ArticlePDF Available

Worth Keeping but Not Exceeding: Asymmetric Consequences of Breaking Versus Exceeding Promises


Abstract and Figures

Promises are social contracts that can be broken, kept, or exceeded. Breaking one's promise is evaluated more negatively than keeping one's promise. Does expending more effort to exceed a promise lead to equivalently more positive evaluations? Although linear in their outcomes, we expected an asymmetry in evaluations of broken, kept, and exceeded promises. Whereas breaking one's promise is obviously negative compared to keeping a promise, we predicted that exceeding one's promise would not be evaluated more positively than merely keeping a promise. Three sets of experiments involving hypothetical, recalled, and actual promises support these predictions. A final experiment suggests this asymmetry comes from overvaluing kept promises rather than undervaluing exceeded promises. We propse this pattern may reflect a general tendency in social systems to discourage selfishness and reward cooperation. Breaking one's promise is costly, but exceeding it does not appear worth the effort.
Content may be subject to copyright.
Social Psychological and Personality Science
The online version of this article can be found at:
DOI: 10.1177/1948550614533134
2014 5: 796 originally published online 8 May 2014Social Psychological and Personality Science
Ayelet Gneezy and Nicholas Epley
Worth Keeping but Not Exceeding: Asymmetric Consequences of Breaking Versus Exceeding Promises
Published by:
On behalf of:
Society for Personality and Social Psychology
Association for Research in Personality
European Association of Social Psychology
Society of Experimental and Social Psychology
can be found at:Social Psychological and Personality ScienceAdditional services and information for Alerts:
What is This?
- May 8, 2014OnlineFirst Version of Record
- Jul 18, 2014Version of Record >>
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
Worth Keeping but Not Exceeding:
Asymmetric Consequences of Breaking
Versus Exceeding Promises
Ayelet Gneezy
and Nicholas Epley
Promises are social contracts that can be broken, kept, or exceeded. Breaking one’s promise is evaluated more negatively than
keeping one’s promise. Does expending more effort to exceed a promise lead to equivalently more positive evaluations? Although
linear in their outcomes, we expected an asymmetry in evaluations of broken, kept, and exceeded promises. Whereas breaking
one’s promise is obviously negative compared to keeping a promise, we predicted that exceeding one’s promise would not be
evaluated more positively than merely keeping a promise. Three sets of experiments involving hypothetical, recalled, and actual
promises support these predictions. A final experiment suggests this asymmetry comes from overvaluing kept promises rather
than undervaluing exceeded promises. We suggest this pattern may reflect a general tendency in social systems to discourage
selfishness and reward cooperation. Breaking one’s promise is costly, but exceeding it does not appear worth the effort.
judgment and decision making, social cognition, social judgment, interpersonal processes, impression formation
It is often said that promises are easier to make than to keep.
Roosevelt promised the same for World War II, and George
Bush Sr. promised ‘‘no new taxes.’’ All were promises broken
while still in office. Although notorious for breaking promises,
politicians are not alone (Buehler, Griffin, & Ross, 1994;
Robinson & Rousseau, 1994). Contractors may promise to
complete projects on time and on budget but end up doing
neither. Employers may promise working conditions that go
unfulfilled. And friends may promise to help us exercise, clean
the apartment, and pick up groceries, only to leave us hungry
while exercising alone in a dirty apartment.
Clearly, not all promises are broken. Promises can be kept,
or even exceeded. Bush could have reduced taxes, contractors
could finish faster and cheaper, and employers could provide
even better working conditions (Lester, Turnley, Bloodgood,
& Bolino, 2002).
This research examines the social consequences of exceed-
ing versus breaking promises, compared to keeping promises.
We define a promise as a commitment to perform some specific
action made by one person to another. Promises are different
than having a belief or expectation about another’s behavior
because promises are inherently interpersonal, whereas a belief
or expectation is intrapersonal. Existing research demonstrates
many negative consequences from broken versus kept pro-
mises, including reduced trust, diminished satisfaction, and
increased turnover (see Rousseau, 1995, for a review). How-
ever, existing research does not measure whether putting in
additional effort to exceed a promise produces symmetrically
more positive consequences than merely keeping a promise.
Does doing more than one promised lead to even more positive
evaluations than simply keeping one’s promise?
We predicted it would not. In particular, we predicted that
exceeding one’s promise would not be evaluated more posi-
tively than merely keeping one’s promise. This predicted
asymmetry could emerge for two reasons.
First, keeping a promise could be overvalued compared to a
linear relationship between outcomes and evaluations. Overva-
luing kept promises is consistent with their contractual nature,
where keeping a promise provides relational value of trust and
reliability beyond the outcome’s objective value. Indeed, social
systems that place a premium on fairness may promote
mutually beneficial cooperation within groups (Bowles &
Gintis, 2003; Cosmides, 1989; Gneezy & Fessler, 2012; Tooby
& Cosmides, 1996).
Second, exceeding a promise could be relatively undervalued
compared to a linear relationship between outcomes and evalua-
tions. This undervaluation is consistent with the gain/loss
University of California San Diego, San Diego, CA, USA
University of Chicago, Chicago, IL, USA
Corresponding Author:
Ayelet Gneezy, Rady School of Management, 3W119 Wells-Fargo Hall, 9500
Gilman Drive, La Jolla, CA 92093-0553, USA.
Email: or
Social Psychological and
Personality Science
2014, Vol. 5(7) 796-804
ªThe Author(s) 2014
Reprints and permission:
DOI: 10.1177/1948550614533134
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
asymmetry predicted by Prospect Theory (Kahneman & Tversky,
1979), in which gains from reference points (such as a promise)
produce relatively little impact compared to losses from that ref-
erence point (Thaler, 1985). It is also consistent with the general
tendency for negative outcomes to produce stronger psychologi-
cal consequences than positive outcomes (Baumeister, Brat-
slavsky, Finkenauer, & Vohs, 2001; Rozin & Royzman, 2001).
This may be particularly true in moral domains where a negative
action, such as lying, is seen as more diagnostic of a person’s char-
acter than a negative action is in competence (or ability) domains,
such as performing poorly on a test (Martijn, Spears, Van der
Pligt, & Jakobs, 1992; Skowronski & Carlston, 1987).
We tested the asymmetry prediction in three sets of experi-
ments (Experiments 1a–3). A final experiment tested the
mechanism for our observed asymmetry: Whether it occurs
because keeping a promise is overvalued or because exceeding
promises are undervalued (Experiment 4). Because promises
arise in varying contexts that cannot be captured in any single
paradigm, we adopted a multimethod approach utilizing
hypothetical (Experiments 1a and 1b), remembered (Experi-
ments 2a, 2b, and 4), and actual promises (Experiment 3). Each
method has strengths and weaknesses. Hypothetical surveys
examine contexts that cannot be recreated experimentally but
utilize imagination rather than actual behavior. Memory is
notoriously imperfect, but it benefits from ecological validity
by examining actual events from real life. Experiments are the
gold standard for causality, but a constructed laboratory con-
text may also raise concerns about ecological validity. Testing
our hypotheses using all of these methods increases confidence
in the robustness of any observed results and addresses weak-
nesses from one methodology that is not contained in another.
Experiments 1a and 1b: Imagined Promises
Participants in Experiments 1a and 1b read hypothetical scenar-
ios in which they were asked to imagine receiving a promise
that was exceeded, kept as promised, or broken. We used these
scenarios to investigate how promise outcomes were likely to
influence promise receivers’ happiness and trust in the promise
maker (1a), and the impressions that promise receivers would
form of promise makers (1b).
Experiment 1a Procedure
Sixty-two University of California San Diego (UCSD) under-
graduates read a one-paragraph scenario in which another stu-
dent promised to help them by reading their term paper and
giving them comprehensive feedback. After reading details
about the circumstance, participants in the broken promise con-
dition read that ‘‘your friend returned the paper ...with com-
ments that were actually far less extensive than
promised ...and instead included only a general uninformative
comment at the top ...’’ Participants in the kept promise con-
dition instead read that ‘‘Your friend returned the paper ...with
comments that were exactly as promised ...’ Participants in the
exceeded condition read that ‘‘your friend returned the
paper ...with comments that were actually far more extensive
than promised ...catching problems with style, grammar, and
the gener al flow ...[and] also identified two major flaws ...and
offered very helpful alternative to strengthen the paper.’
Participants predicted how happy they would be, how likely
they were to trust the promise maker in the future, and how
interested they would be to help that individual in the future
(on 1–11 scales).
Experiment 1b Procedure
We approached 60 UCSD undergraduates in various locations
across campus that agreed to participate in exchange for candy.
Participants read the same scenario described in Experiment 1a
and then predicted the impression they would form of promise
makers on 27 personality traits (on 0–10 scales). Of these, nine
traits were of interest, measuring receivers’ perceptions of the
promise maker’s selfishness (unfair, unkind, and selfish), fair-
ness (just, fair, and equitable), and generosity (charitable, gen-
erous, and kind; see Appendix for the complete list).
Results and Discussion
Experiment 1a
We averagedparticipants’ responses to the three measuresinto an
overall positivity composite (a¼.95). Not surprisingly, partici-
pants evaluated breaking a promise more negatively (M¼4.40,
standard deviation [SD]¼1.51) than keeping a promise (M¼
9.70, SD ¼1.27), t(59) ¼11.90, p<.01, d¼3.75. More impor-
tant, participants did not evaluate exceeding a promise (M¼
9.74, SD ¼1.57) more positively than keeping it, t(59) ¼.08,
Experiment 1b
We averaged participants’ responses to the three selfish (a¼
.84), three fair (a¼.86), and three generous (a¼.88) items into
a composite for each trait type. As shown in Figure 1, partici-
pants rated promise makers as more selfish, less fair, and less
generous after breaking a promise than after keeping it, ts(57)
¼7.10, 4.05, and 4.58, respectively, ps<.01,ds¼5.31, 3.01,
and 3.07. More important, they rated promise makers as equally
(un)selfish, fair, and generous when they exceeded versus kept
their promise, ts(57) ¼.03, .13, and 1.52, respectively, ps > .1,
ds¼.02, .08, and 1.10 (see Figure 1).
We believe these two experiments suggest a psychological
asymmetry between breaking and exceeding a promise.
Whereas breaking a promise is evaluated negatively, exceeding
a promise is evaluated no more positively than merely keeping
a promise. One alternative explanation, however, is that this
asymmetry is an experimental artifact produced by a ceiling
effect on evaluations in the kept and exceeded conditions.
We address this potential artifact in three ways here and
revisit this issue later where applicable. First, we conducted
additional analyses of all experiments using both linear and
Gneezy and Epley 797
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
Tobit regressions. Tobit regressions (or censured regressions)
treat observations at the ceiling of a measure as equivalent to
that observation or greater. Tobit regressions are designed spe-
cifically to address ceiling effects, testing the effect of an inde-
pendent variable on a dependent variable if the measure’s
ceiling were removed. In all experiments (see Table 1), the lin-
ear and Tobit regressions yield nearly identical results, incon-
sistent with a ceiling effect artifact.
Second, not all measures yield the same percentage of obser-
vations at ceiling (see Table 1). In Experiment 1a, for instance,
30–35%of observations were at ceiling in the kept and exceeded
conditions. In Experiment 1b, only 5–10%of observations were
at ceiling in these conditions. However, both experiments show
the same patterns. If our results were constrained by a ceiling
effect, then results should vary as average responses move fur-
ther from the extreme point of a measure. But they do not.
Finally, a ceiling effect suggests that evaluations are
restricted in some conditions more than in others. If so, then the
variance in evaluations should be smaller in the conditions
restricted by a ceiling effect compared to conditions that are not
restricted by a ceiling effect. However, an inspection of the SDs
in Experiments 1a and 1b, as well as in all of the experiments
that follow, shows that the variance in evaluations is not system-
atically smaller in the kept and exceeded conditions than in the
broken conditions. This is inconsistent with a ceiling effect con-
straining evaluations in some conditions.
An artifact based on ceiling effects does not appear to
explain the results of Experiments 1a and 1b, nor the results
of the experiments that follow. Instead there appears to be an
asymmetry in evaluations of broken versus exceeded promises.
Breaking a promise is evaluated negatively compared to keep-
ing a promise, but exceeding one’s promise does not yield
markedly more positive evaluations than merely keeping a
Experiments 2a and 2b—Recalled Promises
Experiments 2a and 2b asked participants to recall promises.
Recalled promises have more ecological validity than hypothe-
tical promises, because they are memories of actual events
rather than imagined scenarios. Recalled promises are also
important because our memories of how others treated us may
influence decision making more than (forgotten) real-time eva-
luations (Wirtz, Kruger, Scollon, & Diener, 2003).
Participants in Experiment 2a recalled three promises: one
broken, one kept, and one exceeded. Experiment 2b provided
a replication using a between-participants’ manipulation of the
promise outcome, and an additional measure of how much
effort participants thought promise makers expended.
Experiment 2a Procedure
Fifty-nine University of Chicago undergraduates recalled pro-
mises someone made to them that were broken, kept, and
exceeded. For each event, participants described its details,
then reported their relationship with the promise maker, then
reported their happiness with the person’s behavior (1 ¼not
at all happy,11¼extremely happy), and finally reported how
difficult it was to recall the event (1 ¼notatall,11¼
Experiment 2b Procedure
Forty-five University of Chicago master of business adminis-
tration students completed a procedure similar to Experiment
2a, with three exceptions. First, participants recalled only one
promise (kept, broken, or exceeded). Second, after writing
down the details of the event, participants reported both how
happy they were at the time of the event (5¼extremely
unhappy,5¼extremely happy) and how happy they were
thinking back on that event (5¼extremely unhappy,5¼
extremely happy). Finally, participants reported how much
effort they thought promise makers invested in keeping the
promise (0 ¼none at all,10¼very much).
Results and Discussion
Experiment 2a
Two participants did not recall broken and exceeded promises,
and three additional participants did not recall an exceeded
promise. Analyzing only those who recalled all three events
(N¼54), participants reported feeling significantly less happy
when recalling broken versus kept promises (M
SD ¼1.53; M
¼8.04, SD ¼2.01), t(53) ¼15.34, p<.01,
d¼5.78. Participants also recalled being happier, although
to a lesser extent (only 13.3%of the size), when remembering
exceeded (M¼8.81, SD ¼2.43) versus kept promises, t(53) <
2.29, p<.05,d¼.77. Notice that relatively few observations
were at ceiling in the kept condition (9.5%, see Table 1), and that
the variance in responses was directionally larger in the
Broken Kept Exceeded
Impression of Promise Makers (1-11)
Figure 1. Promise receivers’ impressions of promise makers following
a hypothetical broken, kept, or exceeded promise (Experiment 1b).
798 Social Psychological and Personality Science 5(7)
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
exceeded promises condition, again suggesting that a ceiling
effect is not artifactually suppressing evaluations in the kept and
exceeded promises conditions. Including all participants in the
analyses does not alter these results meaningfully.
Experiment 2b
Participants’ recalled happiness at the time of the event and
thinking back on the event were highly correlated (r¼
.79), and so we averaged them into a composite for the fol-
lowing analyses. Participants were less happy recalling bro-
ken (M¼2.62, SD ¼1.35) versus kept promises (M¼
3.65, SD ¼1.74), t(42) ¼10.21, p<.01, d¼3.15. More
important, participants were no happier recalling exceeded
(M¼3.25, SD ¼1.70) versus kept promises, t(42) ¼.70,
ns,d¼.23. Figure 2 shows this same pattern for both
reported happiness measures.
Participants’ estimates of promise makers’ effort followed
the same pattern: participants believed promise makers
expended less effort in broken versus kept promises (Ms¼
1.67 vs. 6.82, respectively), t(42) ¼5.64, p<.01, d¼2.25, but
did not expend significantly more effort in exceeded (M¼
7.69) versus kept promises, t(42) ¼1.02, ns d ¼.32. If exceed-
ing a promise required more effort than keeping it, then prom-
ise receivers failed to recognize it.
Experiment 3—Actual Promises
Imagination and memory allowedusinExperiments1a2b
to measure evaluations of events that we could not replicate
in an actual experiment, but imagination is not reality and
memory is prone to distortions, calling into question the
generalizability of these experiments to actual promises that
might arise in social interactions. Although all existing stud-
ies of promise-keeping that we are aware of rely on either
imagined or recalled promises (Conway & Briner, 2002),
we sought in Experiment 3 to create a real promise that is
broken, kept, or exceeded under experimental conditions.
In Experiment 3, promise makers were instructed to make
a promise to another participant that they were subsequently
instructed to break, keep, or exceed in a clearly specified
way. This design allowed for experimental control over the
promises’ outcomes. In addition, this experimental method
allowed us to control the actual effort required under each
promise outcome and to ensure that the objective cost asso-
ciated with breaking a promise equaled the objective gain
from exceeding it. Finally, this method allows clear causal
inferences about the effects of a promise’s outcome on
Table 1. Regression Analyses and Percentage of Responses at Ceiling for Experiments 1a–4.
Experiment Comparison Linear Regression Tobit Regression % at Ceiling
1a Broken–Kept 5.30 (p< .01) 5.63 (p< .01) Kept 31.6
Kept–Exceed 0.04 (p¼.93) 0.06 (p¼.54, ns) Exceeded 33.3
1b Fairness Broken–Kept 2.65 (p< .01) 2.74 (p< .01) Kept 10
Kept–Exceed 0.08 (p>.1) 0.13 (p> .1) Exceeded 5
Generosity Broken–Kept 3.02 (p< .01) 3.11 (p< .01) Kept 5
Kept–Exceed 1.00 (p> .1) 1.13 (p> .1) Exceed 15
2a Broken–Kept 5.00 (p< .01) 5.34 (p< .01) Kept 9.3
Kept–Exceed 0.78 (p¼.05) 1.19 (p< .05) Exceeded 31.5
2b Broken–Kept 6.27 (p< .01) 6.73 (p< .01) Kept 33.3
Kept–Exceed .40 (p¼.45, ns).49 (p¼.75, ns) Exceeded 27.7
3 Broken–Kept 1.54 (p< .01) 2.60 (p< .01) Kept 28
Kept–Exceed .10 (p¼.88, ns).37(p¼.63, ns) Exceeded 29.33
4 Expectations Broken–Kept 4.08 (p< .01) 4.47 (p< .01) Kept 23.5
Kept–Exceed 1.92 (p< .01) 2.76 (p< .01) Exceeded 70
Promises Broken–Kept 6.87 (p< .01) 8.89 (p< .01) Kept 35.7
Kept–Exceed 0.26 (p> .1) 0.23 (p> .1) Exceeded 64.3
Broken Kept Exceeded
Happy (-5 to 5)
Happy Then
Happy Now
Figure 2. Promise receivers’ recalled happiness at the time of, and
now thinking back on, a broken, kept, and exceeded promise
(Experiment 2a).
Gneezy and Epley 799
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
One hundred and twenty University of Chicago undergraduates
(N¼60) participated in dyads. One participant in each dyad
was randomly assigned to be a promise receiver and the other
a promise maker. Promise receivers learned that they would
solve 40 puzzles (counting zeros in a table of numbers) and
be paid per puzzle solved. They learned that they would not
have enough time to solve all of the puzzles, but that they
would be paired up with another participant (the promise
maker) who would decide whether or not to help them com-
plete more puzzles.
Promise makers learned that they could choose to help the
other participant solve tables. If willing to help, they were told
they would promise the other participant to solve 10 puzzles.
We informed all participants that the promise makers’ help
would only benefit the promise receivers. All promise makers
agreed to help, walked to the other participant’s room, and
verbally promised to solve 10 tables following a script provided
by the experimenter.
To manipulate promise keeping, the experimenter then
instructed promise makers to solve 10 tables (as promised),
5 tables, or 15 tables. Promise makers worked for 7 min, dur-
ing which promise receivers completed a personality inven-
tory (the Five Factor Model; John, Naumann, & Soto,
2008). Next, the experimenter delivered 5, 10, or 15 (pre)-
solved tables to promise receivers (with the remaining
unsolved), ostensibly completed by the promise maker.The
experimenter said, ‘‘while you were filling out the personality
test, your partner solved 5 tables, instead of the 10 tables he or
she promised to solve (or 10 tables, exactly as he or she prom-
ised; or 15 tables, instead of the 10 tables he or she promised
to solve).’
Promise receivers then completed a survey measuring the
effort they thought promise makers expended in keeping the
promise (1 ¼very little effort,11¼extreme effort), to what
extent they thought promise makers intended to keep their
promise (1 ¼not at all,11¼very much), how happy they
were (5¼extremely unhappy,5¼extremely happy), and
how grateful they were for their help (1 ¼notatallgrateful,
11 ¼very much grateful). Promise receivers were then
allowed to solve as many tables as they wanted. Just before
being paid, promise receivers completed the same survey
measures again to test whether their final evaluations
depended on their earnings.
Results and Discussion
We reverse scored and transformed happiness ratings to a 1–11
scale. Reported happiness and gratitude were highly correlated
both before and after being paid (r¼.79 and .80, respectively).
We analyze before and after ratings separately below to test for
possible main effects or interactions.
A 2(Evaluation: before vs. after payment) 3(Promise: bro-
ken, kept, exceeded) analysis of variance (ANOVA) on overall
positivity yielded only a significant main effect for promise,
F(2, 114) ¼6.37, p<.01, Z
¼.10, with no main effect of eva-
luation or interaction (ps > .1). Again, promise receivers eval-
uated a broken promise less positively (M¼8.32, SD ¼2.58)
than a kept promise (M¼9.88, SD ¼1.80), t(57) ¼2.34, p<
.05, d¼.74, but did not evaluate exceeded promises more posi-
tively (M¼9.77, SD ¼2.05) than kept promises, ns,d¼.06.
The same ANOVA on perceived effort yielded only a signif-
icant main effect for promise, F(2, 114) ¼20.06, p<.01, Z
.26, with no main effect of evaluation or interaction (ps > .1).
Participants believed promise makers expended less effort
when the promise was broken (M¼6.76, SD ¼2.13) versus
kept (M¼9.55, SD ¼1.68), t(57) ¼4.99, p<.01, d¼1.45,
but did not believe promise makers expended more effort when
the promise was exceeded (M¼9.50, SD ¼1.37) versus kept,
t(57) ¼.09, ns,d¼.03.
Finally, the same ANOVA on intent to keep the promise
yielded only a significant main effect for Promise, F(2, 114)
¼7.57, p<.01, Z
¼.11, with no main effect of evaluation
or interaction (ps > .1). Participants assumed a stronger inten-
tion to keep the promise when it was kept (M¼10.18, SD ¼
1.23) versus broken (M¼8.31, SD ¼2.09), t(57) ¼2.96, p
< .01, d¼1.12, but did not attribute stronger intentions when
a promise was exceeded (M¼9.45, SD ¼2.45) versus kept,
t(57) ¼1.14, ns,d¼.40.
This experiment replicated an asymmetry in evaluations
of broken versus exceeded promises using actual promises.
Unlike the preceding experiments, we observed this asym-
metry despite the obvious (and symmetrical) increase in
effort required of promise makers to exceed a promise than
to keep it. Exceeding a promise clearly required more effort
and produced more benefit than merely keeping a promise,
and yet promise receivers evaluated them equally.
Experiment 4—Promises Versus Reference
The asymmetry observed in our experiments could emerge
through at least two mechanisms. First, kept promises could
be evaluated more positively than a perfectly linear relation-
ship between outcomes and evaluations would produce. Pro-
mises on this account serve as social contracts, in which the
value of keeping a promise goes beyond the objective benefit
provided, thereby increasing its perceived value beyond its
objective benefit. Second, exceeding a promise could be eval-
uated less positively than a linear relationship would produce.
Promises could serve as reference points such that gains from a
reference point have less impact on evaluations than losses
from a reference point (Kahneman & Tversky, 1979). On this
account, keeping a promise is evaluated in line with its objec-
tive outcome but exceeding a promise is undervalued compared
to objective benefit.
We tested these mechanisms by comparing promises with
expectations. Although both promises and expectations pro-
duce a reference point in judgment, they differ because a
promise is interpersonal—a commitment made from one
800 Social Psychological and Personality Science 5(7)
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
person to the other—whereas an expectation is intraperso-
nal—a belief held by only one person (i.e., the ‘‘receiver’’).
Promises are therefore social contracts whereas expectations
are merely reference points of comparison for evaluating
outcomes. We therefore predicted that expectations would
produce a relatively linear relationship between outcomes
and evaluations, providing a benchmark of comparison for
the asymmetry observed between promised outcomes and
evaluations. If keeping one’s promise is overvalued, then
a kept promise should be evaluated more favorably than a
met expectation. If, however, exceeding one’s promise is
undervalued, then an exceeded promise should be evaluated
less positively than an exceeded expectation.
UCSD undergraduates (N¼143) were randomly assigned to
recall a time when someone broke, kept, or exceeded a
promisemadetothem,or a time they expected someone
to do something that the other person fell short of, met,
or exceeded. Participants assigned to the promises condi-
tions received instructions similar to those in Experiment
2a and 2b—to recall a promise someone made to them that
was broken (someone did less than promised), kept (some-
one did exactly as promised), or exceeded (someone did
more than promised). Participants in the expectations condi-
expected someone to do something and the person did less
than expected, did exactly what was expected, or did more
than expected. After describing the event in writing, partici-
pants reported how happy (5¼extremely unhappy,5¼
extremely happy) and how disappointed (1 ¼not at all dis-
appointed,11¼extremely disappointed)theywerewiththe
person’s behavior.
Results and Discussion
We reverse scored and transformed disappointment ratings to
ings to create a positivity composite (r¼.84). A 2 (Context:
expectation vs. promise) 3 (Outcome: broken, kept/met,
exceeded) ANOVA on this composite revealed a significant
effect for outcome, F(2, 137) ¼117.93, p<.01, Z
qualified by the predicted interaction, F(2, 137) ¼5.20, p<
.01, Z
Participants’ responses to expectations were linearly
related to the outcome but responses to promises were not
(Figure 3). When recalling expectations, participants were
more positive when another person met versus fell short
of their expectation, t(69) ¼6.24, p<.01, d¼1.53, and
were also more positive when expectations were exceeded
versus met, t(69) ¼2.99, p<.01, d¼1.04. When consid-
ering promises, however, an asymmetry emerged: partici-
pants were less positive when a promise was broken than
when kept, t(68) ¼10.77, p<.01, d¼4.68, but were not
more positive when a promise was exceeded versus kept,
t(68) ¼.45, ns,d¼.14. More important, participants were
more positive when a promise was kept than when an
expectation was met, t(42) ¼3.19, p<.01, d¼1.10, the
only instance where evaluations of promises and expecta-
tions diverged.
These results are inconsistent with the reference point
mechanism derived from Prospect Theory, and instead support
the social contract mechanism in which keeping promises is
relatively overvalued. Promise receivers appear to place a pre-
mium on keeping one’s promise, a premium that comes from
fulfilling a social contract, with no added value from exceeding
one’s promise.
General Discussion
Four sets of experiments demonstrate an asymmetric evalua-
tion of promises. Although broken promises are evaluated
negatively, exceeded promises are not evaluated more posi-
tively than merely keeping one’s promise. Promise receivers
consistently failed to recognize the additional effort required
to exceed a promise but recognized a lack of effort when
breaking a promise, even when effort and benefits associ-
ated with exceeding versus breaking a promise were clearly
This pattern does not seem to reflect a simple ceiling effect
on evaluations in the kept and exceeded conditions. Tobit
regressions yield the same results as linear regressions, the
percentage of evaluations at ceiling varies across our experi-
ments but the same pattern emerges consistently, and the var-
iance in evaluations is not systematically smaller in the
exceeded promises condition than in the broken promises
Broken Kept Exceeded
Positivity (-5 to 5)
Figure 3. Promise receivers’ evaluations of broken, kept, or exceed
promises versus expectations (Experiment 4).
Gneezy and Epley 801
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
Instead, Experiment 4 suggests the asymmetry in evalua-
tions is caused by overvaluing kept promises, suggesting the
contractual nature of a promise provides value above and
beyond a promise’s tangible outcome. This premium on keep-
ing one’s promise may reflect a broader social system meant
to sustain cooperation for everyone’s benefit, one highly sen-
sitive to signals of trustworthiness (Almenberg, Dreber, Api-
cella, & Rand, 2011). Indeed, existing research suggests that
social systems tend to be hypersensitive to signals of dishon-
esty and unfairness but do not necessarily reward generosity
(Alexander, 1987; Keysar, Converse, Wang, & Epley, 2008;
Klein & Epley, 2014; Tooby & Cosmides, 1996).
We believe this observed ‘‘fairness premium’’ may reflect a
more general pattern in social judgment. Consider a follow-up
experiment testing this possibility. We asked 50 MTurk parti-
cipants to rate the desirability of each trait used in Experiment
1b on scales ranging from 1 (extremely undesirable)to10
(extremely desirable). We used the same nine traits to create
measures of selfishness (unfair, unkind, and selfish; a¼.91),
fairness (just, fair, and equitable; a¼.71), and generosity
(charitable, generous, kind; a¼.81). Consistent with a fairness
premium, participants evaluated selfish traits (M¼3.63, SD ¼
2.37) as significantly less desirable than fairness traits (M¼
7.55, SD ¼1.59), t(49) ¼8.83, p<.01, d¼4.43, but did not
rate generosity traits as more desirable (M¼7.62, SD ¼1.57)
than fairness traits, t(49) ¼.35, ns,d¼.09. Generous traits
were not evaluated more positively than fair traits, but selfish
traits were evaluated more negatively than fair traits. We know
of one other result consistent with these findings. In an experi-
ment involving a dictator game (Almenberg et al., 2011), par-
ticipants could punish or reward another person for being
selfish, fair, or generous to another participant. Results showed
that participants punished others for being selfish, but rewarded
them equally for being fair or generous. Placing a premium on
fairness, but no additional benefit for generosity, may represent
a more general pattern of behavior in a social system designed
to enable cooperation among unrelated individuals.
Although our experiments assessed promises between
individuals, we expect this asymmetry extends to promises
involving more abstract relationships, such as between cus-
tomers and companies or employees and employers (Con-
way & Briner, 2002). Businesses may work hard to
exceed their promises to customers or employees (Lester
et al., 2002), but our research suggests that this hard work
may not produce the desired consequences beyond those
obtained by simply keeping promises. As an initial test of
this possibility, we asked [university] undergraduates to
imagine that they had bought tickets to a concert from an
online company. They imagined purchasing tickets for Row
10, but then either received worse tickets than promised
(Row 11, 13, or 15), better tickets than promised (Row 9,
7, or 5), or exactly what was promised (Row 10). Partici-
pants then indicated how satisfied they would be, how likely
they were to recommend the company to a friend, and how
likely they were to use the same seller in the future (a¼
.97). Once again, participants were more negative when
they received worse tickets than promised (regardless of
how much worse; M¼2.05, SD ¼.74) than when they got
exactly what was promised (M¼7.97, SD ¼1.34), t(86) ¼
16.32, p<.01. Participants were no more positive, however,
when they received better tickets than promised (regardless
of how much better) compared to exactly what was prom-
ised. In fact, they were somewhat more negative (M¼
7.20, SD ¼2.02), t(86) ¼2.03, p<.05.
Promises can be hard to keep, and promise makers
should spend their effort keeping them wisely. The results
of our experiments suggest that it is wise to invest effort
in keeping a promise because breaking it can be costly, but
it may be unwise to invest additional effort to exceed one’s
promises. When companies, friends, or coworkers put forth
rewarded. But when they expend extra effort in order to
exceed those promises, their effort appears likely to be
802 Social Psychological and Personality Science 5(7)
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
We thank Jasmine Kwong, Entzu Lin, Chau Ngo, Silvia Saccardo, and
Beverly Tam for assistance conducting this research; Katherine
Burson and Karsten Hansen for data analysis assistance; Tal Eyal,
Tom Gilovich, Elizabeth Keenan, Alex Imas, and John Skowronski
for helpful comments. Experiments 2a and 2b were included in
Ayelet Gneezy’s PhD dissertation.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to
the research, authorship, and/or publication of this article.
The author(s) received no financial support for the research, authorship,
and/or publication of this article.
1. In all experiments, we collected evaluations from the perspective
of both promise receivers and promise makers. The promise mak-
er’s perspective tested a distinct hypothesis about whether promise
makers would underestimate the negative consequences of break-
ing a promise. All but one experiment supports this hypothesis sig-
nificantly. Although interesting, the promise makers’ evaluations
are not central to our hypotheses about an asymmetry in evalua-
tions for promise receivers. Because they neither qualify nor add
to the results from promise receivers, we present the promise mak-
ers’ procedures and results in the Online Supplemental Material.
Supplemental Material
The online data supplements are available at
Alexander, R. D. (1987). The biology of moral systems. London, Eng-
land: Aldine.
Almenberg, J., Dreber, A., Apicella, C. L., & Rand, D. G. (2011).
Third party reward and punishment: group size, efficiency and
public goods. Psychology and punishment (pp. 73–92). Haup-
pauge, NY: Nova Publishers.
Baumeister, R. F., Bratslavsky, E., Finkenauer, C., & Vohs, K. D.
(2001). Bad is stronger than good. Review of General Psychology,
5, 323–370.
Bowles, S., & Gintis, H. (2003). Origins of human cooperation. In P.
Hammerstein (Ed.), Genetic and cultural evolution of cooperation
(pp. 429–443). Cambridge, MA: MIT Press.
Buehler, R., Griffin, D., & Ross, M. (1994). Exploring the ‘‘plan-
ning fallacy’’: Why people underestimate their task completion
times. Journal of Personality and Social Psychology,67,
Conway, N., & Briner, R. B. (2002). A daily diary study of affective
responses to psychological contract breach and exceeded promises.
Journal of Organizational Behavior,23, 287–302.
Cosmides, L. (1989). The logic of social exchange: Has natural selec-
tion shaped how humans reason? Studies with the Wason selection
task. Cognition,31, 187–276.
Gneezy, A., & Fessler, D. T. (2012). Conflict, sticks and carrots: War
increases prosocial punishments and rewards. Proceedings of the
Royal Society B,279, 219–223.
John, O. P., Naumann, L. P., & Soto, C. J. (2008). Paradigm shift to the
integrative Big-Five trait taxonomy: History, measurement, and
conceptual issues. In O. P. John, R. W. Robins, & L. A. Pervin
(Eds.), Handbook of personality: Theory and research (pp.
114–158). New York, NY: Guilford Press.
Kahneman, D., & Tversky, A. (1979). Prospect theory: An analysis of
decisions under risk. Econometrica,47, 313–327.
Keysar, B., Converse, B. A., Wang, J., & Epley, N. (2008). Recipro-
city is not give and take: Asymmetric reciprocity to positive and
negative acts. Psychological Science,19, 1280–1286.
Klein, N., & Epley, N. (2014). The topography of generosity: Asym-
metric evaluations of prosocial actions. Manuscript submitted for
publication. University of Chicago, Chicago, IL.
(2002). Not seeing eye to eye: Differences in supervisor and
subordinate perceptions of and attributions for psychological
contract breach. Journal of Organizational Behavior,23,
Martijn, C., Spears, R., Van der Pligt, J., & Jakobs, E. (1992). Nega-
tivity and positivity effects in person perception and inference:
Ability versus morality. European Journal of Social Psychology,
22, 453–463.
Robinson, S. L., & Rousseau, D. M. (1994). Violating the psychologi-
cal contract: Not the exception but the norm. Journal of Organiza-
tional Behavior,21, 525–546.
Rousseau, D. M. (1995). Psychological contracts in organizations:
Understanding written and unwritten agreements. Thousand Oaks,
CA: Sage.
Table A1. Full List of Traits, Experiment 1b.
Alert Capable Fair Efficient Friendly Brilliant
Kind Selfish Practical Generous Imaginative Skillful
Unkind Progressive Artistic Charitable Lucky Energetic
Open-minded Humble Equitable Sophisticated Unfair Wise
Sympathetic Just Talented
Gneezy and Epley 803
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
Rozin, P., & Royzman, E. B. (2001). Negativity bias, negativity dom-
inance, and contagion. Personality and Social Psychology Review,
5, 296–320.
Skowronski, J. J., & Carlston, D. E. (1987). Social judgment and
social memory: The role of cue diagnosticity in negativity, positiv-
ity, and extremity biases. Journal of Personality and Social
Psychology,52, 689–699.
Thaler, R. H. (1985). Mental accounting and consumer choice.
Marketing Science,4,199214.
Tooby, J., & Cosmides, L. (1996). Friendship and the banker’s paradox:
Other pathways to the evolution of adaptations for altruism. In W.
Runciman, J. Maynard Smith, & R. Dunbar (Eds.), Evolution of
social behavior patterns in primates and man. Proceedings of the
British Academy (Vol. 88, pp. 119–143).
do on spring break? The role of predicted, online, and
remembered experience on future choice. Psychological Sci-
ence,14, 520–524.
Author Biographies
Ayelet Gneezy is an Associate professor of behavioral sciences and
marketing at UCSD. Her research addresses a wide variety of ques-
tions pertaining to consumer behavior such as behavioral pricing and
social preferences. In her research, professor Gneezy often collabo-
rates with firms and integrates field experiments to test her
Nicholas Epley is the John T. Keller Professor of Behavioral Science
at the University of Chicago Booth School of Business. His research
is focused on the experimental study of social cognition, perspective
taking, and intuitive human judgment.
804 Social Psychological and Personality Science 5(7)
at UNIV OF CHICAGO LIBRARY on July 18, 2014spp.sagepub.comDownloaded from
... However, there is mixed support as to whether exceeding one's promises actually generates the purported benefits. The work of Gneezy and Epley (2014) indicates that failure to keep promises has a significant detrimental impact on relationships, while exceeding promises only has minor positive effects. In fact, under-promising may hurt a firm by signaling a quality level below what is adequate for some customers (Chakravarti et al., 2006;Koku, 1997) leading to them removing the firm from consideration. ...
... First, a theoretical contribution to the disconfirmation paradigm in services and sales studying the joint effect of the initial expectation levels, the promise outcome and the expectation inducing medium is provided. This approach also contributes to the promise-keeping theory in psychology (Gneezy and Epley, 2014) by relying on the role of ideal versus minimal expectations to explain how exceeding promises may lead to better perceptions compared to keeping promises. Second, the results provide a caveat to the viability of the well-regarded UPOD strategy in services and reconcile some of the findings in the marketing and psychology literatures. ...
... Recent evidence from the psychology literature shows that exceeding promises are not necessarily better than keeping promises. Drawing on the prospect theory (Kahneman and Tversky, 1979), Gneezy and Epley (2014) study asymmetric consequence effects of keeping, breaking and exceeding promises. The authors find that it is worth keeping promises, yet exceeding promises are not significantly better than keeping promises. ...
Purpose The paper aims to provide a theoretical and empirical examination of the relationship between service expectation management, expectation inducing agent and customer satisfaction. Design/methodology/approach Based on the disconfirmation paradigm in services and the promise-keeping premise in psychology, the hypotheses are developed and empirically tested using three experiments that manipulated expectations, expectation inducing agent and service outcome. Findings The findings provide reconciliation to the previous studies in services and show that effectiveness of expectation management strategy depends on the individual expectation thresholds and the expectation inducing agent. If customers patronize a firm expecting more, then over-delivering on the service promise results in a significant benefit. However, for those customers whose mental expectation threshold is exceeded, keeping promises is as effective as exceeding promises. Practical implications The practical implication of this paper is that services managers should be cognizant of the mental expectation threshold of customers and be wise in utilizing the under-promise, over-deliver strategy. Originality/value Using a threshold approach, this paper introduces a new perspective to service practitioners who are trying to manage expectations in a highly variable business environment. It also benefits service researchers who are trying to enhance the understanding of service expectation management.
... Derogatory terms such as "do-gooder" or "goodness-junkie" suggest that helping, at times, does not only "not improve", but even worsen the helper's reputation (Minson & Monin, 2012;Monin, 2007;Monin et al., 2008;Silver et al., 2021). For example, although people like helpers and promisekeepers, they do not like those who help too much or those who exceed their promises (Gneezy & Epley, 2014;Klein & Epley, 2014;Pleasant & Barclay, 2018). ...
Full-text available
Charitable giving, volunteering, climate-friendly choices, and most recently changing one’s lifestyle to stop the spread of the coronavirus are all examples of prosocial behavior. Prosociality can be investigated from different perspectives including the “who-question” (which people are more likely to help), and the “when-question” (which situational factors stimulate helping?), but in this article we focus primarily on the “why-question” (which emotions and cognitions motivate helping?)Specifically, this article tries to organize and synthesize literature related to emotions, thoughts, and beliefs (i.e. psychological mechanisms) that motivate or demotivate human helping behavior. To do this, we present a new typology including four overarching interrelated categories, each encompassing multiple subcategories.(1) Emotions: (a) emotional reactions elicited by the need situation such as empathic concern/sympathy, (b) positive or negative attitudes toward the beneficiary or the requester, (c) incidental mood. (2) Moral principles: (a) personal responsibility, (b) fairness-concerns, (c) aversion towards causing harm. (3) Anticipated impact: (a) self-efficacy (e.g. “can I make a difference?”) and (b) response-efficacy (e.g., “is this cause/project efficient and worthwhile?”). (4) Anticipated personal consequences: (a) material, (b) social and (c) emotional costs and benefits that the helper expects will follow if she helps or if she does not help. Increased knowledge about the “who” (e.g. individual differences in demography or personality) and “when” (situational antecedents such as characteristic of those in need, or type of solicitation) can surely help predict and even increase prosociality, but we argue that to understand the psychology of helping we need to also consider the psychological mechanisms underlying prosocial decisions (the “why-question”).We compare our typology against related theoretical frameworks, and present the pros and cons with different methodological approaches of testing psychological mechanisms of helping, with the aim to help researchers and practitioners better organize and understand the many psychological factors that influence prosocial decisions.
... One exception is Krull, Seger and Silvera [43] who manipulated the amount of help and the "willingness to help" and found that amount of help only predicted impressions if the help was given willingly (e.g. while smiling). Also, research investigating cooperation in economic games suggest that evaluations are no more positive when people help more rather than less [44][45][46], and that we sometimes want to avoid overly generous people as we believe they establish undesired norms [47]. In a recent opinion article [27], Anderson et al. suggest that moral praise is less magnitude-dependent than moral blame, meaning that harming much is considered worse than harming some whereas helping much is considered equally good as helping some. ...
Full-text available
Impression of helpers can vary as a function of the magnitude of helping (amount of help) and of situational and motivational aspects (type of help). Over three studies conducted in Sweden and the US, we manipulated both the amount and the type of help in ten diverse vignettes and measured participants' impressions of the described helpers. Impressions were almost unaffected when increasing the amount of help by 500%, but clearly affected by several type of help-manipulations. Particularly, helpers were less positively evaluated if they had mixed motives for helping, did not experience intense emotions or empathy, or if helping involved no personal sacrifice. In line with the person-centered theory of moral judgment , people seem to form impressions of helpers primarily based on the presumed underlying processes and motives of prosociality rather than its consequences.
... However, they fail to judge unintended, obligation-upholding side effects to be equally praiseworthy (Knobe 2003a;2003b;Leslie et al. 2006). People also more readily incorporate the magnitude of the consequences of proscriptive obligations than prescriptive obligations into their judgments of the agent (Gneezy & Epley 2014;Klein & Epley 2014). Together, this research suggests that judgments of proscriptive obligations incorporate factors related to causality, intentionality, and consequences more than do judgments of prescriptive obligations. ...
Satisfying one's obligations is an important part of being human. However, people's obligations can often prescribe contradictory behaviors. Moral obligations conflict (loyalty vs. fairness), and so do obligations to different groups (country vs. family when one is called to war). We propose that a broader framework is needed to account for how people balance different social and moral obligations.
... Actor-observer asymmetries. By adopting a social exchange perspective on transgressions, the current research also provided a noteworthy extension to the idea that social roles govern perceptions and attributions (e.g., Flynn & Lake, 2008;Gneezy & Epley, 2014;Keysar, Converse, Wang, & Epley, 2008;Zhang & Epley, 2012). Whether givers or receivers, actors or observers, or victims and transgressors, an important implication of this research is that victims and perpetrators-just as other people in these social roles-are not different kinds of people. ...
We investigate the possibility that victims and transgressors are predictably miscalibrated in their interpretation of a transgression, and that this has important implications for the process of forgiveness. Across 5 studies, we find that victims underestimate how much transgressors desire forgiveness. This is driven by a 2-part mediating mechanism: First, victims are more likely than transgressors to see the transgression as intentional, and second, this causes victims to believe transgressors feel less guilty than transgressors report feeling. Ultimately, this chain of asymmetries stymies the processes of forgiveness because victims tend to withhold forgiveness from those who actually desire it. The predicted effect emerged in the context of scenario studies (Studies 3 and 5), a real transgression that occurred in the lab (Study 4), transgressions from participants' pasts (Study 1), and transgressions from the same day (Study 2). In Study 4, we describe a new procedure in which 1 participant commits a real transgression against another participant, providing an effective means for researchers to study real-time transgressions from the perspective of both parties involved. Furthermore, in Study 5, we found that when victims were encouraged to empathize with the transgressor, the asymmetries were attenuated, suggesting a means of overcoming this impediment to forgiveness. (PsycINFO Database Record
Full-text available
Homelessness is an increasing social challenge both locally and globally. Social problems associated with street homelessness and services provided to homeless people by social workers are well documented in South Africa. Little is known though concerning the collaborative role of social workers and homeless people in addressing street homelessness. This study explores the role of social workers in promoting social justice through a collaborative purpose amongst the homeless people in addressing street homelessness. A qualitative, explorative, descriptive, and contextual study is undertaken with 14 homeless participants aged 21 to 63, and eight social work participants aged 25 to 38. This study is informed by an amalgamation of two theories, indicating the collaboration theory and the structural theory. Data are collected through semi-structured interviews, analysed by following Tesch’s eight steps (Creswell, 2009:186). The data are collected and verified, employing the Lincoln and Guba’s model of trustworthiness. The ethical considerations adhered throughout this study are informed consent, confidentiality and anonymity, management of information and debriefing. The major findings identify first, poverty and unemployment as major homelessness causes. Secondly, deprived, or non-existing family support, family or marital breakdown, substance use, migration of individuals from rural or foreign countries to city centres, as major contributing factors towards homelessness. Thirdly, homeless people are more vulnerable to personal harm than the housed population. Fourthly, the society and derogatory labels, such as crazy, nyaopes, and criminals are often used to describe these individuals. Finally, homelessness places homeless people at substantial risk of elevated mental health conditions. Implications for social work and recommendations for future research are presented. Key concepts: Collaborative, role, social workers, homeless people, homeless, homelessness, poor, underprivileged, street homelessness.
How do people judge whether someone deserves moral praise for their actions? In contrast to the large literature on moral blame, work on how people attribute praise has, until recently, been scarce. However, there is a growing body of recent work from a variety of subfields in psychology (including social, cognitive, developmental, and consumer) suggesting that moral praise is a fundamentally unique form of moral attribution and not simply the positive moral analogue of blame attributions. A functional perspective helps explain asymmetries in blame and praise: we propose that while blame is primarily for punishment and signaling one’s moral character, praise is primarily for relationship building.
Is it particularly human to feel coerced into fulfilling moral obligations, or is it particularly human to enjoy them? I argue for the importance of taking into account how culture promotes prosocial behavior, discussing how Confucian heritage culture enhances the satisfaction of meeting one's obligations.
Full-text available
Prosociality is considered a virtue. Those who care for others are admired, whereas those who care only for themselves are despised. For one's reputation, it pays to be nice. Does it pay to be even nicer? Four experiments assess reputational inferences across the entire range of prosocial outcomes in zero-sum interactions, from completely selfish to completely selfless actions. We observed consistent nonlinear evaluations: Participants evaluated selfish actions more negatively than equitable actions, but they did not evaluate selfless actions markedly more favorably than equitable actions. This asymptotic pattern reflected monotonic evaluations for increasingly selfish actions and insensitivity to increasingly selfless actions. It pays to be nice but not to be really nice. Additional experiments suggest that this pattern stems partly from failing to make spontaneous comparisons between varying degrees of selflessness. We suggest that these reputational incentives could guide social norms, encouraging equitable actions but discouraging extremely selfless actions. (PsycINFO Database Record (c) 2014 APA, all rights reserved).
Full-text available
In this article, we report two experiments that examine a cue-diagnosticity explanation (Skowronski & Carlston, 1986) for the extremity and negativity biases that are commonly found in impression-formation research. The first experiment demonstrates that negative behaviors are generally perceived as more diagnostic than positive behaviors when the former are morality related (honesty-dishonesty) but that these perceptions are reversed when negative behaviors are ability related (intelligence-stupidity). Furthermore, as predicted, extreme behaviors relating to either morality or ability were perceived as more diagnostic than were moderate behaviors. As a consequence of these perceptions, it was predicted that the impression-formation task used in Experiment 2 would detect negativity and extremity biases for the morality judgment, but positivity and extremity biases for the ability judgment. The data confirmed these predictions, providing support for the cue-diagnosticity approach and posing problems for alternative approaches to judgment biases. Experiment 2 also found recall to be characterized by negativity biases for stimulus behaviors relating to both morality and ability. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Full-text available
Tested 3 hypotheses concerning people's predictions of task completion times: (1) people underestimate their own but not others' completion times, (2) people focus on plan-based scenarios rather than on relevant past experiences while generating their predictions, and (3) people's attributions diminish the relevance of past experiences. Five studies were conducted with a total of 465 undergraduates. Results support each hypothesis. Ss' predictions of their completion times were too optimistic for a variety of academic and nonacademic tasks. Think-aloud procedures revealed that Ss focused primarily on future scenarios when predicting their completion times. The optimistic bias was eliminated for Ss instructed to connect relevant past experiences with their predictions. Ss attributed their past prediction failures to external, transient, and specific factors. Observer Ss overestimated others' completion times and made greater use of relevant past experiences. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Full-text available
The classical definition of altruism in evolutionary biology requires that an organism incur a fitness cost in the course of providing others with a fitness benefit. New insights are gained, however, by exploring the implications of an adaptationist version of the 'problem of altruism,' as the existence of machinery designed to deliver benefits to others. Alternative pathways for the evolution of altruism are discussed, which avoid barriers thought to limit the emergence of reciprocation across species. We define the Banker's Paradox, and show how its solution can select for cognitive machinery designed to deliver benefits to others, even in the absence of traditional reciprocation. These models allow one to understand aspects of the design and social dynamics of human friendship that are otherwise mysterious. (PsycINFO Database Record (c) 2012 APA, all rights reserved)
Introduction Contracting A Modern Dilemma Contract Making The Contract Makers Contemporary Contracts Violating the Contract Changing the Contract Business Strategy and Contracts Trends in the New Social Contract
The present paper deals with negativity and positivity effects in trait inferences and impression formation. In the first experiment we tested the suggestion of Skowronski and Carlston (1987) that in the domain of morality negative information is more diagnostic, will therefore receive more weight and result in a negativity effect whereas in the domain of abilities, positive information is more diagnostic resulting in positivity effects. Results of our first experiment support these predictions: negative behavioural information leads to more certain inferences concerning morality and positive behavioural information leads to more certain inferences concerning ability. In a second experiment, we investigated the relative weight of positive versus negative ability-and morality-related traits in an impression formation task. We counterposed traits from both morality and ability domains to see which was the most dominant in determining evaluative impressions. Findings of this second experiment showed strong negativity effects but also revealed that information related to morality is more influential in forming an evaluative impression than equally extreme information related to ability. Theoretical implications of these findings are discussed.