THE EXPERT’S CURSE: SHIFTING TO NEGATIVE FEEDBACK
STACEY R. FINKELSTEIN
AYELET FISHBACH
2
Stacey Finkelstein is a Doctoral Candidate and Ayelet Fishbach is a Professor of Behavioral
Science and Marketing at the University of Chicago, Booth School of Business. Correspondence
concerning this article may be addressed to Stacey R. Finkelstein or Ayelet Fishbach, The
University of Chicago, Booth School of Business, 5807 S. Woodlawn Ave, Chicago IL 60637.
Email: [email protected] or [email protected].
3
Companies focus a large proportion of their marketing communication on providing feedback to
and soliciting feedback from consumers. This article explores what type of feedback consumers
seek and give. We predict and find a shift from positive to negative feedback seeking and giving
as people gain expertise. We document this shift in foreign language classes, where advanced
students sought more negative and less positive feedback than beginners, and in a novel learning
task, where a larger proportion of learners sought negative (vs. positive) feedback as they
progressed on the task. A similar shift existed for feedback giving, as people provided more
negative feedback to more experienced team members. We examine a motivational account for
the shift in feedback: positive feedback increases novices’ goal commitment and negative
feedback increases experts’ sense of insufficient progress.
4
Feedback is essential for goal pursuit. Without it, individuals would not know whether
and how much to invest in pursuing their goals and what actions they should take (Ashford,
Blatt, and Van de Walle 2003; Frey and Ruble 1987; Kruglanski 1990; Swann and Read 1981;
Miller and Ross 1975; Wood 1989). Accordingly, many social roles are associated with
providing feedback: educators, coworkers, and coaches all provide feedback on individuals’
performance. In addition, a large proportion of marketing communication involves providing
feedback to and soliciting feedback from consumers. For example, language programs provide
feedback on consumers’ mastery of a foreign language, skin product manufactures inform
customers that they can improve their appearance by learning how to apply makeup a certain
way or by using a specific skin care regimen, and personal trainers give clients feedback on their
exercising and nutrition habits. Moreover, marketers often solicit feedback from consumers on
the effectiveness of their services or products, in which case consumers assume the role of
feedback providers (Ofir and Simonson 2001). For example, stores often invite costumers to rate
the quality of their purchases, and some websites specialize in soliciting ratings of services and
products from consumers to promote success in a competitive market (Peters and Waterman,
1982; Oliver, 1997). Thus understanding what feedback people seek and give is important.
Generally speaking, we distinguish between positive feedback on accomplishments,
strengths, and correct responses, and negative feedback on lack of accomplishments,
weaknesses, and incorrect responses. 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 corrective actions (see, e.g., Dweck and Leggett
1988). For example, positive feedback will emphasize a student’s correct use of grammar in a
5
foreign language class, whereas negative feedback will emphasize her incorrect use of grammar
and how she can improve.
In this article, we explore whether perceived expertise influences the type of feedback
individuals seek and give. Specifically, we examine whether individuals seek more negative
feedback and less positive feedback as they gain expertise in a domain of goal pursuit, and
whether feedback givers also shift toward giving negative feedback as their receivers gain
expertise.
THEORETICAL BACKGROUND
Feedback is essential for self-regulation. When consumers acquire a new skill or seek to
improve an existing skill, both positive and negative feedback can allow for realistic selfassessment and calibration of their efforts (Carver and Scheier 1998; Higgins 1987; Maheswaran
and Meyers-Levy, 1990). Clearly, additional reasons exist for why consumers might seek
feedback, namely, to enhance or maintain their positive view of themselves (Russo, Meloy and
Medvec 1998; Tormala and Petty, 2004). For example, loyal consumers are more open to
positive information about favored products than unfamiliar products, which enables them to
maintain a positive view of themselves as the type of person who uses top products (Ahluwalia,
Burnkrant and Unnava 2000; see also Wood, Rhodes and Biek 1985). However, when consumers
seek to acquire a new skill or improve an existing one, the motivation to enhance a positive view
of themselves is often secondary to the motivation to realistically assess their skills and gain a
sense for which direction they should pursue (Trope 1986).
With the objective of accurate self-assessment in mind, both (constructive) positive and
negative feedback on one’s performance are potentially useful. Moreover, in order to improve,
6
people might differentially attend to positive and negative feedback over the course of gaining
experience or expertise on a goal. For example, a person who is looking into mastering a new
language may desire different types of feedback at different points over the course of learning
the language in order to maintain the motivation to memorize new vocabulary and grammar
rules.
A SHIFT TOWARD NEGATIVE FEEDBACK
Our main proposition is that as people gain expertise in pursuing a goal, they shift toward
seeking negative feedback. At least two potential factors create such a shift and they potentially
operate conjointly to enable it. We next review these factors and identify the conditions under
which we can demonstrate the operation of our proposed motivational factor as responsible for
the shift in feedback seeking.
First is the informational account. Positive feedback might be more informative for
novices—those who are less likely to perform a task well—whereas negative feedback might be
more informative for experts—those who are unlikely to perform poorly (Ashford and Tsui
1991; Tesser 1988). For instance, a novice piano player is less likely to play a piece of music
correctly; she is likely to make many mistakes. For this player, hearing she played a series of
notes correctly is more informative because she rarely plays the right note at the right time. Thus,
this feedback has high information value compared with negative feedback that she played some
notes incorrectly. On the other hand, an expert piano player is unlikely to miss notes. Hearing he
missed some notes is rare and has more information value than hearing he did not miss other
notes.
7
Whereas the informational account could potentially create a shift toward seeking
negative feedback as people gain expertise, it is only valid to the extent that novices and experts
are evaluated on a similar scale. If different scales are applied, whether the frequency of
successful performance is higher for experts compared with novices becomes unclear. For
example, to the extent that an expert piano player expects to be evaluated based on his ability to
express his emotions, his likelihood of succeeding should not be higher than the novice pianist,
who expects to be evaluated based on her ability to play the right notes.
The second account for the shift is motivational. It suggests the meaning people derive
from feedback changes over time and that people seek either positive or negative feedback
depending on its meaning and its ability to serve as a motivational tool that allows them to focus
on tasks at hand. Specifically, feedback can either inform individuals of 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; Vroom 1964; Förster, Liberman and Higgins 2005). In this case,
positive feedback on one’s accomplishments 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 confirms 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 increases motivation
because it signals insufficient progress. For example, a student 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 progress at a more sufficient pace.
8
We predict that the meaning with which people imbue feedback on goal pursuit changes
as they gain expertise, and thus their way of motivating themselves shifts. Specifically, people
shift from evaluating commitment to monitoring progress as they gain expertise. Compared with
experts, novices feel uncertain about their level of commitment. Consequently, positive feedback
is more effective as it increases people’s confidence that they are able to pursue their goals, and
it increases their expectancy of attaining the goal. In this regard, positive feedback encourages
individuals to internalize or integrate new goals into their self-concept, thus increasing their
commitment to pursue the goal on subsequent occasions (Ryan and Deci 2000). However,
experts’ commitment is more secure and their focus is on monitoring their progress. Positive
feedback signals they can relax their efforts, whereas negative feedback motivates exerting
effort. Importantly, this analysis suggests hearing the same feedback—for example, ―you can
improve‖—can either signal commitment or progress depending on a person’s expertise level.
When novices hear they can improve, they infer that their goals are not valuable or that their
expectancy of attainment is low; that is, they infer low commitment. In contrast, when experts
hear the same feedback, they infer that they have not invested enough effort toward pursuing
their goals.
A related account suggests expertise acts as a buffer, protecting people from the
detrimental impact of receiving negative feedback on their motivation to adhere to their goals
(Linville 1987; Raghunathan and Trope, 2002; Trope and Neter 1994). Consistent with our
analysis, negative feedback is less likely to lower people’s commitment, including their
expectations for success and the value they place on pursuing their goals, to the extent that they
are experts and thus do not infer commitment from feedback. However, whereas previous
research predicts an increase in tolerance of negative feedback, we further predict that experts
9
will actively seek negative feedback and reject positive feedback as a way to facilitate their
motivation to pursue specific goals. Thus, whereas our key comparison concerns experts versus
novices (within each type of feedback), we further predict that experts who want to motivate
themselves will prefer negative over positive feedback.
Taken together, we assume positive feedback has a greater impact on novices than on
experts, whereas negative feedback has a greater impact on experts than on novices (see also,
Louro, Pieters and Zeelenberg 2007). Therefore, we predict that experts will actively seek more
negative feedback and less positive feedback compared with novices. We further expect a similar
shift toward negative feedback in the feedback individuals give to others as a function of the
recipient’s expertise.
PRESENT RESEARCH
We report three studies that test the hypotheses that experts shift to negative feedback
because the meaning individuals derive from feedback changes with expertise. Across these
studies, we either measured (study 1) or manipulated (studies 2-3) participants’ expertise levels.
In order to manipulate expertise, we had participants in study 2 seek feedback at the beginning of
a learning session, when they felt relatively inexperienced, versus at the end of the session, when
they felt relatively more experienced. In study 3, we manipulated the perceived expertise of a
team member by presenting him as an experienced versus inexperienced member of the team.
Specifically, in study 1, we explore the impact of expertise on the feedback-seeking
behavior of students in beginning versus advanced level French classes. In study 2, we examine
feedback-seeking behavior over time, as participants gain expertise with a task. We further test
for the information contained in feedback to demonstrate that novices (vs. experts) infer high
10
commitment from positive feedback, whereas experts (vs. novices) infer insufficient progress
from negative feedback. In study 3, we explore whether a similar shift toward negative feedback
occurs when people give feedback to an inexperienced versus experienced coworker on a
presentation. We also examine whether people view their feedback as conveying different
information for experts versus novices.
STUDY 1: LANGUAGE CLASSES
Consumers invest a large amount of resources (e.g., time, money) acquiring new skills
such as learning a new language and, in the process, seek positive and negative feedback on their
performances. To explore the impact of expertise, we investigated feedback seeking among
American 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 addition, compared with beginners, advanced students would express greater interest in
learning from an instructor who teaches using a style that emphasizes their mistakes and how
they can improve.
Method
Eighty-seven undergraduate students volunteered to participate in the study after their
French class. This study employed a 2 (expertise: beginner vs. advanced level French class
students) × 2 (feedback: positive vs. negative) mixed design where expertise was a betweensubjects factor and feedback was a within-subjects factor.
11
The experimenter surveyed American students at beginning and advanced level French
classes. Beginners primarily take classes focused on conversational and grammatical skills and
learn material designed to help them communicate at a basic level, whereas advanced level
(expert) students primarily take classes designed at reading classic French literature in French
and writing papers that offer insightful analyses of the text in French.
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 or conjugating new verbs and
how they can fix those mistakes‖ (negative feedback).
As a measure of feedback seeking, students rated their interest in taking a class with an
instructor who teaches using each particular style (7-point scales; 1 = not at all, 7 = very 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, advanced students in literature classes indicated they had
studied French for a longer time (M = 78.64 months, SD = 43.38) than beginners in oral classes
(M = 25.29 months, SD = 27.05; t(72) = 5.52, p < .001).
12
To test for the hypothesis, we compared participants’ interest in taking a class with an
instructor who uses a style emphasizing 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) = .05), suggesting participants’ interest in positive feedback and their interest in
negative feedback were largely independent of each other. An expertise × feedback repeated
measures ANOVA yielded a main effect for feedback, indicating participants preferred an
instructor use a style emphasizing negative feedback on what mistakes they make and how they
can improve, (F(1, 79) = 6.43, p < .02). We found no main effect for expertise. The analysis also
yielded the predicted expertise × feedback interaction, (F(1, 79) = 7.31, p < .01). Contrast
analysis revealed that beginners expressed greater interest than advanced students in an instructor
who uses a style that emphasizes what they do well (M = 4.96, SD = 1.15, versus M = 4.25, SD =
1.47; t(79) = 2.35, 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 (M = 5.45, SD = 1.22, versus M = 4.92, SD = 1.29; t(79) = 1.76, p =
.08; see figure 1).
These results are consistent with the hypothesis that novices seek positive feedback more
than experts, presumably because it allows them to infer greater commitment, whereas experts
seek negative feedback more than novices, presumably because it allows them to infer
insufficient progress. We further predict that experts prefer negative over positive feedback; that
is, expertise not only provides a buffer against the impact of negative feedback (Linville 1987;
Raghunathan and Trope, 2002; Trope and Neter 1994) but rather, as people gain expertise, they
actively seek negative feedback to motivate themselves. In this regard, another contrast analysis
revealed that advanced students expressed greater interest in an instructor who uses a style that
13
emphasizes how they can improve (M = 5.45, SD = 1.22) than in an instructor who uses a style
that emphasizes what they do well (M = 4.25, SD = 1.47; t(54) = 4.44, p < .001); hence advanced
students preferred constructive negative feedback over constructive positive feedback.
In addition to their course enrollment, we evaluated participants’ 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 their interest in an instructor who uses a style that
emphasizes how they can improve (r(77) = .31, p < .01). Similarly, the longer students were
enrolled in French classes, the lower their interest in an instructor who uses a style that
emphasizes what they do well, though this effect was marginal (r(77) = -.19, p < .10). These
correlations support the hypothesis beyond the possible effect of the content of the course, oral
versus literature, which could have also influenced what type of teaching style students prefer.
That is, students in oral classes might prefer an instructor who uses a style that emphasizes what
they do well when learning how to speak the language, whereas students who read and write in
French might prefer learning from an instructor who uses a style that emphasizes how they can
improve their skills.
Study 1 demonstrates a shift toward seeking negative feedback with gained expertise. We
offer a motivational account for these findings: expertise reduces the focus on assessing
commitment and increases the focus on monitoring progress as a source of motivation to invest
in the goal. Notably, this analysis echoes Dweck and Leggett’s (1988) research on learning
goals, which finds that students who believe they can improve their performance (incremental
theorists) benefit from negative feedback because it encourages them to work harder, whereas
14
students who believe their ability is fixed (entity theorists) show decreased motivation from the
same negative feedback because it signals low ability and reduces goal commitment.
This study uses a correlative design, which poses a possible limitation for inferring that
expertise caused the shift to negative feedback seeking. In addition, we have yet to show that
expertise changes the meaning individuals derive from feedback. Accordingly, in study 2, we
explore more directly how expertise influences one’s focus on assessing commitment versus
monitoring progress. In the next study, we manipulate participants’ perceived expertise by
having them complete multiple trials while learning a new skill. We examine how the relative
emphasis on seeking positive or negative feedback changes over time as one gains expertise
while learning a new task. We further test for the meaning individuals derive from feedback as a
function of expertise.
STUDY 2: LEARNING A NEW TASK
Study 2 examines how feedback-seeking behavior changes over time. We manipulated
expertise by tracking native-English speaking participants with no prior experience speaking or
writing in German over multiple trials where they learned a new skill, typing in German.
Specifically, level of expertise increased during the course of the task such that participants felt
they were novices on the first trial of the task and relative experts on the last trial. After each
trial, we measured participants’ interest in receiving positive versus negative feedback regarding
their performance. We predicted an increase in participants’ likelihood of seeking negative (over
positive) feedback as they advanced on the task and thus gained experience.
To demonstrate the motivational role of feedback, in a follow-up to study 2, we further
tested for the meaning individuals derive from feedback as a function of their expertise. We
15
predicted that compared with experts, novices will be more likely to infer from positive feedback
that their learning goal is important or valuable, whereas compared with novices, experts will be
more likely to infer from negative feedback that they should increase their efforts.
Method
Twenty-six undergraduate students participated for monetary compensation. This study
utilized a 6 (expertise: trial number one – trial number six) within-subjects design. The
dependent variable was the feedback participants sought after each trial, positive or negative.
The experimenter recruited participants who had no prior experience speaking or writing in
German to participate in a study regarding how people learn an unfamiliar task, typing in
German. Participants learned they would be engaging in a cognitive flexibility task and that the
researchers would assess their performance on learning a new skill.
Following recruitment, all participants read 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.‖ The experimenter told participants 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. Participants then saw a sample of what the screen set-up
looked like and where they would enter their text.
Next, participants read their performance on the trials would be measured in a few ways:
how quickly they typed the passage (their speed), whether they finished the passage, and the
accuracy of their typing as measured by the match between what they typed and the words in the
passage, as well as by the match between what they typed and the punctuation and capitalization
16
of the words in the passage. Participants further read that as with many learning tasks, they
would have a chance to choose what feedback they would like to receive about their performance
at different points in the study, specifically, before moving to the next trial. Participants read that
the feedback would be individually tailored to their performance.
Next, participants began the task. The passages contained German text from folk stories
or songs. 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. Participants 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 versus negative feedback. Participants had 30
seconds to complete each typing task. Once the time had passed, the program automatically
moved to the next screen.
We limited feedback seeking to one piece of information to tap into the tradeoffs people
often make when seeking feedback. Specifically, 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 participants
received on each trial was predetermined and equally informative. The feedback was ambiguous
enough so that participants felt (and reported in their debriefing) it had been personally tailored
17
to their performance. For example, participants who chose to receive positive feedback on trial
one 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 comparison, 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.‖ Overall, participants completed
six trials and sought feedback six times.
Results and Discussion
We coded a participant’s choice of feedback as a binary variable (1 = chose negative
feedback, 0 = chose positive feedback). In accordance with the hypothesis, a binary logistic
regression on choice of feedback revealed a significant linear trend indicating participants were
more likely to seek negative feedback as they progressed through the trials (beta = .21), (Wald χ2
(1) = 21.62, p < .01; see figure 2). Specifically, whereas only 52 percent of participants sought
negative feedback after their first trial, 74 percent sought negative feedback after their second
trial, 63 percent sought negative feedback after their third trial, 67 percent sought negative
feedback after their fourth trial, 71 percent sought negative feedback after their fifth trial, and 82
percent 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. In a departure from study 1, in this
study, we manipulated expertise by exploring feedback seeking at different points in time as
people learned a new task. These two studies thus complement each other, and they further
18
demonstrate the subjective nature of expertise: those in beginner classes had objectively more
experience performing their task than experts in study 2, who completed six trials of typing in
German. However, in the participants’ perception, those students in beginner classes felt
relatively inexperienced, whereas those who were approaching the end of the German typing task
felt as experienced as they would get.
Confirming that experts seek more negative and less positive feedback, we next test for
the meaning contained in feedback. We assume that compared with experts, novices focus on
assessing their commitment, that is, whether their goals are valuable or feasible. In this case,
positive feedback serves as a signal for high commitment. In contrast, we assume that compared
with novices, experts are focused on monitoring their progress; that is, they focus on whether
they have invested enough effort in pursuing their goals. In this case, negative feedback serves as
a signal that they have made insufficient progress toward their goal.
Study 2: Follow-up
We conducted a follow-up study to confirm the predicted inferences from feedback.
Participants (N = 232) completed a typing task similar to the one in the main study with a few
minor adjustments. The study employed a 2 (expertise: novice vs. expert) × 2 (feedback: positive
vs. negative) × 2 (meaning: commitment vs. progress) between-subjects design.
To ensure participants had enough time to garner sufficient expertise on the task, they
completed 15 typing trials. We randomly assigned participants to receive either positive or
negative 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. Assigning
participants to positive versus negative feedback allowed us to control for self-selection into
19
feedback when testing for the meaning individuals derive from feedback (though it also meant
we could not test for feedback seeking). The feedback participants received in this study was the
same as the feedback they received in the main study. After receiving their feedback, participants
completed another trial (trial 2 for novices and trial 15, the last trial, for experts) before rating
their feelings of making progress toward the goal (7-point scale; 1 = I feel like I have made
sufficient progress on the task, 7 = I feel like I have made insufficient progress on the task) or
their commitment to perform well on the task (7 point scale; 1 = I care about my typing skills
very little, 7 = I care about my typing skills very much). In both of these scales, higher numbers
reflect higher motivation to pursue the task at hand.
Follow-up Study: 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 interaction (F(1, 225) = 3.98, p < .05). We found neither significant main effects nor
2-way interactions. Figure 3 displays the results. Specifically, novices were more likely than
experts to indicate positive feedback signaled they care about their typing skills (i.e.,
commitment, M = 4.67, SD = 1.06,versus M = 3.68, SD = 1.95; t(56) = 2.42, p < .02). On the
other hand, novices were not more likely than experts to infer they were committed when they
received negative feedback (M = 4.50, SD = 1.53, versus M = 4.53, SD = 1.61; t(58) < 1, ns).
Additionally, experts were more likely than novices to indicate they made insufficient progress
when they received negative feedback (M = 4.93, SD = 1.16, versus M = 3.96, SD = 1.58; t(55) =
2.64, p < .02). However, experts were not more likely than novices to infer they had made
20
insufficient progress when they received positive feedback (M = 4.55, SD = 1.67, versus M =
4.15, SD = 1.41; t(56) < 1, ns).
We predict that giving novices (those at their first trial) positive feedback will increase
their performance motivation more than giving them negative feedback. In addition, giving
experts (those at their 14th trial) negative feedback will increase their performance motivation
more than giving them positive feedback. As a measure of performance motivation, we coded the
number of words each participant typed on the trial that followed the feedback: trial two for the
novices and trial 15 for the experts. Analysis of the number of words participants typed on trial
two revealed that novices exhibited better performance when they received positive feedback (M
= 19.05 words, SD = 3.72) than when they received negative feedback (M = 17.32 words, SD =
4.04; t(111) =2.38, p < .02). Additionally, analysis of the number of words participants typed on
trial 15 revealed experts exhibited better performance when they received negative feedback on
the previous trial (M = 16.16, SD = 4.53) compared with experts who received positive feedback
on the previous trial (M = 14.47, SD = 4.05; t(114) = 2.12, p < .04). Notably, we cannot compare
experts to novices in these analyses because we observe a general depletion of resources as
participants progress on the task: they typed more words on trial two than fifteen. Nonetheless,
the comparison between positive and negative feedback implies positive feedback increases
motivation initially and negative feedback increases motivation subsequently.
The findings from study 2 provide evidence for the hypothesis that as people gain
experience, they switch towards seeking negative feedback. In the follow-up study, we find
further support for our proposed motivational account, namely, that compared with experts,
novices focus on assessing their commitment, of which positive feedback serves as a signal for
high commitment, whereas compared with novices, experts focus on their pace of progress, in
21
which case negative feedback signals one has not invested enough effort. Indeed, novices
performed better when they received positive (vs. negative) feedback and experts performed
better when they received negative (vs. positive) feedback.
The first two studies demonstrate that individuals shift from seeking positive feedback to
seeking negative feedback as they gain expertise. Although consumers often seek information
about their performance (e.g., from physical trainers and coaches), they also often assume the
role of feedback giver, as when they provide feedback to service providers or organizations
about their services or products. In study 3, we ask whether a similar shift to negative feedback
occurs when people give feedback, such that people give more negative feedback to experts than
to novices. We predict that feedback givers shift toward providing more negative feedback to
experts than to novices because they view their initial feedback as a way to encourage novices to
take on a goal, and they view their subsequent feedback as signaling that recipients of feedback
need to increase their efforts. In a follow-up to study 3, we test this hypothesis.
STUDY 3: COWORKER EVALUATION
Study 3 explores how perceived expertise of another person influences the type of
feedback one gives to an expert versus a novice. We manipulated perceived expertise of one’s
coworker by emphasizing that a participant’s task was to evaluate a new member of the team
who has been with the company for two weeks (novice) versus an experienced member of the
team who has been with the company for two years (expert). The variable of interest pertained to
the feedback participants gave their coworker regarding his strengths and weaknesses. We
predicted that participants would provide more negative feedback but not less positive feedback
to the expert versus novice coworker.
22
Next, in a follow-up study, we explored the reason why feedback givers increase the
share of negative information in their feedback. We predicted that those who provided feedback
to an expert coworker would view their feedback as a way of signaling their coworker could
invest more effort in preparing for their upcoming presentation (feedback signaling their
coworker’s need to make progress), compared with those who provided feedback to novices.
Method
Forty-nine undergraduate students participated in return for monetary compensation. This
study utilized a 2 (expertise: novice versus expert) × 2 (feedback: positive versus negative)
mixed-design with expertise as a between-subjects factor and feedback as a within-subjects
factor.
The experimenter recruited participants for a coworker evaluation study. Their task was
to assume they were part of a team at a firm that was pitching a new product to a potential client
and that their job was to evaluate a coworker’s practice presentation for the client and to provide
feedback to their coworker. To ensure the practice presentation was similar to the final
presentation, participants read that their coworker’s practice pitch was filmed in front of a mock
audience.
We manipulated the perceived expertise of the participant’s coworker. Specifically, those
we assigned to the novice condition read that their coworker was ―a new member of the team
who has been around for two weeks,‖ whereas those we assigned to the expert condition read
that their coworker was ―an experienced member of the team who has been with the firm for two
years.‖ Participants read that their coworker had been asked to deliver part of the team’s pitch to
23
a new client and that their job was to evaluate the practice presentation of the pitch and give
feedback to their coworker on his strengths and weaknesses.
Next, all participants watched the same video clip featuring an entrepreneur, their
coworker, who was pitching a product, ―Novaflo,‖ designed to eliminate the likelihood of one’s
toilet, sink, or bathtub overflowing. After providing a demonstration of the product using a
bathtub filled with running water that stopped once the sensor for the product was triggered, the
entrepreneur outlined what his product and the projected market were before asking whether the
clients had any questions.
Participants then provided their open-ended feedback about their coworker’s (a) strengths
and (b) weaknesses in delivering the pitch. We counterbalanced the order of these questions.
Participants wrote, for example, ―you paused for too long and said umm too many times‖
(negative feedback) or ―you were confident, and your speed of delivering the pitch was just
right‖ (positive feedback). Participants then evaluated the overall presentation of the presenters
(7-point scale; 1 = the presentation was not very good, 7 = the presentation was very good) and
completed demographic items.
Results and Discussion
Two independent judges counted the number of positive and negative pieces of
information participants provided in their feedback. Inter-rater agreement on the number of
positive and negative pieces of information was high, α = .93.
A repeated-measures ANOVA of number of items yielded the predicted interaction (F(1,
47) = 4.93, p < .04). No other results were significant. Specifically, contrast analysis revealed
that participants provided more negative feedback to experts (M = 1.28, SD = 1.37) than to
24
novices (M = .36, SD = .62; t(47) = 3.07, p < .01). In comparison, participants provided similar
levels of positive feedback to novices (M = 2.27, SD = 1.10) and experts (M = 2.11, SD = 1.71; t
< 1; see figure 4).
Importantly, even though participants gave more negative feedback to the expert
presenter, their overall evaluation of the expert’s presentation was similar (M = 4.70, SD = 1.15)
to those of participants who evaluated a novice presenter (M = 4.69, SD = 1.29; t < 1). Thus,
even though participants did not evaluate the expert and novice coworkers’ presentations
differently, they nonetheless provided more harsh feedback to the expert coworker than to the
novice coworker.
The results from study 3 demonstrate that feedback providers shift toward giving more
negative feedback as the recipient of the feedback gains expertise. This shift toward providing
experts with more negative feedback than novices occurs because individuals view their
feedback as a way of signaling to experts their rate of progress as opposed to viewing their
feedback as a way of increasing commitment. In a follow-up study, we test this hypothesis by
asking whether feedback providers view their feedback as signaling progress more so for experts
than for novices.
Study 3: Follow-up
We predicted that the reason feedback providers give more negative feedback to experts
than to novices is that they wish to signal their coworker’s rate of (insufficient) progress. In a
follow-up study designed to test this hypothesis, participants (N = 111) completed the same
coworker evaluation task as in study 3, with a few minor adjustments. As in the main study, the
experimenter told participants they were part of a team pitching a product to a new client and
25
that before the team could give its pitch to the client, they had taped a practice presentation in
front of a mock audience To ensure a particular video clip did not drive our effect, participants
evaluated a different clip of an entrepreneur pitching a product to an audience than in the main
study.
Before providing their feedback on their coworker’s strengths and weaknesses,
participants rated their agreement with the following statements: (a) ―I give feedback to signal to
my coworker he has a lot of work he needs to accomplish before the actual presentation‖ (i.e.,
the feedback provides a signal of recipient’s progress) and (b) ―I give feedback to my coworker
to increase his confidence in his presentation‖ (i.e., the feedback secures the recipient’s
commitment; 7-point scales, 1 = strongly disagree with the statement, 7 = strongly agree with the
statement).
Follow-up Study: Results and Discussion
In support of the hypothesis, the ANOVA on ratings revealed the predicted expertise
feedback interaction (F(1, 107) = 6.74, p < 02). No main effects were significant. Contrast
analysis revealed participants who gave feedback to experts were more likely than those who
gave feedback to novices to see their feedback as a way to provide information about their
coworker’s progress (M = 5.03, SD = 1.3, versus M = 4.27, SD = 1.29; t(56) = 2.23, p = .03). In
comparison, those who gave feedback to experts were not more likely than those who gave
feedback to novices to view their feedback as a way to instill their coworker with commitment
(M = 3.92, SD = 1.26, versus M = 4.39, SD = 1.13; t(51) = 1.44, p < .16).
We intended the rest of the analysis to replicate the results of the main study 3 with the
new coworker. Independent judges’ analyzed the feedback for the amount of positive and
26
negative pieces of information given to the coworker (α = .81). The repeated-measures ANOVA
on this measure revealed a main effect for feedback (F(1, 105) = 48.26, p < .001), indicating
participants provided more negative than positive feedback to their coworker, as well as the
predicted feedback
expertise interaction (F(1, 105) = 5.15, p < .03). No other effects were
significant. Specifically, contrast analysis revealed participants provided more negative feedback
to an expert coworker (M = 4.04, SD = 1.98) than to a novice (M = 3.28, SD = 1.77; t(105) =
2.11, p < .04). However, participants were not more likely to provide positive feedback to a
novice (M = 2.35, SD = 1.22) than to an expert (M = 2.21, SD = 1.14; t(105) < 1, ns).
Our motivational account predicts that individuals who view their feedback as providing
a signal for insufficient progress will give more negative and less positive feedback to their
coworker. To explore this hypothesis, we created an index for the positivity of feedback by
subtracting the number of negative pieces of information from the number of positive pieces of
information participants provided. In support of this analysis, the more likely participants were to
view their feedback as a signal for low progress, the less positive their feedback was (r(54)= .29, p < .04). However, there was no significant correlation between participants views of
feedback providing a signal for high commitment and the positivity of their feedback (r(53) = .20, p < .15). Thus, consistent with the results from the study, we find an effect for negative
feedback but not for positive feedback.
Taken together, the findings from study 3 demonstrate that feedback givers shift toward
giving more negative feedback to experts than to novices. As the follow-up to study 3
demonstrates, one reason feedback providers give feedback is to signal to someone learning a
new skill that he should invest more resources toward pursuing his goal (feedback signaling
one’s rate of progress). Thus feedback givers appear to strategically tailor their feedback to the
27
expertise level of the receiver, as those giving feedback to experts (vs. novices) view their
feedback as a way of signaling their coworker can improve his presentation.
Notably, these findings echo previous research which demonstrates that when consumers
know they will provide an evaluation of a product or service experience, they provide more
negative feedback than those who do not anticipate evaluating the product or service (Ofir and
Simonson 2001). Our motivational account would suggest that this shift towards negative
feedback when consumers expect to evaluate a product or service reflects their perception that
the marketers’ of the product or service have greater expertise if they asked them to evaluate the
product in advance.
GENERAL DISCUSSION
Feedback is central to marketing communication, as marketers give consumers feedback
regarding their skills as well as receive feedback from them. Hence the current article
investigates when individuals are likely to seek and give positive feedback on strengths versus
negative feedback on weaknesses. We predict a shift toward negative feedback as people gain
expertise, because the meaning individuals derive from feedback changes such that negative
feedback increases motivation to adhere to a goal. In support of our prediction, we find that
novices infer from feedback whether their goal is valuable or feasible (commitment), whereas
experts infer from feedback whether their pace of pursuing an already valuable goal is sufficient
(progress). We further find that feedback providers view their negative feedback as signaling an
expert needs to invest more resources to acquire her desired skill. As a consequence of this
motivational mechanism, novices are more likely than experts to seek positive feedback on their
accomplishments, whereas experts are more likely than novices to seek negative feedback on
28
their lack of accomplishments. Additionally, feedback providers are more likely to give negative
feedback to those they consider an expert rather than a novice.
Results from three studies support these hypotheses. In study 1, we measure actual
expertise and show that novice French students, those in beginning-level language classes,
express greater interest than experts in learning from an instructor who uses a style that
emphasizes what they do well (their strengths), whereas expert French students, those in
advanced-level language classes, express greater interest than novices in learning from an
instructor who uses a style that emphasizes their mistakes and how they can improve (their
weaknesses). Study 2 demonstrates that people shift from seeking positive to seeking negative
feedback as they gain expertise learning a new skill, and that novices (vs. experts) focus on
assessing commitment, for which positive feedback signals high value or feasibility, whereas
experts (vs. novices) focus on their rate of progress, for which negative feedback signals the need
to invest more effort. Finally, the findings from study 3 demonstrate that a similar shift to
negative feedback occurs when providing feedback to others, negative feedback signals to
experts—as opposed to novices—they should increase their effort investment rather than lower
their commitment.
The Subjective Nature of Expertise
Across three studies, we either measured or manipulated expertise. The subtle
manipulations we used indicate expertise is a subjective sense of knowing a relatively large
amount about the task at hand. For example, in study 2, participants felt they were experts
mainly because they were about to finish the learning task. Similarly, in study 3, participants
29
perceived someone as an expert merely because they were told this person has been on the team
for a relatively long period of time.
Many factors might influence subjective feelings of expertise, and marketers can
influence these feelings to increase the impact of their feedback. One such factor might involve
power dynamics in how individuals interact with feedback. Receiving feedback from a person in
a position of higher power might make one question his own commitment and feel like a relative
novice. For example, a person who recently hired a personal trainer and gave that person the
power to dictate her diet and exercise regime may be reminded of the fact that, relative to her
trainer, she has much to learn about diet and exercise. Thus she is likely to feel like a novice
relative to her trainer. In this instance, we would predict that the novice will be more likely to
focus on assessing her commitment and thus will seek and respond more to positive feedback,
compared with a person who has more power.
Although power may indeed influence one’s sense of expertise, and, accordingly, the
feedback an individual seeks, note that across various types of power dynamics in our studies,
we consistently find a shift toward negative feedback with gained expertise. Specifically, study 1
utilizes a design where students receive feedback from an instructor who, presumably, is in a
position of higher power. In study 2, participants engage in a situation where no power dynamics
are involved, as they receive feedback from an unfamiliar computer program. Finally, in study 3,
participants give feedback to a coworker who presumably has the same amount of power as they
do.
Taken together, we expect factors that influence one’s sense of expertise to, accordingly,
influence one’s likelihood of seeking and giving positive versus negative feedback. Specifically,
if individuals are endowed with the sense that they are novices, they will be more likely to seek
30
positive feedback. However, if individuals are endowed with a sense of expertise, they will be
more likely to seek negative feedback.
Marketing Implications
The current findings have specific implications for marketers of learning and skill
acquisition products. For instance, our findings suggest marketers should design their feedback
with a consumer’s expertise level in mind. To illustrate, health clubs should instruct their trainers
to give new clients positive feedback about the things they do well (e.g., they have good form on
a particular exercise) and experienced clients negative feedback about the areas they can improve
(e.g., they can improve their form on a particular exercise). Similarly, weight loss programs
should emphasize that new attendees 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 frequent attendees can monitor their diets a bit closer if they would like to lose
weight.
These findings have further implications for how marketers, as well as educators and
social agents, can encourage people to adhere to their goals. 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, accordingly, provide
feedback that increases their motivation. Indeed, previous research attests that a negative
relationship exists between self-efficacy and performance as people progress on a task,
suggesting that providing more negative feedback on how consumers can improve as they gain
experience (and thus lowering one’s sense of self-efficacy) is more effective than providing
31
positive feedback on consumers’ strengths (Vancouver and Kendall 2006). One caveat to this
recommendation is that consumers should focus 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.
Finally, these findings have implications for soliciting feedback from consumers (Peters
and Waterman, 1982; Ofir and Simonson 2001; Oliver, 1997). Marketers should realize that the
valence of feedback they are getting may reflect their perceived expertise in addition to the level
of their services. For instance, our findings suggest a company that is considered an expert on
environmental issues should expect consumers to provide feedback on how it can improve its
environmental habits, for instance, by reducing packaging materials. Our findings also suggest
consumers give this feedback to encourage the company to act further on its environmental
mission.
32
REFERENCES
Ahluwalia, Rohini, Robert E. Burnkrant and H. Rao Unnava (2000), ―Consumer Response to
Negative Publicity: The Moderating Role of Commitment,‖ Journal of Marketing
Research, 37 (May), 203-14.
Ashford, Susan J., Ruth Blatt and Don Vande Walle (2003), ― Reflections 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 Seeking,‖ 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, NE:
University of Nebraska Press, 69-164.
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 Cognitive Approach to Motivation
and Personality,‖ Psychological Review, 95 (April), 256-73.
Feather, Norman T (1982), ―Actions in Relation to Expected Consequences: An Overview of a
Research Program,‖ in Expectations 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.
33
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,‖ Reading, MA: Addison-Wesley
Förster, Jens, Nora Liberman and E. Tory Higgins (2005), ―Accessibility from Active and
Fulfilled Goals,‖ Journal of Experimental 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.
Kruglanski, Arie W. (1990), ―Motivations for Judging and Knowing: Implications for Causal
Attribution,‖ in Handbook of Motivation and Cognition: Foundations of Social
Behavior,” Vol 2, (eds.) E. Tory Higgins and Robert M Sorrentino, New York: Guilford
Press, 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 & Task Performance,‖
Upper Saddle River, NJ: Prentice Hall.
Louro, Maria J., Rik Pieters and Marcel Zeelenberg (2007), ―Dynamics of Multiple-Goal
Pursuit,‖ Journal of Personality and Social Psychology, 2 (August), 174-93.
Maheswaran, Durairaj and Joan Meyers-Levy (1990), ―The Influences of Message Framing and
Issue Involvement,‖ Journal of Marketing Research, 27 (August), 361-7.
34
Miller, Dale T. and Michael Ross (1975), ―Self Serving Biases in Attribution of Causality: Fact
or Fiction,‖ Psychological Bulletin, 82 (March), 213-25.
Miller, George A., Eugene Galanter and Karl H Pribram (1960), ―Plans and the Structure of
Behavior,‖ New York: Henry Holt.
Ofir, Chezy and Itamar Simonson (2001), ―In Search of Negative Customer Feedback: The
Effect of Expecting to Evaluate on Satisfaction Evaluations,‖ Journal of Marketing
Research, 38 (May), 170-82.
Oliver, L. Richard (1997), ―Satisfaction: A Behavioral Perspective on the Consumer,” New
York: McGraw-Hill
Peters, Thomas J and Robert H. Waterman (1982), ―In Search of Excellence: Lessons from
America’s Best Run Companies,” New York: Harper & Row.
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), 43852.
Ryan, Richard M and Edward L. Deci (2000), ―Self-Determination Theory and the Facilitation of
Intrinsic Motivation, Social Development, and Well-Being,‖ American Psychologist, 55
(January), 68-78.
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
35
Advances in Experimental Social Psychology, Vol 21, (ed.) Leonard Berkowitz, 181-227.
Tormala, Zakary L. and Richard E. Petty (2004), ―Source Credibility 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-Assessment Theories of
Achievement Motivation: A Reply,‖ Motivation and Emotion, 10 (September), 247-61.
Trope, Yaacov & Efrat Neter (1994), ―Reconciling Competing Motives in Self-Evaluation: The
Role of Self-Control in Feedback Seeking,‖ Journal of Personality and Social
Psychology, 66 (April), 646-57.
Vancouver, Jeffrey B. and Laura N. Kendall (2006), ―When Self-Efficacy Negatively Relates to
Motivation and Performance in a Learning Context,‖ Journal of Applied Psychology, 91
(September), 1146-53.
Vroom, Victor H (1964), ―Work and Motivation,‖ New York: Wiley.
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
Consequences, Richard E. Petty and Jon A. Krosnick, eds. Mahwah, NJ: Lawrence
Erlbaum Associates, 283-13.
36
FIGURE 1: INTEREST IN FEEDBACK AS A FUNCTION OF EXPERTISE LEVEL AND
INSTRUCTOR WHO EMPHASIZES POSITIVE VERSUS NEGATIVE FEEDBACK
(STUDY 1)
Interest in Instructor
6
5.5
5
Novice
(Beginning
Level Class)
Expert
(Advanced
Level Class)
4.5
4
3.5
3
Style Emphasizes Positive
Feedback
Style Emphasizes Negative
Feedback
37
FIGURE 2: PERCENTAGE OF PARTICIPANTS WHO SOUGHT NEGATIVE FEEDBACK
AS A FUNCTION OF EXPERTISE LEVEL (STUDY 2)
Seeking Negative Feedback
90%
70%
50%
30%
1
2
3
4
Trial
Expertise
5
6
38
FIGURE 3: INFERENCES ABOUT PROGRESS ON AND COMMITMENT TO LEARNING
TO TYPE IN GERMAN AS A FUNCTION OF FEEDBACK AND EXPERTISE LEVEL
(STUDY 2)
Novice (After First Trial)
5.5
Expert (Before Last Trial)
5
Ratings
4.5
4
3.5
3
2.5
2
Positive Feedback Negative Feedback
Inferences of Commitment
Positive Feedback Negative Feedback
Inferences of Progress
39
FIGURE 4: POSITIVE AND NEGATIVE FEEDBACK GIVEN TO A COWORKER AS A
FUNCTION OF EXPERTISE LEVEL (STUDY 3)
Pieces od Information Given
5
4.5
Inexperienced
Coworker
4
Experienced
Coworker
3.5
3
Provides Positive Feedback
Negative Feedback