Content uploaded by Nicholas Epley
Author content
All content in this area was uploaded by Nicholas Epley on Nov 06, 2018
Content may be subject to copyright.
Many Hands Make Overlooked Work: Over-Claiming of Responsibility
Increases With Group Size
Juliana Schroeder
University of California, Berkeley
Eugene M. Caruso and Nicholas Epley
University of Chicago
Logically, group members cannot be responsible for more than 100% of the group’s output, yet claims
of responsibility routinely sum to more than 100%. This “over-claiming” occurs partly because of
egocentrism: People focus on their own contributions, as focal members of the group, more than on
others’ contributions. Therefore, we predicted that over-claiming would increase with group size because
larger groups leave more contributions from others to overlook. In 2 field studies, participants claimed
more responsibility as the number of academic authors per article and the number of MBA students per
study group increased. As predicted by our theoretical account, this over-claiming bias was reduced when
group members considered others’ contributions explicitly. Two experiments that directly manipulated
group size replicated these results. Members of larger groups may be particularly well advised to consider
other members’ contributions before considering their own.
Keywords: judgment and decision making, egocentrism, over-claiming, biases, groups
Supplemental materials: http://dx.doi.org/10.1037/xap0000080.supp
Individuals often overestimate their relative contributions to
collaborative endeavors. Whether completing chores with one’s
spouse, writing an article with collaborators, playing a team sport,
or designing a product within an organization, individuals’ claims
of responsibility for their group’s output routinely sum to more
than 100%, revealing “over-claiming” at the group level (Brawley,
1984;Kruger & Gilovich, 1999;Kruger & Savitsky, 2009;Ross &
Sicoly, 1979;Thompson & Kelley, 1981; for a review, see Leary
& Forsyth, 1987). Believing one deserves credit that others fail to
acknowledge can be a source of dissatisfaction and conflict in
groups. Although such “over-claiming” is reliable across many
experiments, relatively little is known about what moderates its
magnitude. Here we propose what could be an important moder-
ator: group size. Specifically, we propose that over-claiming will
tend to increase as the size of a group increases. We predict this
effect because over-claiming is produced, at least in part, by
egocentrically focusing on one’s own contributions more than on
others’ contributions. This egocentrism should yield more over-
claiming as group size increases because there are more contribu-
tions from others to overlook.
Accurately allocating responsibility for group outcomes in-
volves considering both one’s own and others’ contributions.
Although the formula for calculating such relative comparisons is
straightforward, getting accurate inputs into the formula is not.
Anything that influences how much people notice, remember, and
credit their own versus others’ contributions to a group outcome
should therefore influence over-claiming. Noticing, recalling, and
therefore, crediting one’s own contributions to a group is obvi-
ously easier than doing so for others.
1
As the size of a group
increases, it becomes increasingly difficult to notice, recall, and
therefore, credit the full extent of others’ contributions. Among
married couples, for instance, there is only one additional person to
1
Comparative judgments, such as the amount one contributes to a group
compared with others, involve three distinct cognitive stages: first, recruit-
ing information about oneself and others; second, forming absolute eval-
uations of the contributions of oneself and others; and third, weighting the
absolute evaluations to form a comparative judgment (Chambers & Wind-
schitl, 2004). Egocentric biases are likely to be present at each of these
stages, thereby increasing over-claiming in one of several different ways.
For example, during the information-recruitment stage, self-relevant infor-
mation might be more accessible and more available than equivalent forms
of other-relevant information, causing self-relevant information to loom
larger than other-relevant information (Kuiper & Rogers, 1979;Markus,
1977;Rogers, Kuiper, & Kirker, 1977;Ross & Sicoly, 1979). During
absolute valuation, people might use or select different standards when
evaluating themselves versus others. During comparative judgment, people
may attend more to self-relevant absolute evaluations, anchor more on
self-assessments (Kruger, 1999), and feel more confident in their self-
assessments and thereby rationally weight them more (Chambers & Wind-
schitl, 2004;Klayman & Burson, 2002;Windschitl, Kruger, & Simms,
2003). Our hypotheses do not depend on the precise nature of egocentrism
underlying a responsibility judgment.
This article was published Online First February 25, 2016.
Juliana Schroeder, Haas School of Business, University of California,
Berkeley; Eugene M. Caruso and Nicholas Epley, Booth School of Busi-
ness, University of Chicago.
We thank Kaushal Addanki, Cody Bengtson, Michael Bremmer, Emilie
Chen, Jesus Diaz, Eric Guo, Jill Jacobbi, Jasmine Kwong, Paul Lou,
Catherine Matteson, Adam Picker, Jenna Rozelle, Emily Shaw, Nia Sotto,
Sherry Tseng, Eli Tyre, and Rebecca White for assistance conducting
experiments, Max Bazerman for helpful comments on this research, and
the Booth School of Business for financial support.
Correspondence concerning this article should be addressed to Juliana
Schroeder, Haas School of Business, University of California, Berkeley,
2220 Piedmont Avenue, Berkeley, CA 94720. E-mail: jschroeder@
haas.berkeley.edu
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
Journal of Experimental Psychology: Applied © 2016 American Psychological Association
2016, Vol. 22, No. 2, 238–246 1076-898X/16/$12.00 http://dx.doi.org/10.1037/xap0000080
238
consider when evaluating responsibility. Although responsibility
claims in married couples may significantly exceed 100% when
summed together, they do not exceed it by much (i.e., M⫽
103.1%; Ross & Sicoly, 1979). However, in a larger group, such
as a family of five, there are four other group members to consider.
Each family member would be well aware of his or her own
contribution but less aware of others’ contributions. We therefore
predict more over-claiming in a large group, such as a family of
five, than in a smaller group, such as a married couple.
A phenomenon as robust as over-claiming is not produced only
by a single mechanism of egocentrism but instead is produced by
multiple independent mechanisms. In particular, egocentrism is
part of a larger class of myopic biases in judgment, including
tendencies to focus only on one item at a time (singularity prin-
ciple; Evans, 2006), to insufficiently incorporate background in-
formation into judgments (focalism; Windschitl, Kruger, &
Simms, 2003), to overweight more concrete entities (generalized-
group theory; McConnell, Sherman, & Hamilton, 1994), and to
consider collaborators as a collective group (support theory; Rot-
tenstreich & Tversky, 1997;Tversky & Koehler, 1994). That is, an
egocentric bias in responsibility allocations emerges because peo-
ple tend to focus on one thing at a time, the self is typically the
focus of our own attention, the self is a concrete entity, and people
often divide the world into what “I do” and what “they do.”
Several alternative mechanisms could also create over-claiming,
but would be inconsistent with our prediction that group size
affects over-claiming. For example, self-serving motives (Miller &
Schlenker, 1985;Schlenker & Miller, 1977), additivity neglect
(Riege & Teigen, 2013;Teigen & Brun, 2011), and using general
standards for local comparisons (LOGE model; Giladi & Klar,
2002) can produce over-claiming. None of these theories, how-
ever, predict that group size systematically increases over-
claiming. Group size should not affect one’s desire to think well of
oneself (e.g., self-serving bias), how people treat a 100% scale
(e.g., additivity neglect), or the likelihood of conflating absolute
evaluations for relative evaluations (e.g., LOGE model).
2
Our
prediction about the effect of group size on over-claiming, there-
fore, stems uniquely from the egocentric tendency to focus more
on one’s own contribution as a focal member of a group than on
others’ contributions. Perhaps more important, our proposed
mechanism also suggests a unique intervention for decreasing the
magnitude of the bias in a group setting: evaluating others’ con-
tributions before evaluating one’s own.
Our prediction that egocentrism underlies over-claiming, and
that over-claiming should increase as group size increases, is
supported by at least three lines of research. First, individuals may
over-claim responsibility not only for activities that reflect posi-
tively on them, as self-serving motives would predict, but also for
activities that reflect negatively on them, as only egocentrism
would predict (Brawley, 1984;Kruger & Gilovich, 1999;Thomp-
son & Kelley, 1981). For example, in a classic demonstration of
over-claiming, married couples over-claimed responsibility not
only for socially desirable items such as doing the chores, but also
for undesirable items such as causing arguments (Ross & Sicoly,
1979; see also Kruger & Gilovich, 1999). These results suggest
that over-claiming stems from a tendency to focus more on one’s
own contributions than on others’ contributions, whether they
reflect positively on the self or not.
Second, experimental manipulations that draw attention to peo-
ple’s own versus others’ contributions can change how they allo-
cate responsibility. Increasing a person’s focus on their own con-
tributions exacerbates their tendency to over-claim for themselves
(Burger & Rodman, 1983;Ross & Sicoly, 1979), whereas increas-
ing focus on others’ contributions diminishes over-claiming
(Caruso, Epley, & Bazerman, 2006;Savitsky, Van Boven, Epley,
& Wight, 2005). Because these interventions directly manipulate
the accessibility of others’ contributions, their effects are clearly
consistent with a reduced egocentric bias. The experiments we
present here provide further independent tests of this intervention
to reduce over-claiming.
Third, in domains extending beyond responsibility allocations,
considering others’ mental states and perspectives seems not to be
an automatic process (Apperly, Riggs, Simpson, Chiavarino, &
Samson, 2006). Rather, it requires motivation and effortful atten-
tional resources. For example, people tend to be more egocentric
when they must respond quickly (Epley, Keysar, Van Boven, &
Gilovich, 2004), are under cognitive load (Lin, Keysar, & Epley,
2010), or are in a positive mood (Converse, Lin, Keysar, & Epley,
2008). This suggests that considering others’ contributions to a
group requires more motivation and cognitive effort than consid-
ering one’s own contributions.
Here we extend these demonstrations of egocentrism in alloca-
tions of responsibility to predict that over-claiming will increase as
a group’s size increases. This mechanism not only predicts what
will increase over-claiming, but also what should decrease over-
claiming in groups. Specifically, if focusing on one’s own contri-
butions and failing to consider others’ contributions creates over-
claiming, then focusing on others’ contributions should reduce it
(Caruso et al., 2006;Savitsky et al., 2005). Therefore, we expect
the most over-claiming among large groups that allocate respon-
sibility without being led to consider others’ contributions. In
contrast, we expect over-claiming to be attenuated in smaller
groups or when people account for others’ contributions. We test
our predictions in both field and laboratory contexts by reanalyz-
ing published data and conducting three novel experiments.
2
We do not reject the possibility that self-serving biases, additivity
neglect, and applying general standards to local comparisons can create
over-claiming. However, these accounts cannot explain why group size
would affect over-claiming. We consider each in turn. First, whether
people only report their own contributions or also report others’ contribu-
tions does not change their motive to think well of themselves. Second, the
additivity neglect model suggests that people do not always treat 100% as
a cut-off unless it is made explicitly clear (Riege & Teigen, 2013;Teigen
& Brun, 2011). We examine participants’ responses in the “other-focused”
conditions in our own experiments and find very few additivity violations
(see Footnote 7). We also included explicit instructions in all of our
contribution measures that 0% meant contributing nothing and 100%
meant contributing everything. Manipulating group size is unlikely to
affect how people interpreted our scale. Finally, the LOGE model suggests
that people make comparisons to more generalized groups rather than to
the specific comparison group. If people were already making a compar-
ison to a large, generalized group, then we would see no effect of increas-
ing group size on their judgments. Furthermore, this mechanism would not
predict consistent over-claiming but rather that people might over- or
under-claim based on people’s perceptions of the broader group.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
239
OVER-CLAIMING INCREASES WITH GROUP SIZE
Re-Analysis of Authorship Data
Caruso et al. (2006, Study 1) emailed surveys to authors of
articles published in major organizational behavior journals with 3
to 6 authors.
3
Authors estimated the percentage that they person-
ally contributed of the overall amount of work, amount of writing,
and amount of thought. Participants in the control condition (n⫽
108) answered these questions only for themselves. To assess
whether drawing attention to others’ contributions would reduce
over-claiming, participants in the other-focused condition (n⫽89)
first listed the initials of each of their coauthors on the article and
the percentage that each coauthor contributed before estimating
their own contributions.
4
Because we did not have responsibility claims from all authors
on all articles, we computed an index of implied responsibility. To
do this, we averaged each participant’s claims of work, thought,
and writing contributed (␣⫽.92) and multiplied this number by
the size of their author group. For example, if an author on a
four-author article claimed to have contributed 30% of the work,
25% of the writing, and 20% of the thought, the implied respon-
sibility would be 100%. This metric provides an appropriate mea-
sure of over-claiming for two reasons. First, it approximates the
standard measure of over-claiming whereby each person’s claimed
contributions in the group are summed and compared to the logical
limit of 100% (Ross & Sicoly, 1979). If our sample included all
people from all groups, our implied measure of responsibility
would be identical to this standard measure. Second, our measure
requires only one assumption: total work in a group cannot logi-
cally exceed 100%. It does not require any assumptions regarding
the actual distribution of work accomplished because we do not
compare each individual’s claim to his or her actual contribution
but rather to the logical limit of 100%.
Results were partially consistent with our hypotheses (see Fig-
ure 1). First, we conducted a linear regression predicting implied
responsibility using group size and dummy variables (0 or 1) for
each author order (second author position, third author position,
fourth author position, and fifth author position) as our indepen-
dent variables. We controlled for author order because author order
correlated with implied responsibility, r⫽⫺0.68, p⬍.001, as one
would expect given author ordering standards in this field. As
predicted, over-claiming increased with group size, ⫽0.16, p⫽
.004. This relationship was statistically significant in the control
condition, ⫽0.19, p⫽.014, and was statistically nonsignificant
in other-focused condition, ⫽0.11, p⫽.129. However, a
separate regression analysis predicting implied responsibility from
the experimental condition (0 ⫽control,1⫽other-focused),
group size, the size by condition interaction, and the author posi-
tion dummy variables, yielded a statistically nonsignificant size by
condition interaction, ⫽⫺0.22, p⫽.354. Finally, in a separate
analysis predicting implied responsibility from experimental con-
dition (0 ⫽control,1⫽other-focused) controlling for author
position, participants in the other-focused condition over-claimed
less than did participants in the control condition, ⫽⫺0.12, p⫽
.018.
Although all of the authors surveyed produced objectively high
quality articles—articles published in top-tier organizational be-
havior journals—larger author groups claimed more responsibility
for their published article than the smaller author groups. Authors
also reported contributing relatively less when they considered
others’ contributions before their own. These findings from a
previously published experiment provide some support for our
hypotheses: over-claiming increases with group size, and consid-
ering others’ contributions reduces over-claiming. We conducted
three new experiments to test our hypotheses more systematically.
Experiment 1: Study Groups
Large author groups claimed more responsibility for their pub-
lished article than small author groups in Caruso et al. (2006).
Experiment 1 used a different field context to test our hypotheses:
MBA study groups. One explanation for the results in the author-
ship study is that indicating the percentage contributed by each
coauthor in the other-focused condition simply made the 100%
contribution limit salient. To explore this possibility, we added a
more implicit other-focused condition in which participants merely
considered their other group members’ contributions before report-
ing their own, but did not explicitly indicate contributions that
summed to 100% (Savitsky et al., 2005). We again hypothesized
that implied responsibility would increase with group size, and that
considering others’ contributions (both explicitly and implicitly)
would reduce over-claiming.
Method
Participants. Participants were 699 MBA students enrolled in
a negotiations course. We sent an online survey to all students
enrolled in the course as part of a learning exercise; 710 students
completed the entire survey, but 11 of them requested to have their
data removed from any analyses that were not related to the course
learning exercise.
Procedure. We randomly assigned participants to a control
(n⫽282), implicitly other-focused (n⫽211), or explicitly other-
3
See the original article for rationale behind selection; 41% of the
sample responded to the e-mailed survey.
4
We do not list survey items irrelevant to our current hypotheses. For all
items, see Supplemental Material.
150%
200%
250%
300%
3456
ytilibisnopseR deilpmI
Number of Authors Listed on Article
Control Survey Condition
Other-Focused Condition
β= 0.19, p= .02
β = 0.11, p = .13
Figure 1. Implied responsibility in the reanalysis of Caruso et al. (2006)
authorship data by the number of authors per article and the survey
condition.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
240 SCHROEDER, CARUSO, AND EPLEY
focused (n⫽206) condition.
5
Participants in the control condition
first answered six measures of work claiming. The first item was:
“Of the total work that your study group did last semester, what
percent of the work do you feel like you personally contributed?”
Items 2– 6 asked participants to indicate the percentage they per-
sonally contributed to: preparation of case write-ups; creative or
intellectual insight; suggestions for how the group should best run;
interesting questions raised; and answers to other group members’
questions. Finally, participants reported their study group size and
each member’s name.
Participants in the implicitly other-focused condition completed
the same questions in a different order, reporting their group size
and each member’s name first and then reporting the percentage
they personally contributed on the six items. Participants in the
explicitly other-focused condition answered in the same order, but
also reported the percentage each group member contributed be-
fore reporting their own contributions.
Results
We could not match each individual with their group members
because students completed surveys anonymously. Therefore, we
computed implied responsibility using the same methodology for
the authorship experiment: We averaged each individual’s claims
on the six items (␣⫽.82) and multiplied this number by the group
size (see Figure 2). Results were consistent with our primary
hypothesis. In a regression that predicted implied responsibility
from group size, a dummy variable for the implicitly other-focused
condition (1 ⫽implicitly other-focused,0⫽control,0⫽explicitly
other-focused), and a dummy variable for the control condition
(1 ⫽control,0⫽implicitly other-focused,0⫽explicitly other-
focused), we found that over-claiming increased as group size
increased ⫽0.19, p⬍.001. Consistent with an egocentric
account of over-claiming, a regression predicting implied respon-
sibility from the implicitly other-focused condition (1 ⫽implicitly
other-focused,0⫽control,0⫽explicitly other-focused) and
explicitly other-focused condition (1 ⫽explicitly other-focused,
0⫽control,0⫽implicitly other-focused) revealed that consider-
ing others’ contributions both explicitly (⫽⫺0.34, p⬍.001)
and implicitly (⫽⫺0.15, p⬍.001) significantly reduced im-
plied responsibility compared to the control condition. Finally,
considering others’ contributions marginally reduced the relation-
ship between group size and responsibility in the control condition
(n⫽282, r⫽.29, p⬍.001) compared with both the implicit
condition (n⫽211, r⫽.13, p⫽.053) and the explicit condition
(n⫽206, r⫽.14, p⫽.042), zs⫽1.83 and 1.71, ps⫽.067 and
.087. In a regression predicting implied responsibility from group
size, control condition (1 ⫽control,0⫽implicitly other-focused,
0⫽explicitly other-focused), implicitly other-focused condition
(1 ⫽implicitly other-focused,0⫽control,0⫽explicitly other-
focused), the control condition by group size interaction, and the
implicitly other-focused condition by group size interaction, we
found a significant control condition by group size interaction, ⫽
0.47, p⫽.003. All other predictors were nonsignificant.
Discussion
Larger MBA study groups claimed more responsibility for their
group’s output than smaller groups, with groups of eight plus
members dramatically claiming more than 140% credit. However,
explicitly—and implicitly— considering group members’ contri-
butions reduced the extent of over-claiming. These results again
demonstrate that over-claiming increases as group size increases.
Being led to focus on others’ contributions significantly reduced
over-claiming, again suggesting that over-claiming comes not
from self-serving motives but rather from a self-centered focus on
one’s own contributions. Even a relatively subtle reminder of
others’ contributions, one that did not mention the 100% contri-
bution limit, significantly reduced over-claiming. These results
make it clear that the tendency to overweight one’s own contribu-
tions to a group comes not from an inability to consider others’
contributions, but rather from a tendency to overlook them unless
prompted to do so.
Experiment 2: Hand Grips
The first two experiments test our hypotheses in naturally oc-
curring groups, but these field experiments have at least three
potential drawbacks: differential selection by members into large
or small groups, unobservable interdependence of the data, and no
objective standard of accuracy. Experiment 2 addresses all of these
concerns by manipulating group size experimentally, analyzing
data from complete groups, and selecting a task with measurable
performance: a handgrip competition. Because we can match in-
dividuals to their groups in this experiment, we computed over-
claiming using the standard measure whereby we sum each group
member’s individual responsibility claim to create a total percent-
age score for each group (Ross & Sicoly, 1979).
Method
Participants. Participants were visitors to the Chicago Mu-
seum of Science and Industry (N⫽339, M
age
⫽35.8 years, 51%
men). Because we conduct analyses at the group level, we ex-
5
See Supplemental Material for an explanation of unequal sample sizes
across conditions.
90%
100%
110%
120%
130%
140%
150%
160%
<4 5 6 7 >8
ytilibisnopse
R deilpmI
Group Size
Control Condition
Implicitly Other-Focused Condition
Explicitly Other-Focused Condition
Accuracy
Figure 2. Implied responsibility as a function of group size and survey
condition in Experiment 1. Error bars represent ⫾1SEM.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
241
OVER-CLAIMING INCREASES WITH GROUP SIZE
cluded three groups in which at least one participant failed to
answer our primary dependent variable.
Procedure. We randomly assigned participants to one of four
conditions in a 2 (group size: 3 vs. 6) ⫻2 (survey: control vs.
other-focused) between-groups design. Once the experimenter as-
sembled a complete group, she explained that this group would be
competing against other groups in a handgrip competition. Each
group member then received a handgrip with a built-in counter
(obscured during the competition). The experimenter explained
that the handgrip competition had four rules: “Rule Number 1: you
have exactly 1 min to grip as much as you can. Rule Number 2:
you can only grip with one hand at a time, but feel free to switch
hands if one gets tired. Rule Number 3: you can grip however you
want but you have to use your hand and not your body or anything
else to do the gripping. Rule Number 4: you can’t help your
teammates by taking their grips for them, but you can feel free to
shout encouragements and strategy to each other.” The experi-
menter further told participants that the winning group would have
the highest average number of team grips, and that each member
of this winning group would receive a $20 Amazon.com gift card.
Participants competed together in a line, so they could each see
each other’s efforts. On the experimenter’s signal, each group
member squeezed the handgrip as many times as possible for 1
min.
After their session, participants completed a survey individually.
In the control condition, participants reported the percentage they
personally contributed to the group’s total. In the other-focused
condition, participants first reported the percentage each other
group member contributed to the total, and then the percentage
they personally contributed. To make it clear that 100% was the
maximum contribution participants could report, we measured
claimed percentage contributions as follows: “What percentage of
the total number do you think you were personally responsible for?
0% means that you contributed none of the total count (you had 0
grips), and 100% means that you contributed all of the total
count.”
6
Results
We summed each individual’s claims of responsibility to create
a total responsibility claim for each group, and then conducted
analyses at the group level (see Figure 3). Consistent with our
primary hypothesis, six person groups (M⫽113.6%, SD ⫽
23.7%) claimed more responsibility than three person groups (M⫽
104.0%, SD ⫽19.6%), F(1, 73) ⫽4.44, p⫽.04, p
2⫽0.06.
Consistent with an egocentric mechanism, control condition
groups (M⫽114.8%, SD ⫽27.0%) claimed more responsibility
than other-focused groups (M⫽102.6%, SD ⫽14.0%), F(1,
73) ⫽7.13, p⬍.01, p
2⫽0.09. Finally, consistent with our
prediction that the most over-claiming should occur in large
groups that allocate responsibility without being led to explicitly
consider others’ contributions, we found that individuals in the six
person control condition (coded as 3) claimed more responsibility
(M⫽122.8%, SD ⫽29.5%) than the other three conditions
combined (each coded as ⫺1), t(74) ⫽3.21, p⬍.01, d⫽0.75.
Our account further suggests that calling attention to others’
contributions should help people adjust their egocentric absolute
assessments of effort into relative assessments, thereby increasing
the accuracy of their assessments. Testing this prediction, we
conducted a hierarchical regression model predicting individual
claims from the actual percentage contribution (at the individual
level). This regression revealed a marginally significant interaction
between survey condition and actual contribution, ⫽⫺0.36, p⫽
.08, such that the relationship between the actual and claimed
percentage was directionally larger in the other-focused condition
(r⫽.71) than in the control condition (r⫽.52).
Because this experiment provides a measure of actual contribu-
tion (e.g., actual grips), we conducted an additional test that could
provide insight into consequences of group size. Specifically, we
tested whether the actual number of handgrips participants per-
formed varied by experimental condition. We predicted that group
size would increase over-claiming, but not actual performance.
Consistent with this prediction, the number of grips per member
did not vary significantly by survey condition, F(1, 73) ⫽0.10 or
group size, F(1, 73) ⫽1.91, or the interaction between these
conditions, F(1, 73) ⫽1.30. Although responsibility claims are
correlated with the actual amount of work performed, group size is
a stronger predictor of claimed responsibility than of actual work.
Discussion
Few people, if any, would claim that their handgrip ability plays
a central role in their lives, nor would they claim that being a good
hand-gripper contributes meaningfully to their sense of self-worth.
And yet, being randomly placed in a large group as part of a
handgrip competition increased over-claiming of responsibility
compared to being randomly placed in a small group. This exper-
iment provides critical causal evidence that being in a large group
increases over-claiming, in a context where alternative mecha-
6
We examined the data to further ensure that participants understood
that 100% was the maximum limit for the scale. Specifically, we examined
how many participants’ reports of their own and their team members’
contributions summed to exactly 100% in the “other-focused” conditions.
There were 19 6-person groups and 21 3-person groups in the other-
focused conditions in Experiment 2. We removed one group who did not
report all of their group members’ contributions, yielding 174 participants.
The majority of participants’ reports summed to exactly 100% (120 out of
174), providing evidence that most participants treated the scale additively.
The average sum of the member reports was 100.1% (SD ⫽7.8%).
80%
90%
100%
110%
120%
130%
140%
Control Condition Other-Focused Condition
ytili
bisno
pse
R puorG
3 person group
6 person group
Figure 3. Claimed responsibility as a function of group size and survey
condition in Experiment 2. Error bars represent ⫾1SEM.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
242 SCHROEDER, CARUSO, AND EPLEY
nisms based on task desirability or self-esteem are unlikely expla-
nations for over-claiming. Replicating Experiment 1, considering
others’ contributions reduced inflated responsibility claims, espe-
cially in the larger group. These results suggest that over-claiming
is produced at least partly by an egocentric focus on one’s own
contributions. Smaller groups may tend to over-claim less because
group members can more easily think about the contributions of
other members, whereas larger groups need a reminder to think
about others’ contributions to reduce over-claiming.
Experiment 3: Remembered Groups
To achieve both high ecological and internal validity, we in-
structed participants to recall real groups that were either relatively
small or large. As an additional test of the underlying mechanism
of egocentrism, we manipulated the relative accessibility of one’s
own versus others’ contributions by having some participants
report their own contributions first and others report their own
contributions last. Finally, in addition to assessing responsibility as
a percentage of total work, we also asked participants to assess
responsibility on a more absolute measure (not bounded by the
logical limit of 100% responsibility). If over-claiming is produced
by focusing egocentrically on one’s own contributions, then the
correlation between the amount of work claimed on the absolute
scale and the percentage of total work claimed should be larger in
the control condition than in the other-focused condition. That is,
participants’ responsibility claims in the control condition should
be based more heavily on the absolute amount of work they
believe they did than in the other-focused condition, where judg-
ments should be more sensitive to the amount of work they did
compared to others.
Method
Participants. An online panel (N⫽1962, M
age
⫽50.2 years,
50% men) maintained by Qualtrics participated.
Procedure. We randomly assigned participants to one of six
conditions in a 2(group size: small vs. large) ⫻3(focus: control vs.
other-focused you first vs. other-focused you last) between-
participants design. We instructed participants to think of a time
when they worked with a group of either 2– 4 (small group) or
5–10 (large group) individuals. Participants reported the exact
number of group members and then described the project briefly.
In the control condition, participants next reported their own
contributions to the group as both a percentage of the total (0 –
100%, entered in a blank box), and as an absolute amount (on a
slider scale ranging from “contributed none of the work”to“con-
tributed all of the work”). We counterbalanced the order of the
percentage and absolute claiming measures across participants.
In the other-focused conditions, participants reported their group
members’ names after describing the project. They then reported
their own contributions either first or last, along with the contri-
butions of each group member (whose names were populated
automatically) on the same percentage and scale items (the order
of which was again counterbalanced).
Results
We computed implied responsibility by multiplying each indi-
vidual’s responsibility claims by the group size. Consistent with
our primary hypothesis, large groups (M⫽171.0%, SD ⫽
148.1%) claimed more implied responsibility than small groups
(M⫽141.1%, SD ⫽90.3%), F(1, 1956) ⫽36.01, p⬍.01, p
2⫽
0.02.
7
Consistent with egocentrism, focus condition affected im-
plied responsibility, F(2, 1956) ⫽102.47, p⬍.01, p
2⫽0.10.
Participants in the control condition claimed more (M⫽204.0%,
SD ⫽143.2%) than participants in the other-focused you first
condition (M⫽144.5%, SD ⫽108.5%), who in turn claimed more
than participants in the other-focused you last condition (M⫽
117.7%, SD ⫽90.6%), ts(1959) ⬎4.21, ps⬍.01, ds⬎0.19.
Individuals in the large group control condition (coded as 5)
claimed more responsibility (M⫽235.8%, SD ⫽164.6%) than the
other five conditions combined (each coded as ⫺1; M⫽141.3%,
SD ⫽106.6%), t(1956) ⫽12.85, p⬍.01, d⫽0.58 (see Figure
4).
8
As further evidence for egocentrism, the amount participants
claimed to have contributed on the scale measure was more
strongly correlated with percentage claims in the control condition
(r⫽.80) than in either of the other-focused conditions (rs⫽0.40),
zs⫽2.81, ps⬍.01. Calling attention to others’ contributions led
participants to rely less on an absolute assessment of the amount
they contributed than on a relative assessment of the percentage
they contributed.
9
Discussion
Using an experimental design with high internal validity that
also captured the ecological validity of real group work, we dem-
onstrated again that over-claiming increases in larger groups. We
also provided two further points of evidence showing that over-
claiming is egocentric. First, asking participants to list other group
members’ contributions, whether first or last, reduced over-
claiming. This addresses the concern that any score listed last
would be trimmed because of the 100% limit. Instead, our findings
imply it is not simply the case that listing one’s own contributions
last per se reduces over-claiming— but rather that the act of
considering group members’ contributions, whether first or last, is
what reduces over-claiming. Second, considering others’ contribu-
tions also reduced reliance on absolute assessments of contribu-
tion. When not reminded of others’ contributions, participants’
relative claims of responsibility were largely a function of how
7
As expected, implied responsibility increases with group size when
actual group size is treated as a continuous variable collapsed across our
two experimental conditions, r(1962) ⫽0.33, p⬍.001.
8
See Supplemental Material for additional analyses.
9
These data allow us to test an alternative explanation for our findings,
which is that divisions of 100% by smaller numbers (e.g., 3–33.3%,
4 –25%, and 5–20%) are more accessible (more fluent and less effortful)
than divisions by larger numbers (e.g., 6 –16.67%, 7–14.29%, and
8 –12.5%). Specifically, people might be more likely to anchor on their
“equal share” for smaller-sized groups because the division is easier. If this
were true, then people in 10-person groups (10% each) should over-claim
less than people in 9-person groups (11.11%). However, if our account is
correct, there should be more over-claiming in groups of 10 compared to
9. The data reveal that 10-person groups over-claim more (n⫽30, M⫽
441.3%, SD ⫽264.0%) than 9-person groups (n⫽9, M⫽373.0%, SD ⫽
190.4%), although this difference was not statistically significant,
t(37) ⫽⫺0.72, p⫽.48. However, the increase in over-claiming from
9-person to 10-person groups was similar to the increase from 8-person to
9-person groups, suggesting that ease of processing did not reduce over-
claiming.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
243
OVER-CLAIMING INCREASES WITH GROUP SIZE
much absolute work they believed they did—suggesting that in-
dividuals were egocentrically relying on absolute assessments to
inform their relative claims. However, after considering others’
contributions, absolute assessments of own contribution did not
predict relative claims as strongly.
General Discussion
In four field and laboratory experiments, with academic authors,
students, museum-goers, and a large-scale national sample, we
demonstrate that over-claiming responsibility for group tasks in-
creases with group size. Whereas classic demonstrations of ego-
centric allocations of responsibility show rather meager over-
claiming never exceeding an average of 110% (e.g., spouses, Ross
& Sicoly, 1979; video game pairs, Kruger & Gilovich, 1999), we
find dramatic over-claiming exceeding 235% in larger groups
(Experiment 3), suggesting that biased responsibility assessments
may be more consequential than previously demonstrated because
prior research has primarily studied small groups. Over-claiming
increases with group size, at least in part because people tend to
assess responsibility by focusing on their own contributions and
overlooking others’ contributions. As groups get larger, there are
simply more contributions from others to overlook. Across our
experiments, over-claiming was consistently highest in large
groups when people only reported their own relative contribution,
and was attenuated in smaller groups or when people explicitly
considered others’ contributions.
Our results make at least two important theoretical contribu-
tions. First, we provide additional evidence that over-claiming
stems from an egocentric focus on one’s own contributions. Over-
claiming is robust partly because it is produced by multiple mech-
anisms, such as self-serving motives to claim credit for positive
outcomes (Leary & Forsyth, 1987;Schlenker & Miller, 1977) and
the tendency to use scales nonadditively (Teigen & Brun, 2011).
Indeed, interventions could shift the relative weight of these dif-
ferent mechanisms for over-claiming. For example, if we rewarded
our participants for their claimed contributions, thereby incentiv-
izing them to claim more, over-claiming might result more from
self-serving motives and as such would probably be less affected
by other-focused interventions designed to combat egocentrism.
However, in the absence of extrinsic incentives to over-claim, our
research provides several unique pieces of evidence consistent
with an egocentric mechanism for over-claiming. In particular,
focusing on other group members’ contributions in all experiments
reduced over-claiming. Extending prior research (Caruso et al.,
2006;Savitsky et al., 2005), we also demonstrated the robustness
of the effect for different types of focus manipulations: asking
participants to explicitly list their group members’ percentage
contributions before or after their own contributions as well as
merely asking participants to list their group members’ names.
These findings suggest that even more subtle manipulations to
increase individuals’ focus on others may be effective to reduce
over-claiming, such as priming an interdependent self-focus
(Brewer & Gardner, 1996).
Our experiments provide two further results that support an
egocentric mechanism. Perhaps most novel, our experiments dem-
onstrated that focusing on others’ contributions not only reduced
over-claiming but also increased the accuracy with which partic-
ipants estimated others’ contributions. Reducing bias in evalua-
tions of one’s own contributions need not automatically increase
accuracy in estimates of others’ relative contributions (Teigen &
Brun, 2011). A coach who learns that she overvalued one player
does not automatically become more accurate in her evaluations of
other players on the team. Furthermore, group members also relied
less on absolute assessments of their own contributions for report-
ing relative contributions when reminded of their group members.
These new findings collectively point to egocentrism as an impor-
tant mechanism for over-claiming, implicating previously uncon-
sidered factors, such as group size, that should moderate respon-
sibility allocations.
Second, our results make a broader theoretical point about the
aggregation of seemingly small psychological effects. More than
two decades ago, Prentice and Miller (1992) cautioned against
using effect size as a measure of importance because effect sizes
depend on the strength of an independent variable and the malle-
ability of a dependent variable. Our results contribute another
cautionary note to their argument. Effects that are objectively
small when studied at the individual level—such as egocentric
biases in dyads—may compound when aggregated across groups.
This is true for small amounts of accuracy in individual judgment,
where aggregating across individuals within a group is responsible
for the well-known “wisdom of crowds” effect (Hastie & Kameda,
2005;Surowiecki, 2004). Our results, in concert with others (Sim-
mons, Nelson, Galak, & Frederick, 2011), demonstrate the same
result for small amounts of bias that, when aggregated across
larger groups, can also reveal the foolishness of crowds.
Finally, our research also has important consequences for
groups across domains, whether in organizations, sports teams, or
academic collaborations. Based on our findings, we expect it
would become increasingly difficult for groups to determine how
to equitably split rewards (and punishments) as group size in-
creases. Just as the team of 3,221 coauthors on a single publication
(ATLAS Collaboration, 2010) will over-claim more than the team
of three coauthors, the former team is also likely to have trouble
deciding how to split a reward. Larger groups may show greater
reliance on equality heuristics (i.e., each group member receiving
the same amount; Messick, 1993) than on equitable distributions of
returns. Perceived violations of equity, a dominant concern in
nearly every social relationship (Walster, Walster, & Berscheid,
1978), are likely to be especially common in larger groups because
individuals will believe they received less compensation than their
deserved share.
0%
50%
100%
150%
200%
250%
300%
Control Condition Other-Focused You First
Condition
Other-Focused You Last
Condition
ytilibisnopseR deilpmI
Small Group
Large Group
Figure 4. Implied responsibility as a function of group size and survey
condition in Experiment 3. Error bars represent ⫾1SEM.
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
244 SCHROEDER, CARUSO, AND EPLEY
Accordingly, larger groups are likely prone to more dissatisfac-
tion than smaller groups— especially when group members’ un-
equal responsibility allocations are made explicit. Consistent with
this proposition is empirical evidence that egocentrism instigates
dissatisfaction in groups: Group members dislike those who appear
to take more credit than they deserve (Forsyth, Berger, & Mitchell,
1981), negotiators often overestimate the likelihood a neutral judge
will agree with their egocentric assessments of fairness (Babcock,
Loewenstein, Issacharoff, & Camerer, 1995), and egocentrism
predicts negotiation impasse (Thompson & Loewenstein, 1992).
The lead author on a two-author article is relatively unlikely to
claim an outsized share of credit, meaning that both authors will
feel suitably appreciated for their work. However, should one of
the 3,221 authors in the ATLAS collaboration claim to have done
the lion’s share of the group’s work, he or she will surely leave
many disaffected colleagues.
Concluding Thought
Whether collaborating on an article, competing in a handgrip
challenge, or completing work for an organization, larger groups
tend to claim more than their fair share of credit compared with
smaller groups. We suggest that over-claiming increases with the
size of the group because overlooking group members’ contribu-
tions becomes easier as group size increases. As a result, members
of larger groups tend to rely more on their egocentric assessments
of contribution to inform relative responsibility allocations, and
also tend to be less accurate in their allocations. Members of larger
groups may be particularly well-advised to remember that many
hands make overlooked work—and to consider others’ contribu-
tions alongside their own.
References
Apperly, I. A., Riggs, K. J., Simpson, A., Chiavarino, C., & Samson, D.
(2006). Is belief reasoning automatic? Psychological Science, 17, 841–
844. http://dx.doi.org/10.1111/j.1467-9280.2006.01791.x
ATLAS Collaboration. (2010). Charged-particle multiplicities in pp inter-
actions at 公s⫽900 GeV measured with the ATLAS detector at the
LHC. Physics Letters, 688, 21– 42.
Babcock, L., Loewenstein, G., Issacharoff, S., & Camerer, C. (1995).
Biased judgments of fairness in bargaining. The American Economic
Review, 85, 1337–1343.
Brawley, L. R. (1984). Unintentional egocentric biases in attributions.
Journal of Sport Psychology, 6, 264 –278.
Brewer, M. B., & Gardner, W. (1996). Who is this “we”? Levels of
collective identity and self representations. Journal of Personality and
Social Psychology, 71, 83–93. http://dx.doi.org/10.1037/0022-3514.71
.1.83
Burger, J. M., & Rodman, J. L. (1983). Attributions of responsibility for
group tasks: The egocentric bias and the actor-observer difference.
Journal of Personality and Social Psychology, 45, 1232–1242. http://dx
.doi.org/10.1037/0022-3514.45.6.1232
Caruso, E., Epley, N., & Bazerman, M. H. (2006). The costs and benefits
of undoing egocentric responsibility assessments in groups. Journal of
Personality and Social Psychology, 91, 857– 871. http://dx.doi.org/10
.1037/0022-3514.91.5.857
Chambers, J. R., & Windschitl, P. D. (2004). Biases in social comparative
judgments: The role of nonmotivated factors in above-average and
comparative-optimism effects. Psychological Bulletin, 130, 813– 838.
http://dx.doi.org/10.1037/0033-2909.130.5.813
Converse, B. A., Lin, S., Keysar, B., & Epley, N. (2008). In the mood to
get over yourself: Mood affects theory-of-mind use. Emotion, 8, 725–
730. http://dx.doi.org/10.1037/a0013283
Epley, N., Keysar, B., Van Boven, L., & Gilovich, T. (2004). Perspective
taking as egocentric anchoring and adjustment. Journal of Personality
and Social Psychology, 87, 327–339. http://dx.doi.org/10.1037/0022-
3514.87.3.327
Evans, J. S. B. T. (2006). The heuristic-analytic theory of reasoning:
Extension and evaluation. Psychonomic Bulletin & Review, 13, 378 –
395. http://dx.doi.org/10.3758/BF03193858
Forsyth, D. R., Berger, R. E., Mitchell, T., & the Other-Serving Claims of
Responsibility on Attraction and Attribution in Groups. (1981). The
effects of self-serving vs. other-serving claims of responsibility on
attraction and attribution in groups. Social Psychology Quarterly, 44,
59 – 64. http://dx.doi.org/10.2307/3033865
Giladi, E. E., & Klar, Y. (2002). When standards are wide of the mark:
Nonselective superiority and inferiority biases in comparative judgments
of objects and concepts. Journal of Experimental Psychology: General,
131, 538 –551. http://dx.doi.org/10.1037/0096-3445.131.4.538
Hastie, R., & Kameda, T. (2005). The robust beauty of majority rules in
group decisions. Psychological Review, 112, 494 –508.
Klayman, J., & Burson, K. A. (2002). Looking for Lake Wobegon: Why
sometimes we’re all below average. Paper presented at the annual
conference of the Society for Judgment and Decision Making, Kansas
City, MO.
Kruger, J., & Gilovich, T. (1999). “Naïve cynicism” in everyday theories
of responsibility assessment: On biased assumptions of bias. Journal of
Personality and Social Psychology, 76, 743–753. http://dx.doi.org/10
.1037/0022-3514.76.5.743
Kruger, J., & Savitsky, K. (2009). On the genesis of inflated (and deflated)
judgments of responsibility. Organizational Behavior and Human De-
cision Processes, 108, 143–152. http://dx.doi.org/10.1016/j.obhdp.2008
.06.002
Kuiper, N. A., & Rogers, T. B. (1979). Encoding of personal information:
Self-other differences. Journal of Personality and Social Psychology,
37, 499 –514. http://dx.doi.org/10.1037/0022-3514.37.4.499
Leary, M. R., & Forsyth, D. R. (1987). Attributions of responsibility for
collective endeavors. In C. Hendrick (Ed.), Review of personality and
social psychology: Vol. 8.Group processes (pp. 167–188). Newbury
Park, CA: Sage.
Lin, S., Keysar, B., & Epley, N. (2010). Reflexively mindblind: Using
theory of mind to interpret behavior requires effortful attention. Journal
of Experimental Social Psychology, 46, 551–556. http://dx.doi.org/10
.1016/j.jesp.2009.12.019
Markus, H. (1977). Self-schemas and processing information about the
self. Journal of Personality and Social Psychology, 35, 63–78. http://dx
.doi.org/10.1037/0022-3514.35.2.63
McConnell, A. R., Sherman, S. J., & Hamilton, D. L. (1994). On-line and
memory-based aspects of individual and group target judgments. Jour-
nal of Personality and Social Psychology, 67, 173–185. http://dx.doi
.org/10.1037/0022-3514.67.2.173
Messick, D. M. (1993). Equality as a decision heuristic. In B. A. Mellers
& J. Baron (Eds.), Psychological perspectives on justice (pp. 11–31).
New York, NY: Cambridge University Press. http://dx.doi.org/10.1017/
CBO9780511552069.003
Miller, R. S., & Schlenker, B. R. (1985). Egotism in group members:
Public and private attributions of responsibility for group performance.
Social Psychology Quarterly, 48, 85– 89. http://dx.doi.org/10.2307/
3033785
Prentice, D. A., & Miller, D. T. (1992). When small effects are impressive.
Psychological Bulletin, 112, 160 –164. http://dx.doi.org/10.1037/0033-
2909.112.1.160
Riege, A. H., & Teigen, K. H. (2013). Additivity neglect in probability
estimates: Effects of numeracy and response format. Organizational
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
245
OVER-CLAIMING INCREASES WITH GROUP SIZE
Behavior and Human Decision Processes, 121, 41–52. http://dx.doi.org/
10.1016/j.obhdp.2012.11.004
Rogers, T. B., Kuiper, N. A., & Kirker, W. S. (1977). Self-reference and
the encoding of personal information. Journal of Personality and Social
Psychology, 35, 677– 688. http://dx.doi.org/10.1037/0022-3514.35.9
.677
Ross, M., & Sicoly, F. (1979). Egocentric biases in availability and
attribution. Journal of Personality and Social Psychology, 37, 322–336.
http://dx.doi.org/10.1037/0022-3514.37.3.322
Rottenstreich, Y., & Tversky, A. (1997). Unpacking, repacking, and an-
choring: Advances in support theory. Psychological Review, 104, 406 –
415. http://dx.doi.org/10.1037/0033-295X.104.2.406
Savitsky, K., Van Boven, L., Epley, N., & Wight, W. (2005). The unpack-
ing effect in allocations of responsibility for group tasks. Journal of
Experimental Social Psychology, 41, 447– 457. http://dx.doi.org/10
.1016/j.jesp.2004.08.008
Schlenker, B. R., & Miller, R. S. (1977). Egocentrism in groups: Self-
serving biases or logical information processing? Journal of Personality
and Social Psychology, 35, 755–764. http://dx.doi.org/10.1037/0022-
3514.35.10.755
Simmons, J. P., Nelson, L. D., Galak, J., & Frederick, S. (2011). Intuitive
biases in choice versus estimation: Implications for the wisdom of
crowds. Journal of Consumer Research, 38, 1–15. http://dx.doi.org/10
.1086/658070
Surowiecki, J. (2004). The wisdom of crowds. New York, NY: Doubleday.
Teigen, K. H., & Brun, W. (2011). Responsibility is divisible by two, but
not by three or four: Judgments of responsibility in dyads and groups.
Social Cognition, 29, 15– 42. http://dx.doi.org/10.1521/soco.2011.29
.1.15
Thompson, L. L., & Loewenstein, G. F. (1992). Egocentric interpretations
of fairness and interpersonal conflict. Organizational Behavior and
Human Decision Processes, 51, 176 –197. http://dx.doi.org/10.1016/
0749-5978(92)90010-5
Thompson, S. C., & Kelley, H. H. (1981). Judgments of responsibility for
activities in close relationships. Journal of Personality and Social Psy-
chology, 41, 469 – 477. http://dx.doi.org/10.1037/0022-3514.41.3.469
Tversky, A., & Koehler, D. J. (1994). Support theory: A nonextensional
representation of subjective probability. Psychological Review, 101,
547–567. http://dx.doi.org/10.1037/0033-295X.101.4.547
Walster, E., Walster, G. W., & Berscheid, E. (1978). Equity: Theory and
research. Boston, MA: Allyn & Bacon.
Windschitl, P. D., Kruger, J., & Simms, E. N. (2003). The influence of
egocentrism and focalism on people’s optimism in competitions: When
what affects us equally affects me more. Journal of Personality and
Social Psychology, 85, 389 – 408. http://dx.doi.org/10.1037/0022-3514
.85.3.389
Received July 14, 2015
Revision received December 30, 2015
Accepted January 9, 2016 䡲
This document is copyrighted by the American Psychological Association or one of its allied publishers.
This article is intended solely for the personal use of the individual user and is not to be disseminated broadly.
246 SCHROEDER, CARUSO, AND EPLEY
