© The Author(s) 2017
Reprints and permissions:
Praise is one of the most commonly used forms of
reward. It is convenient, is nearly effortless, and makes
the recipient feel good. However, praising children for
being smart carries unintended consequences: It can
undermine their achievement motivation in a way that
praising their effort or performance does not (Cimpian,
Arce, Markman, & Dweck, 2007; Kamins & Dweck,
1999; Mueller & Dweck, 1998; see Dweck, 2007). In this
study, we investigated whether the negative conse-
quences of praising children for being smart extend to
the moral domain, by encouraging cheating.
There is some prior work suggesting that evaluative
feedback can influence children’s moral behaviors (Fu,
Heyman, Qian, Guo, & Lee, 2016; Mueller & Dweck,
1998; Zhao, Heyman, Chen, & Lee, 2017). Telling 5-year-
olds (but not younger children) that they have a reputa-
tion for being good leads to a reduction in their
cheating, presumably because they are interested in
maintaining this reputation (Fu etal., 2016). We pro-
pose that telling children that they are smart, a form of
ability praise, may have the opposite effect by motivat-
ing them to cheat to appear smarter. In a study consis-
tent with this possibility, Mueller and Dweck (1998)
found that 10-year-olds exaggerated how well they had
performed after receiving ability praise. However, little
is known about whether ability praise can influence
young children’s moral behavior. The present research
addressed this question by comparing the effects of
ability and performance praise on preschool children’s
Participants were 300 preschool children in eastern
China: one hundred fifty 3-year-olds (age range = 3.08
to 4.00 years, M = 3.62, SD = 0.27; 71 boys, 79 girls)
and one hundred fifty 5-year-olds (age range = 5.01 to
6.00 years, M = 5.38, SD = 0.33; 78 boys, 72 girls). To
measure cheating, we used a version of a well-
established peeking paradigm (see Heyman, Fu, Lin,
Qian, & Lee, 2015), in which an experimenter hides a
playing card (with a number from 3 to 9, excluding 6)
behind a barrier and children guess whether it is greater
or less than 6. The children are told that they can win
a prize if they guess correctly on at least three of the
The session began with a practice trial in which the
children were told that they had guessed correctly. They
were then randomly assigned to three conditions (50
children in each condition): In the ability condition,
children were told, “You are so smart.” In the perfor-
mance condition, they were told, “You did very well
this time.” In the baseline condition, no praise was
The real guessing game, which was identical across
the three conditions, followed this practice trial. On
each trial, the children were instructed not to peek.
Unbeknownst to them, the game was rigged to ensure
success on two of the first five trials and failure on
three. On each trial, the children were told whether
they had been successful, but no praise was given.
During the pivotal sixth trial, the experimenter left the
room for 60 s after eliciting a promise not to peek at
the card. Acts of cheating, defined as any obvious forms
of peeking (i.e., the child got out of his or her seat or
721529PSSXXX10.1177/0956797617721529Zhao et al.Smart Praise
Kang Lee, Dr. Eric Jackman Institute of Child Study, University of
Toronto, 45 Walmer Rd., Toronto, Ontario M5R 2X2, Canada
Praising Young Children for Being
Smart Promotes Cheating
Li Zhao1,2, Gail D. Heyman3,4, Lulu Chen1, and
1Institutes of Psychological Sciences, Hangzhou Normal University; 2Zhejiang Key Laboratory
for Research in Assessment of Cognitive Impairments, People’s Republic of China; 3Department
of Psychology, University of California, San Diego; 4Department of Psychology, Zhejiang
Normal University; and 5Dr. Eric Jackman Institute of Child Study, University of Toronto
Received 3/31/17; Revision accepted 6/28/17
2 Zhao et al.
leaned over the barrier), were recorded by a hidden
A pretest was used to assess basic numerical under-
standing, and the 3 children who failed (all 3-year-olds)
completed an analogous color-guessing task instead
(see the Supplemental Material available online for
more details about the procedure, including the pre-
test). Because our analyses showed that the pattern of
results did not differ when these 3 children who failed
the pretest were excluded, we report results that
included the data from all participants.
Figure 1 shows the cheating rate for each age group
and gender, broken down by condition. There was
more cheating in the ability condition than in the other
conditions overall, in each age group, and within each
gender (ps < .05 for all chi-square tests).
We conducted a binary logistic regression analysis
in which condition, age, gender, and their interactions
were the predictors of cheating. The final model was
significant, χ2(3, N = 300) = 15.68, p = .001, −2 log
likelihood = 399.13, Nagelkerk R2 = .068, and revealed
two significant effects. One was a main effect of condi-
tion (Wald = 10.12, df = 2, p = .006). A priori compari-
sons with the ability condition as the reference showed
that cheating rates were significantly higher in the abil-
ity condition than in the performance and baseline
conditions—performance condition: β = 0.78, SE = 0.29,
Wald = 7.26, df = 1, p = .007, odds ratio = 2.19, 95%
confidence interval (CI) = [1.24, 3.87]; baseline condi-
tion: β = 0.82, SE = 0.29, Wald = 7.93, df = 1, p = .005,
odds ratio = 2.27, 95% CI = [1.28, 4.02]. The other sig-
nificant effect was a main effect of gender; compared
with girls, boys cheated more across the three condi-
tions, β = 0.55, SE = 0.24, Wald = 5.39, df = 1, p = .020,
odds ratio = 1.74, 95% CI = [1.09, 2.77]. No other effects
were significant (ps > .1).
Cheating latency did not differ significantly across
conditions (ps > .1).
We examined how different forms of praise affect
young children’s moral behavior. The results showed
that 3- and 5-year-olds who were praised for their abil-
ity on a single occasion were more likely to cheat than
3-Year-Olds 5-Year-Olds Boys Girls Total
Participants Who Cheated (%)
Fig. 1. Percentage of participants who cheated in each age group, within each gender,
and overall, broken down by condition.
Smart Praise 3
their counterparts who were praised for their perfor-
mance, or not praised at all.
It is likely that ability praise promotes cheating because,
unlike performance praise, it is a generic form of language
that implies the presence of a stable ability (e.g., smart-
ness) that underlies performance (Cimpian etal., 2007).
In our study, ability praise may have motivated children
to cheat in order to uphold the positive trait assessment
or the reputation of being smart (Zhao etal., 2017).
Surprisingly, the effect was the same for 3-year-olds and
5-year-olds. This finding demonstrates that even 3-year-
olds are sensitive to the difference between ability and
performance praise, and that their understanding of
ability overlaps with that of older individuals. This find-
ing also raises questions about what kind of conceptual
understanding is required for ability praise to have
Our findings demonstrate that ability praise can
promote cheating in young children. More broadly,
they build on evidence suggesting that subtle social
cues have the power to shape children’s thinking and
behavior (Bryan, Master, & Walton, 2014) and suggest
that researchers cannot rule out the importance of
socialization effects in young children just because
there are no obvious sources of direct teaching or
James K. McNulty served as action editor for this article.
L. Zhao, K. Lee, and G. D. Heyman developed the study. L.
Chen collected the data. L. Zhao performed the data analysis
and drafted the manuscript. K. Lee and G. D. Heyman pro-
vided critical revisions. All the authors approved the final
version of the manuscript for submission.
We thank Brian Compton for his helpful comments on the
Declaration of Conflicting Interests
The authors declared that they had no conflicts of interest with
respect to their authorship or the publication of this article.
Additional supporting information can be found at http://
All data have been made publicly available via the Open
Science Framework and can be accessed at https://osf.io/
p2z8a/. The complete Open Practices Disclosure for this
article can be found at http://journals.sagepub.com/doi/
suppl/10.1177/0956797617721529. This article has received
the badge for Open Data. More information about the Open
Practices badges can be found at https://www.psychological
Bryan, C. J., Master, A., & Walton, G. M. (2014). “Helping” ver-
sus “being a helper”: Invoking the self to increase helping
in young children. Child Development, 85, 1836–1842.
Cimpian, A., Arce, H. M. C., Markman, E. M., & Dweck, C. S.
(2007). Subtle linguistic cues affect children’s motivation.
Psychological Science, 18, 314–316. doi:10.1111/j.1467-
Dweck, C. S. (2007). The perils and promises of praise.
Educational Leadership, 65(2), 34–39.
Fu, G., Heyman, G. D., Qian, M., Guo, T., & Lee, K. (2016).
Young children with a positive reputation to maintain are
less likely to cheat. Developmental Science, 19, 275–283.
Heyman, G. D., Fu, G., Lin, J., Qian, M. K., & Lee, K. (2015).
Eliciting promises from children reduces cheating.
Journal of Experimental Child Psychology, 139, 242–248.
Kamins, M. L., & Dweck, C. S. (1999). Person versus pro-
cess praise and criticism: Implications for contingent
self-worth and coping. Developmental Psychology, 35,
Mueller, C. M., & Dweck, C. S. (1998). Praise for intelligence
can undermine children’s motivation and performance.
Journal of Personality and Social Psychology, 75, 33–52.
Zhao, L., Heyman, G. D., Chen, L., & Lee, K. (2017). Telling
young children they have a reputation for being smart
promotes cheating. Developmental Science. Advance
online publication. doi:10.1111/desc.12585