to negative feed-
back. They were similarly interested in negative and positive
feedback, which further suggests that people are interested
in constructive negative feedback and that they seek it more
to the extent that they perceive themselves as experts.
In addition to their course enrollment, we evaluated par-
ticipants’ expertise based on the amount of time they had
studied French prior to the study. This variable was highly
skewed; thus we log transformed it. Collapsing across the
types of classes, the longer students had been enrolled in
French classes, the greater was their interest in an instructor
who uses a style that emphasizes how they can improve
(r(77) p.31, p!.01). Similarly, the longer students were
enrolled in French classes, the lower was their interest in
an instructor who uses a style that emphasizes what they do
well, though this effect was marginal (r(77) p⫺.19, p!
.10). These correlations provide further support for our hy-
pothesis and rule out an alternative explanation—that the
different content of the advanced- and beginning-level
courses affected the feedback students sought.
TELL ME WHAT I DID WRONG 000
FIGURE 1
INTEREST IN FEEDBACK FROM AN INSTRUCTOR WHO EMPHASIZES POSITIVE VERSUS NEGATIVE FEEDBACK
AS A FUNCTION OF EXPERTISE LEVEL (STUDY 1)
Study 1 demonstrates that experts seek more negative
feedback and less positive feedback than novices. We
predict that expertise further affects how people respond
to feedback. Accordingly, our next study tested for feed-
back seeking and for how people respond to feedback as
a function of their expertise.
STUDY 2: ENVIRONMENTAL ACTIONS
We conducted study 2 to examine whether expertise in pur-
suing environmentally friendly actions increases interest in
negative feedback on how people can improve their actions
and whether expertise further increases the tendency to re-
spond to negative feedback by donating to an environmental
organization (Greenpeace). We defined expertise as fre-
quency of performing goal-related actions—a concept we
measured by comparing members of environmental orga-
nizations (experts) with those who are not members (nov-
ices). Participants listed things they do to help the environ-
ment (e.g., reducing waste and conserving energy) and
indicated their interest in feedback on either their effective
or ineffective environmental actions. Thus, in this and sub-
sequent studies, we posed a trade-off between seeking pos-
itive and negative feedback. In a later session, we manip-
ulated the feedback participants received (regardless of what
they originally sought) and assessed its impact on their will-
ingness to donate to Greenpeace, an environmental charity.
We predicted that environmental experts would be more
interested in negative feedback and upon receiving negative
feedback would increase their donations more than novices
would.
Method
Eighty-one students (53 women) participated in the study
for a chance to win $25 in a lottery. The study employed
a 2 (expertise: novice vs. expert) #2 (feedback: positive
vs. negative) between-subjects design. We recruited indi-
viduals who were frequent attendees of several environ-
mental organizations on campus (experts) and individuals
who did not participate in any environmental organizations
(novices). They all took part in an online study on envi-
ronmental issues.
We conducted the study in two sessions. In the first ses-
sion, participants listed the things they do to help the en-
vironment. The format was open-ended, and participants
listed about 5–10 things they do—for example, recycling
paper, cans, and plastic bottles and trying not to waste water.
Participants next read that an environmental consultant
would evaluate their responses and was willing to provide
them with feedback on their actions. Because the consult-
ant’s time was limited, the consultant was willing to offer
them feedback on either their effective or ineffective actions
but not both. Participants indicated their feedback choice
(between [a] their actions that are effective for helping the
environment and [b] their actions that are ineffective for
helping the environment) and provided their e-mail ad-
dresses so that the experimenter could contact them for the
second session of the study.
The second session took place 2 weeks after the first to
reduce the likelihood of participants recalling which feed-
back they asked for, a procedure that allowed us to ran-
domize the feedback regardless of what participants sought
000 JOURNAL OF CONSUMER RESEARCH
FIGURE 2
DONATION AMOUNTS TO GREENPEACE AS A FUNCTION OF EXPERTISE LEVEL AND FEEDBACK RECEIVED (STUDY 2)
in the first part of the study. Participants received an e-mail
reminding them of the activities they had previously listed
that indicated how they helped the environment. The e-mail
contained a link to an external website where participants
would purportedly receive feedback on their habits.
Unbeknownst to the participants, the content of the neg-
ative or positive feedback participants received on each trial
was predetermined and equally informative. Because allpar-
ticipants wrote about recycling, the feedback referred to
participants’ recycling habits. Participants assigned to re-
ceive positive feedback about their effective actions read
that their recycling habits reduce the amount of materials
and energy manufacturers need to make goods. Participants
assigned to receive negative feedback about their ineffective
actions read that their recycling habits are ineffective as
certain items are not easily recycled; consequently, they
could improve their habits by taking more care to sort items
before placing them in the recycling bin. Thus, the feedback
framed participants’ own recycling actions either positively
or negatively.
Next, participants were reminded that as payment for the
study, they would be entered in a lottery to win $25. Our
variable of interest was how much of their future earnings,
if they were to win, participants would donate to Green-
peace. Participants were then debriefed and dismissed. In
their debriefings, none of the participants expressed suspi-
cion that the feedback was not individually tailored.
Results and Discussion
In support of the hypothesis, expert environmentalists
sought negative feedback more often (92%) than novices
(74%; x
2
(1) p3.81, pp.05). This result confirms that,
when facing a trade-off in feedback seeking, experts express
a greater interest in negative feedback than novices.
We next explored how experts versus novices respond to
feedback. To measure response to feedback, we examined
whether participants agreed to contribute some of their lot-
tery earnings (if they were to win) to Greenpeace and how
much. In support of the hypothesis, an ANOVA on donation
amounts (we coded no donations as zeros) yielded the pre-
dicted expertise #feedback interaction (F(1, 77) p16.24,
p!.001) and no main effects. Contrast analysis revealed
that experts who received negative feedback agreed to do-
nate more (Mp$8.53, SD p$9.54) than novices who
received the same feedback (Mp$1.24, SD p$2.46; t(44)
p3.75, p!.001). We obtained the reverse pattern for
positive feedback: novices who received positive feedback
donated more (Mp$8.31, SD p$8.81) than experts who
received the same feedback (Mp$2.92, SD p$6.23; t(33)
p2.12, p!.05, see fig. 2).
We found a similar pattern for response rates. We analyzed
the proportion of participants who agreed to donate anything
from their potential future earnings. Experts who received
negative feedback were more likely to donate to Greenpeace
(76%) than were novices who received the same feedback
(30%, x
2
p8.85, p!.01). Conversely, novices who re-
ceived positive feedback were more likely to donate (81%)
than were experts who received the same feedback (37%,
x
2
p7.00, p!.01)
Recall that we randomized the feedback that the partic-
ipants received such that some participants received feed-
back different from what they initially requested (e.g., they
requested negative feedback but received positive feedback).
To test whether our effect was limited to those who received
feedback that matched their preferences, we conducted an-
TELL ME WHAT I DID WRONG 000
other analysis with participants’ original choice—whether
it matched the feedback they received versus not—as an
additional variable. The three-way expertise #feedback #
match interaction was not significant (F!1), indicating a
similar expertise #feedback interaction among those who
received feedback that matched their request and those who
did not.
Study 2 extends our results to the environmental domain:
experts are more interested in negative feedback than nov-
ices. It further yields support for the hypothesis that experts
respond more to negative feedback than novices, as mea-
sured by donations to Greenpeace. Conversely, novices re-
spond more to positive feedback than experts. This lasteffect
can have ironic consequences: we find that after receiving
positive feedback, those who care less about the environ-
ment are more willing to take action.
One alternative explanation is that experts are more confident
in goal attainment and that increased confidence in goal at-
tainment, rather than increased expertise, drives the shifttoward
negative feedback. We propose that experts seek and respond
more to negative feedback because they are more committed
and therefore feedback conveys to them that they have made
insufficient progress and need to work harder to achieve their
environmental goals. By commitment, we mean that experts
value the goal and are confident about their ability to pursue
it (i.e., high sense of self-efficacy; Bandura 1991). However,
experts are not necessarily more confident than novices that
they will achieve the goal and might even be more pessimistic
than novices about goal attainment.
Indeed, in a follow-up, participants from the same en-
vironmental organizations and campus populations (Np
74) indicated how (a) confident and (b) optimistic they were
that humans would respond well to future environmental
crises (1 pnot at all confident/optimistic, 7 pvery con-
fident/optimistic). We asked about the group’s (humans’)
ability to respond to future environmental issues because of
a unique feature of environmental goals, namely, that society
as a whole, rather than one individual, must take action for
goals to be achieved. We collapsed these items (rp.46, p
!.05) and found that members of environmental organi-
zations (experts) were less confident and optimistic than
novices that humans would respond well to future environ-
mental crises (Mp2.13, SD p.66, vs. Mp2.64, SD p
1.09; t(71) p2.06, p!.05). Notably, these results echo
research by Kruger and Dunning (1999), which finds that
those who know the least (often novices) are the most over-
confident in their ability, often to the point that they do not
differ from experts in their perceived ability to pursue a goal
and success expectations.
In studies 1 and 2, we measured expertise through group
affiliation (i.e., course, environmental organization). Be-
cause group affiliation was measured rather than manipu-
lated, it may have been associated with other individual-
difference variables, and our ability to infer that expertise
caused the search for negative feedback is somewhat limited.
Moreover, group affiliation may have been directly asso-
ciated with goal commitment. Accordingly, the rest of our
studies operationalized expertise using a standard definition
of this concept as reflecting frequency of performing goal-
related actions, training, and knowledge (Alba and Hutch-
inson 1987). These operationalizations allow us to distin-
guish between expertise and its consequences for goal
commitment. We further propose that people’s perceptions
that they are experts, rather than their actual knowledge or
experience, drive our effects. Hence, in study 3, we test
whether making people feel like experts will influence them
to seek and respond more to negative feedback than those
who feel like novices.
STUDY 3: BEAUTY PRODUCTS
Study 3 examines how perceived expertise affects consum-
ers’ interest in feedback on their use of beauty products and
how such feedback influences their subsequent use of such
products. We focused on women’s use of nail-care services,
a burgeoning industry that pulls in roughly $1 billion in
revenue per year and is a rapidly growing market (“Market
Trends” 2010). We manipulated perceived expertise using
social comparison information (Schwarz et al. 1985). Be-
cause frequent consumption increases people’s perception
that they are experts in using some products (Bettman and
Park 1980; Kiel and Layton 1981), we asked participants
to rate their frequency of performing nail-care activities us-
ing scales that made them feel that they were either frequent
users (experts) or infrequent users (novices). As our depen-
dent measures, participants in study 3A chose between pos-
itive and negative feedback on their nail-care habits and
participants in study 3B indicated their willingness to pay
for a manicure, a nail-care activity, after receiving positive
versus negative feedback. Unlike study 2, we separated the
samples in studies 3A (seeking feedback) and 3B (respond-
ing to feedback) to avoid situations in which participants
are assigned to receive different feedback than what they
selected. We predicted that those who perceived that they
frequently performed nail-care activities (experts) would
seek more negative feedback than novices (study 3A) and
would respond more to negative feedback by expressing a
higher willingness to pay for a manicure than novices who
received the same negative feedback (study 3B).
Study 3A: Feedback Seeking
Method. Seventy-three women participated in the study
for monetary compensation. We enrolled only women be-
cause nail-care habits and manicures are of broader interest
to women than to men. The study employed a 2 (expertise:
novice vs. expert) between-subjects design.
Participants completed a study on women’s nail-care
habits, presumably as part of a broader study on women’s
habits regarding cosmetic products and services. The first
part of the survey manipulated expertise via social com-
parison scales. Those who were made to feel they that
perform nail-care activities infrequently (novices) re-
sponded to the following questions: (a) How often do you
get a manicure at a nail salon or beauty parlor? (b) How
000 JOURNAL OF CONSUMER RESEARCH
often do you get a pedicure at a nail salon or beauty parlor
(for aand b:1pless than twice a month, 2 p2–3 times
a month, 3 ponce a week)? and (c) How often do you
paint your finger- or toenails by yourself (1 pless than
twice a month, 2 p2–3 times a month, 3 ponce a week)?
Those who were made to feel that they perform nail-care
activities frequently (experts) responded to the same ques-
tions on different response scales (for manicure and pedicure
items: 1 pless than once every 2 years, 2 ponce every
2 years, 3 pmore than once every 2 years; and for do-it-
yourself: 1 pless than once every 2 years, 2 ponce every
2 years, 3 pmore than once every 2 years). All participants
further listed how often they get manicures, pedicures, and
paint their own finger- or toenails (open questions). Using
these scales, novice participants were more likely to choose
responses on the lower end of the scale and thus feel in-
experienced, whereas expert participants were more likely
to choose responses on the higher end of the scale and thus
feel relatively experienced. As a manipulation check, par-
ticipants indicated whether they feel they often perform nail-
care activities (1 pstrongly disagree, 7 pstrongly agree).
After completing the expertise manipulation, participants
took a few minutes to list in a space roughly half a page in
length what they do to maintain the health of their nails.
They listed, for example, that they drink plenty of water,
moisturize their skin, and use sunscreen to avoid excessive
sun exposure. Next, participants read that a beauty con-
sultant would evaluate their responses and would be avail-
able to provide them with feedback on either the nail-care
habits they do well or the way they can improve their nail-
care actions. Participants read that because the consultant’s
time was limited, they could only receive one piece of
feedback. Participants then indicated their choice of either
positive or negative feedback. Our variable of interest per-
tained to participants’ likelihood of choosing positive ver-
sus negative feedback.
Results. In support of the manipulation, participants who
perceived themselves as experts indicated that they feel they
perform nail-care activities more often (Mp5.11, SD p
2.02) than those who perceived themselves as novices (M
p4.03, SD p2.44; t(73) p2.06, p!.05). In support of
our hypothesis, participants who perceived themselves as
experts were more likely to seek negative feedback (100%)
than novices (73.68%, x
2
(1) p10.95, pp.001). Recall
that, as part of the manipulation, participants also provided
open-ended responses regarding how often they get mani-
cures, pedicures, or paint their nails themselves over the
course of a year. We averaged these individual-difference
variables (ap.62) and ran a binary logistic regression of
this variable on participants’ interest in feedback (0 pchose
negative feedback, 1 pchose positive feedback). The re-
gression revealed that the more actual experience women
had with nail care, the more likely they were to choose
negative feedback (bp⫺.12, Wald x
2
p12.66, p!.001).
Notably, as we would expect, actual expertise was similar
across conditions (M
Novices
p11.42 times per year, SD p
14.17, M
Experts
p9.27, SD p10.19; t(71) !1, NS), thus
lending credence to our hypothesis that perceived expertise
(which may or may not correspond to actual expertise)
drives our effect.
As in previous studies, we find that people are generally
open to negative feedback on how they can improve: they
mostly prefer nonthreatening negative feedback over posi-
tive feedback on what they do well. More relevant for the
present investigation, we find that people are more open to
negative feedback if they see themselves as experts in a
domain. This effect extends our previous results on mea-
sured expertise (i.e., French students, environmentalists) to
manipulated expertise, allowing us to infer that the percep-
tion that one is an expert causes the increase in seeking
negative feedback. Next, we test for a similar relationship
between perceived expertise and responding to negative
feedback.
Study 3B: Responding to Feedback
Method. Fifty-three women participated for monetary
compensation. The study employed a 2 (expertise: novice vs.
expert) #2 (feedback: positive vs. negative) between-sub-
jects design.
Participants read similar instructions and completed a sim-
ilar perceived expertise manipulation, in one of the two
conditions, as participants in study 3A. Next they answered
several questions regarding their specific habits. The purpose
of these questions was to increase the perception that the
detailed feedback that followed was reliable and personal.
Specifically, participants listed the following: how many
glasses of unflavored water they drink every day; how many
glasses of coffee, juice, tea, or other flavored beverages they
drink; how many times per week, on average, they mois-
turize their hands (open-ended questions); and how long
their finger- and toenails are (1 pvery short, 5 pvery
long). Finally participants checked off from a list of foods
the items they eat that contain calcium (e.g., milk, cheese,
spinach, beans, tuna, nuts, tofu, oranges, and oats).
Next the participants received positive or negative feed-
back, depending on the experimental condition. Unbe-
knownst to participants, the content of the negative or pos-
itive feedback participants received on each trial was
predetermined and equally informative. Those who were
assigned to receive positive feedback read the following:
that they do a good job drinking plenty of water, which
would keep their nails strong and healthy; that they do a
nice job ensuring their hands are moisturized on a weekly
basis, which would help maintain their skin’s elasticity; that
their nails are at a good length, which would help reduce
breakage and chipping (if participants indicated their nails
are short to medium length), or that being able to grow their
nails so long is a sign of strong, healthy nails (if participants
indicated their nails are long); and, finally, that they eat
plenty of foods with calcium, a habit that keeps their nails
strong and healthy. Overall, the feedback referred to the five
pieces of information participants provided.
In comparison, participants assigned to receive negative
feedback read the following: that they could improve their
TELL ME WHAT I DID WRONG 000
FIGURE 3
WILLINGNESS TO PAY FOR A MANICURE AS A FUNCTION OF EXPERTISE LEVEL AND FEEDBACK RECEIVED (STUDY 3)
nail-care habits by drinking more unflavored and noncaffein-
ated water, which would help keep their nails strong and
healthy; that they could improve their habits by moisturizing
their hands more often to help maintain their skin’s elasticity;
that having short nails is potentially a sign that their nails are
not strong and healthy (if they indicated they had short fin-
gernails or toenails), or that having longer nails makes their
nails more prone to breaking and chipping (if they indicated
they had medium-length to long fingernails or toenails); and,
finally, that they could improve their habits by eating more
foods with calcium, which would help reduce the likelihood
that their nails would break or chip. Thus participants received
five pieces of negative feedback on their nail-care actions.
Our key dependent variable was participants’ interest in
caring for their nails. To measure participants’ interest,
they indicated their willingness to pay for “a professional
manicure at a nail salon or beauty parlor.” We presented
an open-ended question for willingness to pay (no price
range). As a manipulation check, participants then rated
how personalized (1 pnot at all personalized, 7 phighly
personalized) and reliable (1 pnot at all reliable, 7 p
very reliable) the feedback was. Participants were then
debriefed and dismissed. In their debriefings, none of the
participants expressed suspicion that the feedback was not
individually tailored.
Results and Discussion. In support of our manipulation,
participants perceived positive and negative feedback as
equally reliable (M
positive
p5.46, SD p2.98, vs. M
negative
p
4.72, SD p3.06; t(51) p1.26, NS) and personalized
(M
positive
p5.11, SD p3.98, vs. M
negative
p4.39, SD p
4.43; t(51) p1.10, NS). All four means were significantly
higher from the midpoint of the scale (4; all t12.5, all p
1.01), indicating that the feedback, overall, was personal-
ized and reliable.
In support of our hypothesis, the ANOVA on participants’
willingness to pay for a manicure yielded the predicted in-
teraction (F(1, 49) p9.20, p!.01) and no main effects.
Contrast analysis revealed that experts who received neg-
ative feedback expressed a higher willingness to pay for a
manicure (Mp$19.77, SD p$8.21) than novices who
received the same feedback (Mp$12.47, SD p$8.40;
t(24) p2.21, p!.04). We found the opposite pattern for
participants who received positive feedback: novices ex-
pressed a higher willingness to pay for a manicure (Mp
$17.47, SD p$9.05) than experts (Mp$11.67, SD p
$3.89; t(25) p2.24, p!.04; see fig. 3).
As in study 3A, we also had information on individual
differences in expertise from the open-ended questions in-
cluded in the expertise manipulation. We averaged these
individual-difference variables (ap.68) and ran a regres-
sion on participants’ willingness to pay for a manicure. The
regression revealed a main effect of actual expertise (bp
.68; tp4.85, p!.001), indicating that the more actual
expertise women had with nail care, the more they were
willing to pay for a manicure, as well as a main effect for
feedback (bp9.00; tp9.81, p!.001), indicating that
participants were more willing to pay for a manicure after
getting negative feedback. Finally, the analysis yielded the
predicted interaction (bp.37; tp3.93, p!.001), indi-
cating that experts’ greater willingness to pay for a manicure
more often followed receiving negative feedback than pos-
itive feedback.
Taken together, the findings from study 3 demonstrate
that those who perceive themselves as experts seek more
negative feedback than novices. In addition, upon receiving
negative feedback, experts respond more favorably than
novices by investing resources (i.e., willingness to pay) in
000 JOURNAL OF CONSUMER RESEARCH
pursuit of the goal. In contrast, novices respond more than
experts to positive feedback.
We argue that differences in the information that novices
and experts seek underlie the shift toward negative feedback.
Novices focus on assessing commitment, of which positive
feedback is a stronger signal, whereas experts focus on mon-
itoring progress, of which negative feedback is a greater
signal. We have yet to show that expertise changesthe mean-
ing individuals derive from feedback on their goal pursuit.
Accordingly, in study 4, we explored how expertise moves
people from assessing commitment to monitoring progress.
Another goal of study 4 was to explore how people’s interest
in negative feedback increases as they gain expertise with
a task.
STUDY 4: LEARNING A NEW TASK
In study 4, we tracked native-English-speaking participants’
interest in negative feedback as they gained experience in
learning to type in German. We operationalized expertise as
level of training such that it increased as participants pro-
gressed on the learning task: they were novices on the first
trial of the task and relative experts on the last trial. We
predicted an increase in participants’ likelihood of seeking
negative feedback as they advanced on the task.
As in study 3, we divided study 4 into two parts. The
first part tested for feedback seeking. The second part tested
how the meaning of feedback changes as a function of po-
sition in the task. We predicted that novices would be more
likely than experts to infer from positive feedback that their
learning goal was important or valuable (i.e., commitment),
whereas experts would be more likely than novices to infer
from negative feedback that they should increase their ef-
forts (i.e., progress).
Study 4A: Meaning of Feedback over Time
Method. Twenty-six undergraduate students participated
for monetary compensation. This study utilized a 6 (exper-
tise: trial number 1 to trial number 6) within-subjects design.
We recruited participants who had no prior experience
speaking or writing in German to complete a study on how
people learn an unfamiliar task: typing in German. Because
participants were college students, adopting a goal of learn-
ing a new cognitive skill was relatively easy for them.
Participants read that the study assessed people’s ability
to learn a new skill that required cognitive flexibility and
that they would be “typing texts taken from popular German
authors like Rilke and Goethe as well as songs from famous
artists like the Beatles written in German.” Participants read
that they would see text appear on the top portion of the
computer screen and that their task was to duplicate the text
in the space provided in the bottom portion of the screen
—which was left blank with a blinking cursor—in the time
allotted to complete the passage.
Next, participants learned that their performance on the task
would be measured by how quickly they typed the passage
and the accuracy of their typing as measured by the match
between what they typed and the words in the passage. They
further read that, as with many learning tasks, they would have
a chance to choose what individually-tailored feedback they
would like to receive about their performance at different points
in the study, specifically, before moving to the next trial.
We piloted the task to be fun yet moderately challenging.
For example, participants typed the song “I Want to Hold
Your Hand” from the Beatles or a passage from The Sorrows
of Young Werther by Johann Wolfgang von Goethe. They
completed six trials in total. We randomized the order of
the trials across participants to ensure a specific passage did
not drive a participant’s propensity of seeking positive ver-
sus negative feedback. Participants had 30 seconds to com-
plete each typing task. Once the time had passed, the pro-
gram automatically moved to the next screen.
After each trial, participants read: “Now that you have
finished the (number, e.g., “first”) trial, what feedback would
you like to receive on your performance? You can only pick
one piece of information so please choose what you would
most like to know.” Participants chose between receiving
positive feedback about what they had done well or negative
feedback about how they could improve. The item of interest
was the type of feedback, positive or negative, participants
chose as they progressed through the trials and became more
experienced with the task.
Unbeknownst to participants, the content of the negative
or positive feedback they received on each trial was pre-
determined and appeared equally informative. For example,
participants who chose to receive positive feedback on trial
1 read: “After analyzing your response, it appears that you
have good finger placement and that you do a good job
ignoring how you think words should be spelled. This good
finger placement helps your speed and accuracy.” In com-
parison, participants who received negative feedback read:
“After analyzing your response, it appears that you focus
too much on how you think words should be spelled and
that your accuracy is hindered when you add extra letters
to words. You can improve your accuracy by watching your
finger placement.”
Results and Discussion. We coded a participant’s choice
of feedback as a binary variable (1 pchose negative feed-
back, 0 pchose positive feedback). In accordance with the
hypothesis, a binary logistic regression on choice of feed-
back revealed a linear trend indicating that participants were
more likely to seek negative feedback as they progressed
through the trials (bp.21; Wald x
2
(1) p21.62, p!.01;
see fig. 4). Specifically, whereas only 50% of participants
sought negative feedback after their first trial, 74% sought
negative feedback after their second trial, 63% sought neg-
ative feedback after their third trial, 67% sought negative
feedback after their fourth trial, 71% sought negative feed-
back after their fifth trial, and 82% sought negative feedback
after the last trial.
These results are consistent with the hypothesis that as
people gain expertise, they switch from seeking positive
feedback to seeking negative feedback. Confirming that ex-
TELL ME WHAT I DID WRONG 000
FIGURE 4
PERCENTAGE OF PARTICIPANTS WHO SOUGHT
NEGATIVE FEEDBACK AS A FUNCTION OF
EXPERTISE LEVEL (STUDY 4)
perts seek more negative and less positive feedback, we next
tested for the meaning contained in feedback.
Study 4B: Meaning of Feedback
We conducted study 4B to test whether the same feedback
on one’s performance conveys different information for nov-
ices and experts, thus serving different motivational func-
tions. We predicted that novices focus on assessing their
commitment and positive feedback motivates them because
it signals high commitment but that experts focus on mon-
itoring their progress and negative feedback motivates them
because it signals insufficient progress toward their goal.
Method. Two hundred and thirty-two participants (116
women) completed a typing task similar to the one in the
main study but with a few minor adjustments. The study
employed a 2 (expertise: novice vs. expert) #2 (feedback:
positive vs. negative) #2 (meaning: commitment vs. pro-
gress) between-subjects design.
To ensure that participants had enough experience in the
task, they completed 15 typing trials. The rest of the task,
including the feedback content, was similar to that of study
4A except that participants did not seek feedback. We ran-
domly assigned participants to receive either positive or neg-
ative feedback after they completed the first trial, at which
point they felt like relative novices, or before their last trial,
at which point they felt like relative experts.
After receiving their feedback, participants rated either
their feelings of making progress toward the goal (1 p“I
feel like I have made sufficient progress on the task,” 7 p
“I feel like I have made insufficient progress on the task”;
progress inference) or how much they cared about doing
well on the task (1 p“I care about my typing skills on this
task very little,” 7 p“I care about my typing skills on this
task very much”; commitment inference). We followed pre-
vious research (Fishbach and Dhar 2005) in wording these
items. Our progress question captures a sense of making
insufficient progress, whereas our commitment question cap-
tures the value (i.e., caring) component of commitment (see
Value #Expectancy model, Fishbein and Ajzen 1975). In
the context of this task, both a sense of insufficient progress
and a sense of caring about performing well reflect greater
motivation to pursue the task at hand but for different rea-
sons. After providing their answer, participants completed
another trial (trial 2 for novices; trial 15, the last trial, for
experts). Finally, participants were debriefed and dismissed.
In their debriefings, none of the participants expressed sus-
picion that the feedback was not individually tailored.
Results and Discussion. The ANOVA on ratings of
meaning in feedback yielded the predicted expertise (novice
vs. expert) #feedback (positive vs. negative) #meaning
(asked about commitment vs. progress) three-way interac-
tion (F(1, 225) p3.98, p!.05) and no main effects (see
fig. 5). Specifically, novices were more likely than experts
to indicate that positive feedback signaled that they care
about their typing skills (i.e., commitment, Mp4.67, SD
p1.06, vs. Mp3.68, SD p1.95; t(56) p2.42, p!.02).
On the other hand, novices were not more likely than experts
to infer that they were committed when they received neg-
ative feedback (Mp4.50, SD p1.53, vs. Mp4.53, SD
p1.61; t(58) !1, NS). Additionally, experts were more
likely than novices to indicate that they made insufficient
progress when they received negative feedback (Mp4.93,
SD p1.16, vs. Mp3.96, SD p1.58; t(55) p2.64, p!
.02). However, experts were not more likely than novices
to infer that they had made insufficient progress when they
received positive feedback (Mp4.55, SD p1.67, vs. M
p4.15, SD p1.41; t(56) !1, NS).
We predicted that giving novices (those at their first trial)
positive feedback would increase their performance moti-
vation more than giving them negative feedback. In addition,
giving experts (those at their fourteenth trial) negative feed-
back would increase their performance motivation morethan
giving them positive feedback. Since we held time constant
at 30 seconds per trial, we coded the number of words each
participant accurately typed on the trial that followed the
feedback (two vs. 15) as our measure of performance mo-
tivation.
Analysis of the number of words participants typed on
trial 2 revealed that novices exhibited better performance
when they received positive feedback (Mp19.05 words,
SD p3.72) than when they received negative feedback (M
p17.32 words, SD p4.04; t(111) p2.38, p!.02). Ad-
ditionally, on trial 15, experts performed better when they
received negative feedback on the previous trial (Mp
16.16, SD p4.53) compared with experts who received
positive feedback on the previous trial (Mp14.47, SD p
4.05; t(114) p2.12, p!.04). We did not observe better
000 JOURNAL OF CONSUMER RESEARCH
FIGURE 5
INFERENCES ABOUT PROGRESS ON AND COMMITMENT TO LEARNING TO TYPE IN GERMAN
AS A FUNCTION OF FEEDBACK AND EXPERTISE LEVEL (STUDY 4)
performance among experts versus novices, possibly due to
a general depletion of resources as participants progressed.
Indeed, we find that they typed more words on trial 2 than
on trial 15. Nonetheless, the comparison between positive
and negative feedback implies that positive feedback in-
creases motivation initially and that negative feedback in-
creases motivation subsequently.
The findings from study 4 further support our hypothesis
regarding the link between expertise and seeking negative
feedback (study 4A) and between expertise and responding
to the negative feedback (study 4B), this time by following
these trends as people gain expertise on a task. Moreover,
in study 4B, we find that the same positive and negative
feedback mean different things for those who start a task
versus those who are about to finish it. Novices infer com-
mitment more than experts, whereas experts infer a need
for progress more than novices. Interestingly, because ex-
pertise is subjective and also relative, to the extent that
people perceive themselves as approaching the end of a task,
they feel as if they are experts and hence seek negative
feedback on the task after only engaging in it for a few
trials. We can contrast these findings with those of study 1,
where novices have been studying French for a longer time
than this study’s “experts,” and conclude that expertise as
a frame of mind, more than actual knowledge and a set of
skills, influences the shift to negative feedback.
Taken together, in our studies thus far, participants sought
and responded to feedback on their own actions. A related
question is whether people respond in a similar way to feed-
back on shared goals—that is, goals a group of individuals
pursues together. Accordingly, in our final study, we asked
how people respond to feedback on the effectiveness of their
community’s recycling program. By examining shared
goals, we not only extended our investigation to feedback
that is not self-threatening and is less likely to invoke de-
fensive processes (e.g., negative feedback on the community
recycling program is less offensive than negative feedback
on one’s own attempts) but we could further test for feed-
back that is presented as part of a persuasive appeal. We
predicted that experts would be more responsive than nov-
ices to persuasive appeals that emphasize negative aspects
of the present situation.
STUDY 5: FEEDBACK ON SHARED
GOALS: RECYCLING PROGRAMS
Individuals often pursue goals together. For instance, when
people recycle, they understand that other community mem-
bers must also recycle if they are to make a significant dent
in helping the environment. We refer to these types of goals
as shared goals. We capitalized on this unique property of
shared goals to provide participants with feedback on en-
vironmental actions that their city, rather a single person,
performs.
In this study, we defined expertise as knowledge about
environmental issues, and we manipulated participants’ per-
TELL ME WHAT I DID WRONG 000
ceived expertise by asking them easy versus difficult knowl-
edge questions. Participants then read a media message em-
phasizing that their city’s recycling program was highly
effective (positive feedback) or highly ineffective (negative
feedback). They then indicated their attitudes toward the
messages. We predicted that those who perceived they were
knowledgeable about environmental issues (experts) would
exhibit more favorable attitudes toward the negative mes-
sage than those who perceived they were less knowledgeable
(novices). In addition, we predicted that those who perceived
they were novices would exhibit more favorable attitudes
toward the positive media message than experts.
Method
Fifty-two people (19 women) at a downtown laboratory
in a Midwestern city participated in the study for monetary
compensation. This study employed a 2 (expertise: novice
vs. expert) #2 (feedback: positive vs. negative) between-
subjects design.
Participants completed a study on political issues, which
presumably assessed the importance people place on learn-
ing about (a) the economy and (b) the environment. Their
first task was to answer a series of questions on these topics.
All participants indicated their familiarity with 10 events,
five on the economy and five on the environment. Those
assigned to the novice condition were made to feel relatively
less knowledgeable about environmental issues and rela-
tively more knowledgeable about economic issues. In this
regard, we influenced perceived expertise by asking novices
if they were familiar with five environmental issues that
were not highly publicized: the 2009 Istanbul floods, the
2008 Chinese winter storms, the 2009 fires in Australia, the
2009 typhoon in China, and the 2009 cyclone in Myanmar.
On the other hand, novices were asked about five economic
issues they had likely heard of: rising national debt in the
United States, increasing unemployment in the United
States, government bailouts in the United States, the costs
of health care reform, and the 2009 economic summit in
Pennsylvania.
In contrast, those assigned to the expert condition were
made to feel relatively more knowledgeable about environ-
mental issues and less knowledgeable about economic is-
sues. Experts indicated their familiarity with five well-
known environmental issues: Hurricane Katrina, the Kyoto
Protocol, the 2004 tsunami in Thailand, the 2009 California
wildfires, and melting Arctic icecaps. They also indicated
their familiarity with three economic issues that were not
as well publicized—the 2009 economic crisis in Mexico,
the 2009 economic boom in Qatar, and China’s decreased
lending—and two economic issues that were more publi-
cized (from the previous condition)—the costs of health care
reform and the 2009 economic summit in Pennsylvania. We
included the latter two economic issues that participants had
more likely heard of in the environmental-expert condition
based on our pretesting, in which participants were generally
less familiar with economic issues. We wanted to ensure
that participants could recognize a similar number of events
across the expertise conditions and thus feel equally com-
petent. In this way, we manipulated perceived knowledge
of environmental issues without negatively affecting partic-
ipants’ views of themselves as knowledgeable people.
Next, the experimenter moved participants to a new room
to complete a presumably unrelated study. Participants read
that the researchers were interested in how people think
about newsletters written by journalism students from a local
college. Participants assigned to read positive feedback read
a newsletter entitled “City Recycling Program Is an Envi-
ronmental Panacea” that emphasized that another agent,
their city, utilizes a highly effective recycling program. Spe-
cifically, participants read that their city’s recycling program
costs less to operate than waste collection does and was thus
highly cost-effective. Further, they read that, in addition to
reducing waste, the city saved money by getting rid of du-
plicate pick-up routes, due to its history of being a city that
recycles. In contrast, those assigned to read negative feed-
back were given an article entitled “City Recycling Dumped
in Landfills” that emphasized that another agent, their city,
has a highly ineffective recycling program. Specifically, par-
ticipants read that their city paid exorbitant costs to get rid
of recyclables or simply had city employees dump recycl-
ables in public landfills. Further, they read that those at City
Hall claimed that determining what plastics are easily versus
not easily recycled is too complex and that the public needs
to express how much it values recycling to stop the city’s
bad actions.
Upon completion of the article, participants rated it on a
several dimensions. The variables of interest pertained to
participants’ attitudes toward the article. They rated how
persuasive the article was and how useful, convincing, and
diagnostic the information in it was (for all four items, 1 p
not at all persuasive/useful/convincing/diagnostic, 7 pvery
persuasive/useful/convincing/diagnostic).
Results and Discussion
We counted the number of economic and environmental
issues with which the participants were familiar. In support
of the manipulation, participants who were made to feel like
novices had heard of fewer environmental issues; thus they
were less familiar with them (Mp1.89, SD p1.20) than
were those who were made to feel like experts (Mp4.07,
SD p.79; t(50) p8.47, p!.001). In further support of
the manipulation, those who were made to feel like experts
on environmental issues reported knowing of fewer eco-
nomic issues (Mp2.73, SD p1.11) than those who were
made to feel like novices on environmental issues and ex-
perts on economic issues (Mp4.11, SD p.47; t(50) p
6.69, p!.001). Thus the manipulation did not influence
overall competence across both sets of items, as participants
indicated that they had heard of roughly the same amount
of issues, yet, as intended, half of the participants were made
to feel they were relatively knowledgeable about the envi-
ronment.
To further ensure that our expertise manipulation did not
influence mood, we sampled participants from the same sub-
000 JOURNAL OF CONSUMER RESEARCH
FIGURE 6
ATTITUDES TOWARD POSITIVE AND NEGATIVE
MESSAGE AS A FUNCTION OF EXPERTISE
(STUDY 5)
ject population (Np40) to complete the expertise manip-
ulation and then rate their mood on the positive and negative
PANAS scales (Watson, Clark and Tellegen 1988). Those
made to feel like novices on environmental issues (and ex-
perts on economic issues) and those made to feel like experts
on environmental issues (and novices on economic issues)
reported similar levels of positive mood (M
Novices
p1.72,
SD p.69, vs. M
Experts
p1.75, SD p.54; t(38) !1, NS)
and similar levels of negative mood (M
Novices
p1.73, SD p
.69, vs. M
Experts
p1.75, SD p.54; t(38) !1, NS).
To test the main hypothesis, we assessed participants’
attitudes toward the messages by collapsing the four ques-
tions that measured evaluation (ap.82). An ANOVA of
attitudes ratings revealed the predicted expertise #feedback
interaction (F(1, 48) p19.49, p!.001) and no main effects.
Contrast analysis revealed that, among those who read the
media message emphasizing the positive aspects of their
city’s recycling program, those who perceived themselves
as novices exhibited more favorable attitudes toward the
message (Mp5.50, SD p.91) than those who perceived
themselves as experts (Mp3.80, SD p1.32; t(19) p
3.47, p!.01). In contrast, among those who read the media
message emphasizing the negative aspects of their city’s
recycling program, those who perceived themselves as ex-
perts exhibited more favorable attitudes toward the message
(Mp5.56, SD p.83) than novices (Mp4.68, SD p
1.02; t(29) p2.50, p!.02, see fig. 6).
In study 5, we find support for our proposition that ex-
pertise affects the response to feedback even if the feedback
is targeted toward group members pursuing a shared goal
rather than toward an individual. We demonstrate that nov-
ices exhibit more favorable attitudes than experts toward a
media message emphasizing positive feedback, whereas ex-
perts exhibit more favorable attitudes than novices toward
a media message emphasizing negative feedback.
GENERAL DISCUSSION
This article investigates the feedback individuals seek as
well as how they respond to that feedback by changing their
attitudes and behaviors. We predict an increase in negative
feedback as people gain expertise, because the meaning peo-
ple derive from feedback changes such that negative feed-
back increases the motivation to adhere to a goal. In support
of our prediction, we find that novices infer from feedback
whether their goals are valuable (commitment), whereas ex-
perts infer from feedback whether their pace of pursuing
already valuable goals is sufficient (progress). As a conse-
quence of the information in feedback, novices are more
likely than experts to seek positive feedback on their
strengths and alter their behaviors and attitudes when they
get such feedback, whereas experts are more likely than
novices to seek negative feedback on their weaknesses and
alter their behaviors and attitudes when they get this feed-
back.
Results from five studies support these hypotheses. In
studies 1 and 2, we measured expertise and showed that it
was associated with seeking more negative feedback on
one’s performance in a language class (study 1) and recy-
cling habits (study 2). In study 2, we further demonstrated
that novices respond more than experts to positive feedback
by donating to an environmental charity, whereas expert
environmentalists respond more than novices to negative
feedback by increasing their donations. In study 3, we ma-
nipulated expertise: women who felt like nail-care experts
sought more negative feedback on their nail-care habits and
responded more to this feedback by expressing a higher
willingness to pay for manicures than women who felt like
novices. In comparison, women who perceived themselves
as novices responded more to positive feedback. In study
4, we examined how people seek an increasing amount of
negative feedback as they progress on a learning task and
demonstrated that, indeed, novices seek and respond more
to positive feedback because it affirms their commitment to
a goal, whereas experts seek and respond more to negative
feedback because it signals that they have made insufficient
progress and have not invested enough effort toward their
goals. Finally, in study 5, we demonstrated that expertise
influences people’s responses (in particular, their attitudes)
to persuasive appeals that emphasize successful versus un-
successful pursuit of a shared environmental goal. Thus,
negative messages on the city’s ineffective recycling actions
were more persuasive for experts than novices, whereas pos-
itive messages on the city’s effective actions were more
persuasive for novices than experts.
Interestingly, across these studies, we find that people are
generally interested in negative feedback (e.g., in study 3,
100% of those made to feel like experts in caring for their
nails and 74% of those made to feel like novices sought
negative feedback). Thus, whereas some previous research
portrays people as negative-feedback avoiders (Russo et al.
TELL ME WHAT I DID WRONG 000
1998; Tormala and Petty 2004), we identifyconditions under
which they seek and endorse negative feedback. In partic-
ular, negative feedback seems to serve an important function
when it is constructive (rather than detrimental) and when
people desire to acquire new habits or improve existing ones
(rather than enhance their self-image).
Motives Underlying Feedback Seeking
The present research addresses situations in which people
look for feedback to motivate themselves to pursue their
goals and, under these circumstances, we find that experts
seek more negative feedback than novices. Although mo-
tivating oneself is a common and possibly the dominant
motive in feedback seeking, at times people might hold other
motivations for feedback seeking. For instance, people
might want to receive self-enhancing feedback (Tesser
1988), in which case they will prefer positive feedback re-
gardless of their expertise. For example, we would predict
that both experts and novices will seek positive feedback
when in a negative mood as a means for mood improvement.
At other times, people might seek confirming feedback,
which reaffirms their self-view (Swann and Read 1981). If
self-affirmation underlies feedback seeking, we could expect
experts to seek more positive feedback than novices on their
ability to pursue their goal, but experts might seek more
negative feedback than novices on goal attainment because
such feedback will affirm the novices’ greater optimism
about goal attainment. Moreover, people might also look for
feedback to justify goal disengagement, that is, as an excuse
to get out of pursuing a certain goal. In these situations, we
would expect experts to seek positive feedback (signaling
they have put in enough effort) and novices to seek negative
feedback (signaling that the action is not worth doing) be-
cause this feedback would undermine task motivation.
Thinking about this latter motivation to get out of pur-
suing a goal further helps us illuminate the distinction be-
tween expertise and commitment. We propose that expertise
increases commitment but that expertise is not commitment.
Whereas expertise, by definition, increases with experience,
goal commitment often increases but can at times decrease
or remain stable. And whereas committed people, by defi-
nition, desire to continue pursuing a goal, experts might at
times look for reasons to disengage from goal pursuit. For
example, an expert French speaker might look for reasons
to slack off in a required French language class and thus
look for positive feedback suggesting that she has suffi-
ciently progressed in learning the materials, whereas ahighly
committed French major will look for ways to improve her
French and thus look for negative feedback suggesting that
she can improve.
The Subjective Nature of Expertise
Our findings suggest that researchers should think of ex-
pertise as a subjective experience that fluctuates depending
on the context and salient social comparison standards. For
example, women in study 3 felt less experienced with their
nail-care habits when they compared themselves to someone
doing her nails on a weekly basis than when they compared
themselves to someone doing her nails once every 2 years.
Whereas previous research identified the various dimensions
of consumer expertise, including the frequency of perform-
ing actions (Bettman and Park 1980) and prior knowledge
(Hong and Sternthal 2010), we find that the impact of each
of these variables depends on the subjective experience of
the consumer. Whether a person feels that she performs
actions more than another person or feels that she is more
knowledgeable than a salient comparison standard will de-
termine her perception of herself as an expert and will further
influence how she responds to feedback.
Naturally, other dimensions of consumer expertise exist
beyond what we explored in this research. For example, one
such dimension of expertise might be power. Receiving feed-
back from a person in a position of higher power might
make one question his own commitment and feel like a
relative novice. For example, patients who receive advice
from their doctors, who are perceived as higher in power,
might be reminded that they have a lot to learn about diet
and exercise and thus, compared to their doctors, they will
feel like novices on health-related issues. In this instance,
we would predict that the novice (patient) will be more likely
to focus on assessing her commitment and thus will seek
and respond more to positive feedback than will a person
who has more power (the doctor). We would further predict
that, as with other dimensions of expertise, the subjective
experience of power (or powerlessness) influences the ex-
perience of expertise and thus the feedback an individual
seeks.
Although power may indeed influence one’s sense of ex-
pertise, this variable cannot explain the patterns we observed
in our studies because the power of the feedback giver was
held constant in each study. For example, regardless of par-
ticipants’ expertise in study 1, they always sought feedback
from an instructor who had more power than they did, or
as participants progressed through the task in study 4, they
constantly sought feedback from the computer, which was
not an entity with whom they had a power relationship.
One important implication of our finding that expertise is
subjective is that marketers can make the recipients of per-
suasive experts feel relatively experienced or inexperienced
and then tune their feedback to the induced subjective ex-
pertise. Moreover, inducing a sense of expertise among mes-
sage recipients could be particularly useful when marketers
cannot frame a message positively, as when a person’s per-
formance is particularly bad and negative feedback would
hurt her self-esteem—for example, when targeting individuals
who overeat or fail to save.
Marketing and Policy Implications
The present findings have further implications for mar-
keters and members of the media attempting to persuade
people to see their points of view. Our findings attest that
media messages emphasizing positive feedback have greater
impact on novices than experts, whereas media messages
000 JOURNAL OF CONSUMER RESEARCH
emphasizing negative feedback have greater impact on ex-
perts. Thus the present findings add to the already large
body of research exploring when positive versus negative
feedback is more effective (see, e.g., Ahluwalia et al. 2000;
Maheswaran and Myers-Levy 1990) by suggesting that,
when consumers focus on realistically assessing their skills,
negative feedback can, in fact, alter attitudes and behaviors.
Conceivably, then, companies that desire to have consumers
engage more with their product might want to target new
users of their products by telling them how well they already
utilize their sophisticated products and target experienced
users by telling them how they can improve their usage of
such sophisticated products.
Additionally, the current findings have specific implica-
tions for marketers of learning and skill-acquisition prod-
ucts. For instance, our findings suggest that marketers should
design their feedback with a consumer’s expertise level in
mind. To illustrate, health clubs should instruct their trainers
to give positive feedback about the things new clients do
well (e.g., that they have good form on a particular exercise)
and focus on negative feedback about the areas clients can
improve (e.g., they can improve their form on a particular
exercise) when interacting with experienced clients. Simi-
larly, weight-loss programs should emphasize that new at-
tendees have done a nice job monitoring their diet over the
course of the week and that this monitoring will help them
lose weight, but the programs should emphasize that fre-
quent attendees can monitor their diets a bit more closely
if they would like to lose weight.
Finally, these findings have implications for how mar-
keters, as well as educators and social agents, can help en-
courage people to adhere to the goals they set for themselves.
In general, marketers can be more effective in the feedback
they provide by accounting for a person’s level of expertise
in pursuing performance goals. For instance, companies that
offer products designed to aid in skill acquisition should
account for their customers’ sense of expertise and, accord-
ingly, provide feedback that increases their motivation. One
caveat to this recommendation is that consumers should fo-
cus on improving and learning while they acquire new skills
rather than on seeking self-enhancing feedback; otherwise,
negative feedback could be detrimental to their performance.
REFERENCES
Ahluwalia, Rohini, Robert E. Burnkrant, and H. Rao Unnava
(2000), “Consumer Response to Negative Publicity: The
Moderating Role of Commitment,” Journal of Marketing Re-
search, 37 (May), 203–14.
Alba, Joseph W., and J. Wesley Hutchinson (1987), “Dimensions
of Consumer Expertise,” Journal of Consumer Research,13
(March), 411–52.
Ashford, Susan J., Ruth Blatt, and Don Vande Walle (2003), “Re-
flections on the Looking Glass: A Review of Research on
Feedback-Seeking Behavior in Organizations,” Journal of
Management, 29, (6), 773–99.
Ashford, Susan J., and Anne S. Tsui (1991), “Self-Regulation for
Managerial Effectiveness: The Role of Active Feedback Seek-
ing,” Academy of Management Journal, 34 (2), 251–80.
Bandura, Albert (1991), “Self-Regulation of Motivation through
Anticipatory and Self-Reactive Mechanisms,” in Nebraska
Symposium on Motivation, ed. R. A. Dienstbier, Lincoln: Uni-
versity of Nebraska Press, 69–164.
Bettman, James R., and C. Whan Park (1980), “Effects of Prior
Knowledge and Experience and Phase of the Choice Process
on Consumer Decision Processes: A Protocol Analysis,” Jour-
nal of Consumer Research, 7 (December), 234–48.
Brown, Roger, and Camille Hanlon (1970), “Derivational Com-
plexity and Order of Acquisition in Child Speech,” in First
Language Acquisition: The Essential Readings, ed. Barbara
Lust and Claire Foley, Malden, MA: Blackwell, 155–75.
Carver, Charles S., and Michael F. Scheier (1998), On the Self-
Regulation of Behavior, New York: Cambridge University
Press.
Dweck, Carol Sorich, and Ellen L. Leggett (1988), “A Social Cog-
nitive Approach to Motivation and Personality,” Psycholog-
ical Review , 95 (April), 256–73.
Feather, Norman T. (1982), “Actions in Relation to Expected Con-
sequences: An Overview of a Research Program,” in Expec-
tations and Actions: Expectancy-Value Models in Psychology,
ed. Norman T. Feather, Hillsdale, NJ: Erlbaum, 53–95.
Fishbach, Ayelet, and Ravi Dhar (2005), “Goals as Excuses or
Guides: The Liberating Effect of Perceived Goal Progress on
Choice,” Journal of Consumer Research, 32 (December),
370–77.
Fishbach, Ayelet, Ying Zhang, and Minjung Koo (2009), “The
Dynamics of Self-Regulation,” European Review of Social
Psychology, 20 (October), 315–44.
Fishbein, Martin, and Icek Ajzen (1975), Belief, Attitude, Intention,
and Behavior: An Introduction to Theory and Research, Read-
ing, MA: Addison-Wesley.
Fo¨rster, Jens, Nora Liberman, and E. Tory Higgins (2005), “Ac-
cessibility from Active and Fulfilled Goals,” Journal of Ex-
perimental Social Psychology, 41 (May), 220–39.
Frey, Karin S., and Diane N. Ruble (1987), “What Children Say
about Classroom Performance: Sex and Grade Differences in
Perceived Competence,” Child Development, 58 (August),
1066–78.
Higgins, E. Tory (1987), “Self-Discrepancy: A Theory Relating
Self and Affect,” Psychological Review, 94 (July), 319–40.
Hong, Jiewen, and Brian Sternthal (2010), “The Effects of Con-
sumer Prior Knowledge and Processing Strategies on Judg-
ments,” Journal of Marketing Research, 47 (April), 301–11.
Hutchinson, J. Wesley (1983), “Expertise and the Structure of Free
Recall,” in Advances in Consumer Research, Vol. 10, ed.
Richard P Bagozzi and Alice M. Tybout, Ann Arbor, MI:
Association for Consumer Research, 585–89.
Kiel, Geoffrey C., and Roger A. Layton (1981), “Dimensions of
Consumer Information Seeking,” Journal of Marketing Re-
search, 18 (May), 233–39.
Koo, Minjung, and Ayelet Fishbach (2008), “Dynamics of Self-
Regulation: How (Un)Accomplished Goal Actions Affect
Motivation,” Journal of Personality and Social Psychology,
194 (2), 83–95.
Kruger, Justin, and David Dunning (1999), “Unskilled and Una-
ware of It: How Difficulties in Recognizing One’s Own In-
competence Can Lead to Inflated Self-Assessments,” Journal
of Personality and Social Psychology, 77 (December),
1121–34.
Kruglanski, Arie W. (1990), “Motivations for Judging and Know-
ing: Implications for Causal Attribution,” in Handbook of
Motivation and Cognition: Foundations of Social Behavior,
TELL ME WHAT I DID WRONG 000
Vol. 2, ed. E. Tory Higgins and Robert M Sorrentino, New
York: Guilford, 53–92.
Linville, Patti W. (1987), “Self-Complexity as a Cognitive Buffer
against Stress-Related Illness and Depression,” Journal of
Personality and Social Psychology, 52 (April), 663–76.
Locke, Edwin A., and Gary P. Latham (1990), A Theory of Goal
Setting and Task Performance, Upper Saddle River, NJ: Pren-
tice Hall.
Louro, Maria J., Rik Pieters, and Marcel Zeelenberg (2007), “Dy-
namics of Multiple-Goal Pursuit,” Journal of Personality and
Social Psychology, 2 (August), 174–93.
Maheswaran, Durairaj, and Joan Meyers-Levy (1990), “The Influ-
ences of Message Framing and Issue Involvement,” Journal
of Marketing Research, 27 (August), 361–67.
“Market Trends: Nail Care Products” (2010), Packagedfacts.com,
http://www.packagedfacts.com.
Miller, Dale T., and Michael Ross (1975), “Self-Serving Biases in
Attribution of Causality: Fact or Fiction?” Psychological Bul-
letin, 82 (March), 213–25.
Miller, George A., Eugene Galanter, and Karl H Pribram (1960),
Plans and the Structure of Behavior, New York: Holt.
Raghunathan, Rajagopal, and Yaacov Trope (2002), “Walking the
Tightrope between Feeling Good and Being Accurate: Mood
as a Resource in Processing Persuasive Messages,” Journal
of Personality and Social Psychology, 83 (September),
510–25.
Russo, J. Edward, Margaret G. Meloy, and Victoria Husted Medvec
(1998), “Predecisional Distortion of Product Information,”
Journal of Marketing Research, 35 (November), 438–52.
Ryan, Richard M., and Edward L. Deci (2000), “Self-Determi-
nation Theory and the Facilitation of Intrinsic Motivation,
Social Development, and Well-Being,” American Psycholo-
gist, 55 (January), 68–78.
Schwarz, Norbert, Hans J. Hippler, Brigitte Deutsch, and Fritz
Strack (1985), “Response Scales: Effects of Category Range
on Reported Behavior and Comparative Judgments,” Public
Opinion Quarterly, 49 (Fall), 388–95.
Swann, William B., and Stephen J. Read (1981), “Self-Verification
Processes: How We Sustain Our Self Conceptions,” Journal
of Experimental Social Psychology, 17 (July), 351–72.
Tesser, Abraham (1988), “Toward a Self-Evaluation Maintenance
Model of Social Behavior,” in Advances in Experimental So-
cial Psychology, Vol. 21, ed. Leonard Berkowitz, New York:
Academic Press, 181–227.
Tormala, Zakary L., and Richard E. Petty (2004), “Source Cred-
ibility and Attitude Certainty: A Metacognitive Analysis of
Resistance to Persuasion,” Journal of Consumer Psychology,
14 (September), 427–42.
Trope, Yaacov (1986), “Testing Self-Enhancement and Self-As-
sessment Theories of Achievement Motivation: A Reply,”
Motivation and Emotion, 10 (September), 247–61.
Trope, Yaacov, and Efrat Neter (1994), “Reconciling Competing
Motives in Self-Evaluation: The Role of Self-Control in Feed-
back Seeking,” Journal of Personality and Social Psychology,
66 (April), 646–57.
Vroom, Victor H. (1964), Work and Motivation, New York: Wiley.
Ward, Scott (1974), “Consumer Socialization,” Journal of Con-
sumer Research, 1 (September), 1–14.
Watson, David, Lee A. Clark, and Auke Tellegen (1988), “De-
velopment and Validation of Brief Measures of Positive and
Negative Affect: The PANAS Scales,” Journal of Personality
and Social Psychology, 54 (June), 1063–70.
Wood, Joanne V. (1989), “Theory and Research Concerning Social
Comparisons of Personal Attributes,” Psychological Bulletin,
106 (September), 231–48.
Wood, Wendy, Nancy Rhodes, and Micheal Biek (1995),“Working
Knowledge and Attitude Strength: An Information-Processing
Analysis,” in Attitude Strength: Antecedents and Conse-
quences, ed. Richard E. Petty and Jon A. Krosnick, Mahwah,
NJ: Erlbaum, 283–313.
