Self-Deception 1
Running Head: SELF-DECEPTION
The Decay and Prevention of Self-Deception
Zoë Chance
Yale University
Francesca Gino & Michael I. Norton
Harvard University
Dan Ariely
Duke University
We thank Don Moore for providing the trivia questions used in these studies. Correspondence
should be addressed to Zoë Chance, Yale School of Management, 135 Prospect St L-15, New
Haven, CT 06511, [email protected], phone 857-891-5013, fax (203) 815-1557.
Self-Deception 2
Abstract
People demonstrate an impressive ability to engage in self-deception, distorting their negative
behavior to reflect positively on themselves – for example, by cheating on tests and attributing
their improved performance to their own genius. Given the costs of self-deception to both
individuals and organizations, we explore two means by which self-deception can be decreased:
natural decay and a behavioral intervention. We show that self-deception diminishes over time
when self-deceivers are repeatedly confronted with evidence of their true ability (Study 1);
importantly, however, this “learning” fails to make them less susceptible to future self-deception
(Study 2). Given this stickiness of self-deception, we test a behavioral intervention in which we
introduce barriers to self-deception – literally, by making it more difficult to access test answers
in order to cheat – which decreases self-deception (Study 3). Taken together, these studies offer
insight into when and how self-deception can be ameliorated.
Keywords: self-deception, cheating, self-enhancement, debiasing, self-perception, delusion
Self-Deception 3
The Decay and Prevention of Self-Deception
Nothing is easier than self-deceit. For what each
man wishes, that he also believes to be true.
Demosthenes (384–322 BC)
During the California power crisis orchestrated by Enron and other energy companies –
which cost the state some $30 billion – Enron CEO Jeffrey Skilling blamed the situation on state
government, defending his firm to PBS’s FrontLine: “We are the good guys… We are on the
side of the angels.” Prior to its collapse, Enron had been voted the most innovative of Fortune’s
“most admired companies” for six years in a row, and was the largest on Fortune’s list of 50
fastest-growing firms. Even after Enron had declared bankruptcy and Skilling been indicted,
Skilling testified at a Congressional hearing, “I was immensely proud of what we
accomplished. We believed that we were changing an industry, creating jobs, helping resuscitate
a stagnant energy sector, and… trying to save consumers and small businesses billions of dollars
each year. We believed fiercely in what we were doing” (Enron, 2005).
While a first reaction may be to question whether Skilling is simply blatantly lying, we
explore the notion that despite the unethical behavior of his firm, Skilling may have truly
believed that he – and his company – were not just not doing harm, but actually doing good.
Skilling may have engaged in self-deception, the process by which people involved in negative
behavior go beyond merely explaining their actions away and come to believe that their poor
behavior reflects well on them. In one recent investigation, for example, people given the
opportunity to cheat on tests came to believe that they were smarter people, and this self-
deception persisted even when they were given incentives to undo it, such that self-deception
Self-Deception 4
proved costly (Chance, Norton, Gino, & Ariely, 2011). Similar instances of self-deception are
quite common in organizations. For example, managers may come to view the aggressive or
even unethical strategies they used to increase profits as evidence of their creativity and tenacity
in getting the job done – such that their bad behavior ironically makes them rate their managerial
skills even more highly.
Self-deception can not only harm the self-deceived individual but cause collateral
damage as well – even facilitating bankruptcies of firms, states, and affected individuals – but
little research has explored how it might be prevented or ameliorated. Recent research in
behavioral ethics and moral psychology has examined the conditions under which people engage
in unethical behavior and deception and has identified various factors that exacerbate dishonesty
(e.g., Schweitzer, Ordonez, & Douma, 2004; Tenbrunsel & Messick, 2004; Tenbrunsel,
Diekmann, Wade-Benzoni, & Bazerman, 2010).1 We contribute to this body of work by
examining not deception but self-deception, exploring the psychology of how this costly
tendency can be reduced.
Clinging to a Positive Self-view
The notion that people tend to see themselves through rose-tinted glasses has been
explored in a number of research paradigms. People demonstrate a tendency to self-enhance:
most individuals are unable to see themselves through any lens other than a positive one
(Dunning, Leuenberger, & Sherman, 1995). Individuals routinely inflate their current standing
when evaluating themselves on positive attributes such as intelligence, agreeableness, ability or
even ethicality, distorting self-appraisals in order to maintain the most favorable self-view
1 Following prior research (e.g., Jones, 1991), throughout the paper we use the words “unethical,” “immoral,” and “dishonest” interchangeably.
Self-Deception 5
(Alicke, 1985; Taylor & Brown, 1988). Much of the empirical work on biased self-evaluations
has explored the ways in which people perceive themselves to be better than those around them,
from research demonstrating that the overwhelming majority of people see themselves as “better
than average” on a variety of traits (e.g., Alicke, Klotz, Breitenbecher, Yurak, & Vredenburg,
1995) – for example, 94 percent of college professors rated themselves as “above average” Cross
(1977) – to research showing that people see themselves as less susceptible than others to serious
accidents and major illnesses (Weinstein, 1980). In addition to overestimating themselves
compared to their peers, research has also shown that people are frequently motivated to
overestimate their own actual abilities, viewing themselves as better than they truly are (e.g.,
Burson, Larrick, & Klayman, 2006). This upward shift in evaluations of the current self reflects a
fundamental human motive to enhance and maintain a positive self-concept (Greenwald, 1980;
Sedikides & Strube, 1997).
Similarly, research on motivated reasoning has identified several means by which people
maintain positive self-views and ignore or explain away negative information about themselves
(Kunda, 1990; Pyzczynski & Greenberg, 1987). People tend to interpret information in ways that
confirm preexisting beliefs and attitudes about the self (e.g., Lord, Ross, & Lepper, 1979; Ditto
& Lopez, 1992; Swann, Stein-Seroussi, & Giesler, 1992) and about others (e.g., Darley & Gross,
1983; Schaller, 1992). Similarly, people can display creativity in justifying and rationalizing
their questionable behavior and decisions (e.g., Gino & Ariely, 2012; Norton, Vandello, &
Darley, 2004)
Judging oneself as generally better than others indicates an overoptimistic bias, and
motivated reasoning allows people to maintain their preferred self-views, but neither process is
fully self-deceptive. In our view, self-deception is the special case in which negative behavior is
Self-Deception 6
not only explained away (e.g., “I am better than others in general” or “everyone does this”) but
actually leads people to see themselves more positively than they did before the self-deception
occured. Motivated reasoning and better-than-average effects describe situations in which
individuals hold favored beliefs. Self-deception is a special case in which the individuals holding
these positive beliefs should know better – because they just behaved poorly – helping to explain
how people like Skilling can shift from the knowledge that Enron committed fraud for personal
gain to believing they were “on the side of the angels.”
Decreasing Self-Deception
Previous research exploring self-deception shows that individuals who cheat not only
give themselves full credit for their success, but expect continued success in the future. Consider
the paradigm developed by Chance et al. (2011). Participants take an initial test, with some being
given access to the answers, then all participants predict their scores on a future test – for which
no answer key will be available. Participants with access to the answers perform better on the
first test than those without the answers; most importantly, those given the answers deceive
themselves into thinking their superior performance owed to their ability. As a result, they
predict continued superior performance in the second test – a measure of self-deception.
Chance et al. (2011) showed that incentivizing people to avoid self-deception was
ineffective: even when those who cheated on a test were paid to predict their future performance
accurately, self-deception persisted, leading to inflated predictions which proved costly.
However, a different type of influence did affect self-deception. When cheaters were given
special recognition for their high performance scores – much like Enron traders recognized by
Fortune for their success – they became even more self-deceptive. Thus the previous research
Self-Deception 7
suggests that self-deception is sticky (it persists over time, from one event to another) and can be
exacerbated. In the present research, we examine how self-deception might decay over time (the
extent to which the stickiness “sticks”) and an intervention designed not to exacerbate but
ameliorate self-deception.
Although methods for reducing self-deception have not been explored previously, a
number of studies have investigated moderators of unethical behavior more generally. Cheating
increases under the following conditions: when individuals lack the cognitive resources to resist
temptation (Christian & Ellis, 2011; Gino, Schweitzer, Mead, & Ariely, 2010; Mead,
Baumeister, Gino, Schweitzer, & Ariely, 2009); when economic inequity induces feelings of
guilt or envy (Gino & Pierce, 2009; Moran & Schweitzer, 2008); when people feel anonymous,
such as when lights are low or when they wear sunglasses (Zhong, Bohns, & Gino, 2010); when
people feel inauthentic due to wearing counterfeit products (Gino, Norton, & Ariely, 2009); and
in the presence of large sums of cash (Gino & Pierce, 2009). Markedly fewer studies, however,
have investigated potential interventions to reduce cheating. A notable exception is research
showing that cheating becomes less prevalent when people are reminded of moral standards –
religious commandments or a school honor code (Mazar, Amir, & Ariely, 2008; Shu, Gino, &
Bazerman, 2011). Understanding the natural decay of self-deception and designing interventions
to decrease self-deception are crucial in assisting individuals in better monitoring their own self-
deceptive tendencies, and informing managers on how to design simple changes in their
organizations to reduce unethical behavior.
The Present Research
Three studies investigate explore the decay, reinstatement, and reduction of self-
deception. Study 1 observes the time course of the decay of self-deception when an initial act of
Self-Deception 8
self-deception – believing onself intelligent due to cheating on an initial test – is followed by two
rounds of feedback on one’s actual abilities – taking subsequent tests without the answers.
Because the world offers more than one opportunity to cheat, Study 2 explores how the decay of
self-deception can be interrupted when people are given a second opportunity to cheat, and
therefore a second opportunity to self-deceive. Finally, Study 3 tests an intervention designed to
reduce self-deception by employing a physical barrier to self-deception; we draw cheaters’
attention to their cheating by forcing them to exert effort to cheat.
Study 1: Decay of Self-Deception Over Time
Study 1 had two primary goals. First, we tested the extent to which self-deception
would persist in the face of evidence of true ability. Given that experience is crucial for learning
(Argote & Miron-Spektor, 2011), we predicted that self-deception would fade after repeated
experience. Participants completed a battery of four tests of their general knowledge, predicting
their test performance before each test. Some participants – in the answers condition – had access
to an answer key for Test 1, and we expected them to both perform better on Test 1 and predict
better perfomace on Test 2 than those who did not have access to the answers. Importantly,
neither group had an answer key for Tests 2 through 4, therefore their scores on these tests
offered repeated evidence of their true ability. We assessed the extent to which the inflated
predictions of those given the answers on Test 1 would be tempered by their experience taking
subsequent tests without the answers.
Second, we tested whether individual differences in the tendency to self-deceive would
predict the rate of decay. Chance et al. (2011) showed that that individuals high in self-deceptive
Self-Deception 9
enhancement were most likely to inflate their beliefs about their ability after cheating. We
predicted that self-deception would decay least rapidly for these individuals.
Method
Participants. Seventy participants (28 male, Mage = 23.9) from a university and
surrounding community in a northeastern city were paid $20 to complete the experiment as part
of a series of unrelated studies.
Design and procedure. Each participant was assigned to either the control or the answer
condition, and took four general knowledge quizzes with items such as, “What is the only
mammal that truly flies?” (Moore & Healy, 2008). Questions were pretested and configured into
10-item quizzes with trivia of varying topics and difficulty, selected to yield average scores of
approximately five points. For Test 1 of this study, participants in the answers condition had the
answers to all ten questions printed in an answer key at the bottom of the page. For all
particpants, Tests 2-4 did not have answers at the bottom. Participants were given three minutes
to complete each test, and were paid a bonus of $0.25 for each correct answer. After completing
each test, participants scored that test and predicted their score on the next – which they could
clearly see did not have answers at the bottom. Finally, participants completed Paulhus’s (1991)
Self-Deceptive Enhancement and Impression Management scales. In accordance with Paulhus
(1991), we calculated a Self-Deceptive Enhancement score for each participant by reverse-
coding the even-numbered questions and then scoring one point for each statement rated “six” or
“seven” (out of seven) on the first 26 items (e.g., “I never regret my decisions,” α=.69). The
remaining 16 items (e.g., “I have never pretended to be sick to avoid work or school”) were
aggregated in a similar fashion into an Impression Management score (α=.66).
Self-Deception 10
Results and Discussion
As expected, participants who had access to the answers for Test 1 scored higher on that
test than did those in the control group (Manswers = 7.89, SD 2.37 vs. Mcontrol = 4.44, SD 1.86;
F(69) = 45.31, p <.001). More importantly, the answers group also showed evidence of self-
deception, expecting to score higher than the control group on Test 2 despite lacking an answer
key for that test (Manswers = 6.28, SD = 1.78 vs. Mcontrol = 5.44, SD = 1.28; F(69) = 5.03, p = .03).
Because the two randomly-assigned groups possessed similar ability, there was no difference in
actual scores on Tests 2 through 4 (all ps > .3). Given the equivalence of the two groups’
performances, we could therefore test how long it would take the inflated estimates of self-
deceivers to return to the level of the control group. Failing to fully incorporate the evidence of
their true ability (i.e., their Test 2 score), the answers group again predicted marginally higher
scores than the control group on Test 3 (Manswers = 5.72, SD = 1.67 vs. Mcontrol = 5.00, SD = 1.91;
F(69) = 2.86, p =.10), and finally, adjusted to make their predictions equivalent to the control
group on Test 4 (Manswers = 5.53, SD = 2.16 vs. Mcontrol = 4.97, SD = 1.69; F(69) = 1.43, p = .24).
Note, however, that even for predictions of Test 4, participants in the answers group still tended
to predict higher scores, suggesting the staying power of self-deception.
We next explored whether people’s general tendency to engage in self-deception would
relate to the rate at which self-deception decayed. Self-Deceptive Enhancement was correlated
with overpredictions on the second test in the answers condition (r = .40, p = .02), but not the
control condition (r = .07, p = .59). A median split on Self-Deceptive Enhancement revealed
that, as predicted, high self-enhancers were driving the irrationally optimistic predictions in the
answers group. High self-deceivers significantly overpredicted their scores on Test 2 compared
to those in the control condition (MhighSDE = 6.58, SD = 1.83 vs. Mcontrol = 5.44, SD = 1.28; t(51) =
Self-Deception 11
2.65, p = .01). This difference in predictions persisted on Test 3 (MhighSDE = 5.95, SD = 1.65 vs.
Mcontrol = 5.00, SD = 1.91; t(51) = 1.82, p = .07). Eventually, even the high self-deceivers became
more realistic, predicting scores still directionally higher than the control group, but not
significantly so (MhighSDE = 5.74, SD = 2.30 vs. Mcontrol = 4.97, SD = 1.70; t(51) = 1.38, p = .17).
Low self-deceivers in the answers group, on the other hand, were indistinguishable from the
control group in their predictions on all three tests (all ts < .87, all ps .39 or greater).
Impression Management showed no significant relationship to overpredictions on any of
the tests (all rs < .27, all ps > .13), suggesting that the inflated predictions we observe are
primarily a private process rather than a strategy to impress others (such as the experimenter).
These results demonstrate that self-deception eventually fades after repeated exposure to
counterevidence in the form of less-than-stellar scores on tests without answers. The adjustment
was not immediate, however, and required multiple exposures – particularly for those with a
natural tendency to self-deceive. Waiting for the effects of self-deception to dissipate naturally,
then, holds only limited promise in attempting to decrease its prevalence.
Study 2: A Second Chance to Cheat
Study 1 demonstrated that self-deception can fade over time, and that this process
happens more slowly for some people than others. In everyday life, however, people are not
presented with only one opportunity to cheat – and only one opportunity to self-deceive. In Study
2 we explored whether the learning about true ability in the face of feedback that we observed in
Study 1 would persist even when people were given a second chance to cheat on Test 3, or
whether people would cheat – and then self-deceive – yet again. We predicted that those with the
Self-Deception 12
answer keys would cheat on Tests 1 and 3, and that their inflated scores would buffer them from
the knowledge of their true ability, re-awakening self-deception.
Participants. One hundred forty-eight participants (68 male, Mage = 23.0) from a
university and surrounding community in a northeastern city were paid $20 to complete the
experiment as part of a series of unrelated studies.
Design and procedure. Participants were assigned to either the control or the answer
condition, and took four ten-item general knowledge quizzes as in Study 1. In the answers
condition, Tests 1 and 3 had the answers to all ten questions printed at the bottom of the page.
Tests 2 and 4 did not have answers. Participants looked over each test before predicting their
score, and were then given three minutes to complete the test. For each question answered
correctly, they were paid a bonus of $0.25.
Results and Discussion
Replicating the result of Study 1, participants with the answer key for Test 1 scored
higher on that test than did those in the control group (Manswers = 7.65, SD = 2.01 vs. Mcontrol =
5.03, SD = 1.92; F(147) = 13.92, p <.001). The answers group also showed evidence of self-
deception, expecting to score higher than the control group on Test 2 as well, again despite
lacking an answer key for that test (Manswers = 6.06, SD = 2.23 vs. Mcontrol = 4.97, SD = 1.88;
F(147) = 10.18, p <.001). Of course, because the two groups possessed similar ability, there was
no difference between conditions in scores on Test 2 (Manswers = 5.55, SD = 2.09 vs. Mcontrol =
5.34, SD = 1.70; F(147) = .44, p = .51). The answers group, seeing an answer key on Test 3,
predicted higher scores than the control group (Manswers = 6.95, SD = 2.29 vs. Mcontrol = 4.51, SD
= 1.89; F(147) = 48.76, p <.001), and again answered more questions correctly (Manswers = 7.46,
SD = 2.14 vs. Mcontrol = 4.68, SD = 1.61; F(147) = 77.31, p <.001). In their predictions for the
Self-Deception 13
final test, did they carry forward their learning from having taken Test 2 without the answers, or
did the new opportunity to cheat wipe away that learning? Self-deception appeared to dominate:
those in the answers group again predicted better performance than the control group for Test 4
(Manswers = 5.75, SD = 2.10 vs. Mcontrol = 4.79, SD = 1.92; F(147) = 8.25, p =.005). Again,
however, both groups performed equally well on Test 4 (Manswers = 5.28, SD = 1.98 vs. Mcontrol =
4.88, SD = 1.96; F(147) = 1.46, p = .23).
A final comparison between the answer group’s predictions for Test 2 and Test 4 is
instructive. Their predictions for Test 2 were informed only by their performance on Test 1,
which was inflated by cheating. Their predictions for Test 4, on the other hand, should have been
determined by their performance not only on Tests 1 and 3 (where they had the answers) but also
by Test 2 (where they did not). Although their predictions were directionally lower for Test 4
than Test 2 (M Test2 = 6.06, SD = 2.23 vs. MTest4 = 5.75, SD = 2.10), the difference was only
marginal (t(79) = 1.16, p = .1). Thus the learning about their true ability on Test 2 appeared to be
disregarded or discounted as an input into their final predictions in favor of the self-deception in
which they engaged after Tests 1 and 3.
Study 3: Barriers to Self-Deception
Although feedback can eventually reduce self-deception (Study 1), Study 2 shows that
having an additional opportunity to cheat, with resulting superior performance, seems to erase
the memory of the previous legitimate feedback. Self-deception is therefore sticky: once it has
occurred, it is difficult to debias. As a result, reducing self-deception may require an intervention
closer to the initial act of cheating. In the case of our paradigm, this translates into making the
use of the answers on Test 1 more salient.
Self-Deception 14
We conducted our third study by computer to allow for a condition in which the answers
to the first test could be temporarily revealed by using the mouse to move the cursor to the
bottom of the screen. This slight modification still permitted cheating, but drew increased
attention to the act – making the act of cheating more salient, we predicted, would make cheating
harder to attribute to one’s innate genius. Our logic follows from related research by Mazar,
Amir, and Ariely (2008) which shows that people are comfortable cheating to the extent to which
it doesn’t feel too much like actively cheating. We expected that participants in the “reveal
answers” condition would report lower scores than those in the standard answers condition on
the first test, and also make more conservative predictions on the second—despite having access
to all the answers.
Method
Participants. Three hundred forty-eight participants (age and gender information not
collected) were university and community members in a northeastern city who were paid $20 to
complete the experiment as part of a series of unrelated studies.
Design and procedure. Participants were assigned to one of three conditions to take a
computerized IQ test with eight questions (e.g. “If a man weighs 75% of his own weight plus 42
pounds, how much does he weigh?”). In the control condition, participants saw the questions
only, with no answers. In the answers condition, the answers to all eight questions were visible at
the bottom of the screen. In the reveal answers condition, participants saw the same screen as the
control group, and were told that moving the mouse down to the bottom of the screen would
reveal the answers and moving it back up would hide them again. This methodology allowed us
to count the number of times participants peeked at the answers.
Self-Deception 15
Participants were given four minutes to complete the test, then asked to score themselves
and report their scores. Next, they were asked to predict how well they would do on a similar
eight-item computerized test without answers. Finally, they completed the second test. During
this task, they had 30 seconds to respond to each question before the program advanced them to
the next.
Results and Discussion
First, we confirmed that participants in the reveal answers condition did take the
opportunity to peek at the answers: 84 percent peeked at least once, and the mean number of
peeks was 5.83, SD = 4.35. This number was significantly, though weakly, related to number of
correct answers (r(128) = .18, p = .04).
As predicted, scores on the first test differed by condition (F(2, 347) = 3.09, p < .05). The
self-reported scores of those in the answers and control conditions followed the expected pattern,
with the answers group scoring higher than control (Manswers= 4.48, SD = 1.85 vs. Mcontrol = 3.95,
SD = 1.72; t(110)= 2.17, p = .03). The reveal answers group reported significantly lower scores
than the standard answers group (Mreveal = 3.98, SD = 1.83; t(236) = 2.11, p = .04). Thus, when
attention was drawn to peeking, participants were more aware of which problems they could
solve without assistance and did not claim undue credit.
Predictions for the second test also differed by condition (F (2, 344) = 3.23, p < .04),
with participants in the reveal answers condition carrying their modest predictions forward to
Test 2 (Mreveal = 3.93, SD = 1.67), such that their predictions were lower than those of
participants in the answers condition (Manswers = 4.42, SD = 1.70; t(235) = 2.23, p = .03): despite
equal access to the answers on Test 1, participants in the “reveal answers” condition did not
show evidence of self-deception in their predictions for Test 2. These low expectations in the
Self-Deception 16
reveal answers condition are particularly striking in light of the fact that participants in the
control condition actually predicted improvement on Test 2 (Mcontrol = 4.41, SD = 1.71). (The
amount of time allotted for the first test was relatively short in this experiment, such that the
control group may have expected practice to improve their performance.) Actual performance on
the second test did not differ by condition, (Mcontrol = .86, SD = 1.02; Manswers = .96, SD = 1.17;
Mreveal = .91, SD = 1.07; F(2, 345) = .23, p = .79).
These findings suggest that increasing awareness of the link between having answers to a
test present and one’s performance on that test can reduce both cheating and the self-deception
that arises from it.
General Discussion
People misbehave: They lie, they cheat, they steal, they betray. Yet most people view
themselves as good and moral individuals (Aquino & Reed, 2002) and seem to be resistant to
shining a critical moral light on their behavior. How do people resolve this apparent
inconsistency between their bad behavior and positive self-image? Their innate psychological
tendency to engage in self-deception (Tenbrunsel & Messick, 2004). Recognizing one’s own
self-deception is difficult (Bok, 1978; 1989), and correcting the ingrained mechanisms that
engender it is a difficult task (Tenbrunsel & Messick, 2004). The first step is to recognize its
pervasive presence and acknowledge that no one is immune to it. Tenbrunsel and Messick (2004,
p. 234) suggest that “by confronting the tendency toward self-deception head on, we will be
more likely to reduce its prevalence than if we ignore it and act as if it did not exist.” Yet
individuals themselves are ill-equipped to confront their own self-deception after the fact, since
Self-Deception 17
they are unaware of it; as observed in previous research, incentives often fail to reduce self-
deception (Chance et al., 2011).
The current research explores two means by which people’s self-deception may be
decreased, one addressing the issue after the initial self-deception has occurred, and the other
attempting to prevent self-deception from happening in the first place. Our studies demonstrate
that time and repetitions of unbiased feedback can reduce self-deception and eventually eliminate
it (Study 1); however, the tendency re-emerges when the opportunity for cheating presents itself
again (Study 2). Interventions early on in the process, providing barriers to self-deception, may
prevent self-deception and decrease cheating as well (Study 3).
In our studies, we tested for self-deception using a specific set of tasks, similar to school
or university test situations in which students might have the opportunity to cheat. While we
examined the impact of self-deception on people’s beliefs about their future performance, self-
deception in contexts like these may also impact people’s subsequent behavior. For instance, it
may lead them to spend less time preparing for future tests, thus reducing their learning in
addition to hampering the possibility of good future performance. Future research could examine
this type of behavioral consequence, both in the context of academic cheating as well as in other
contexts where people have the opportunity to inflate their performance by cheating and then
deceive themselves. Importantly, the tools for debiasing or reducing self-deception may be
different in contexts outside of academic cheating. Further research would enhance our
understanding of how to prevent both cheating and self-deception in organizations and society
more broadly.
Self-Deception 18
Theoretical Contributions
Although the construct of self-deception has a long history in psychology, decision
making research and organizational behavior, the nature of the process by which self-deception
takes place is still subject to debate (Gur & Sackeim, 1979; Mele, 1997; Sackeim & Gur, 1978;
1979; Trivers, 2000). Our research contributes to this literature by showing that when people are
made more aware of the fact that they are cheating, they are subsequently less likely to self-
deceive (Study 3). In addition, we show that while self-deception does occur rapidly – it takes
just one episode of cheating to change people’s beliefs about their ability – there is some decay
over time, suggesting that self-deception may provide temporary “boosts” to the self-concept, but
that these boosts may be relatively short-lived given some feedback from the environment (Study
1). At the same time, we show that sensitivity to feedback is highly contingent on the extent to
which it enables self-deception; feedback that suggests one is a genius is given more weight than
feedback suggesting one’s true abilities (Study 2). As a result, it appears as though people are
“primed” to self-deceive, awaiting opportunities to inflate their self-views, and only grudgingly
and occasionally adjusting them downward.
In addition, our research contributes to the behavioral ethics literature by highlighting the
importance of understanding psychological drivers to unethical behavior, their powerful and
robust effects, and the difficulty of eliminating them. Perhaps most importantly, the existing
research on the individual psychology of ethical decision making takes mostly a “one-shot”
approach by examining the effect of one or more specific variables on one subsequent unethical
act, or a global approach by assessing very general beliefs about the self such as people’s broad
tendency to view themselves in a positive light (e.g., Alicke, 1985; Chambers & Windschitl,
2004; Moore & Kim, 2003; Taylor & Brown, 1988). Along these lines, Chance et al. (2011)
Self-Deception 19
documented how a single act of cheating can lead to self-deception in the short term. The current
investigation explores ethics over a longer time course, examining how an initial unethical
behavior changes people’s beliefs in the longer term – and how those beliefs are impacted by
both receiving more information about one’s character over time (Study 1) and being provided
with subsequent opportunities to engage in further unethical behavior (Study 2).
Organizational Implications
Learning how to prevent or diminish self-deception is crucial because cheating is both
commonplace and costly (Callahan, 2004; Hollinger & Langton, 2007). Evidence shows that the
majority of high school students admit having cheated during the past year (Josephson Institute,
2010), and the majority of resumes contain misleading information (HireRight, 2012). According
to the U.S. Chamber of Commerce, nearly 75 percent of workers steal from their employer,
causing fully 30 percent of all corporate bankruptcies – and most of the perpetrators of employee
theft are managers (Burke & Cooper, 2010). If even some of this dishonesty is facilitated by self-
deception, then understanding self-deception may improve social welfare.
How might organizations erect barriers to self-deception? In the documentary Enron: The
Smartest Guys in the Room, a former Enron trader confessed that it was ambiguity that allowed
him to participate in the firm’s questionable activities, without admitting to himself what he was
doing was wrong: “If I had questions, I didn't ask them, because I didn't want to know the
answer. You know, I didn't want confirmed what I suspected might be true, that what I was
doing was, in fact, unseemly or was at least unethical, if not worse.” In Shu et al. (2012), signing
an honesty pledge before having the opportunity to cheat reduced cheating by making ethics
salient during the decision process. Like our “mouse peeks” intervention, this psychological
Self-Deception 20
barrier decreased the likelihood that individuals could cheat mindlessly – and mindless cheating
offers an optimal condition for self-deception, since it implies that part of the self is unaware of
one’s actions.
Indeed, Mazar et al. (2008) demonstrated that people are most comfortable cheating when
they are unaware of the fact that they are cheating, when there is sufficient ambiguity in their
behavior. In organizational settings, this kind of ambiguity is often present: does stealing pens
from work make me a bad person, or does it in fact show that I am the very best employee at the
firm because I’ll use them to work at home this evening? At the same time, there are clearly
contexts in which rules for good conduct are clear; in such settings, cheating may not lead to
self-deception. Future research could explore how self-deception following cheating varies
depending on the level of ambiguity in the situation, thus deepening our understanding of the
conditions under which self-deception is most likely to occur.
Conclusion
It is perhaps not surprising that economic incentives can encourage cheating and self-
deception in laboratory studies (e.g. Gino & Pierce, 2009) and also in real world situations like
the Enron environment – people like money and will behave unethically to obtain more of it.
However, as noted earlier, those same economic incentives have proven to be much less effective
at undoing self-deception; once deception has taken place, money fails to reduce self-deception
(Chance et al., 2011). In such cases, and as we have shown here, when economic incentives miss
the mark, psychological interventions may prove more fruitful.
Self-Deception 21
References
Alicke, M. D. (1985). Global self-evaluation as determined by the desirability and controllability
of trait adjectives. Journal of Personality and Social Psychology, 49, 1621-1630.
Alicke, M. D., Klotz, M. L., Breitenbecher, D. L., Yurak T. J., & Vredenburg, D. S. (1995).
Personal contact, individuation, and the better-than-average effect. Journal of Personality
and Social Psychology, 68, 804-825.
Argote, L., & Miron-Spektor, E. (2011). Organizational learning: From experience to
knowledge. Organization Science, 22(5), 1123-1137.
Aquino, K. F., & Reed, A., II (2002). The self-importance of moral identity. Journal of
Personality and Social Psychology, 83, 1423–1440.
Burke, R. J. & Cooper, C. L. (2010) Risky Business: Psychological and Behavioural Aspects of
Risk. Surrey, England: Gower Publishing, Ltd.
Burson, K. A., Larrick, R. P., & Klayman, J. (2006). Skilled or unskilled, but still unaware of it:
How perceptions of difficulty drive miscalibration in relative comparisons. Journal of
Personality and Social Psychology, 90(1), 60-77.
Callahan, D. (2004). The cheating culture: Why more Americans are doing wrong to get ahead.
New York: Harcourt.
Chambers, J. R., & Windschitl, P. D. (2004). Biases in social comparative judgments: The role
of non-motivated factors in above-average and comparative optimism effects.
Psychological Bulletin, 130, 813-838.
Chance, Z., Norton, M., Gino, F., & Ariely, D. (2011). Temporal view of the costs and benefits
of self-deception. Proceedings of the National Academy of Sciences, 108(3), 15655-
15659.
Self-Deception 22
Christian, M. S., & Ellis, A. P. J. (2011). Examining the effects of sleep deprivation on
workplace deviance: A self-regulatory perspective. Academy of Management Journal, 54,
913-934.
Colvin, C. R., & Block, J. (1994). Do positive illusions foster mental health? An examination of
the Taylor and Brown formulation. Psychological Bulletin, 116, 3–20.
Cross, P. (1977). Not can but will college teachers be improved? New Directions for Higher
Education, 17, 1–15.
Darley, J. M., & Gross, P. H. (1983). A hypothesis-confirming bias in labeling effects. Journal of
Personality and Social Psychology, 44(1), 20-33.
Ditto, P. H., & Lopez, D. F. (1992). Motivated skepticism: Use of differential decision criteria
for preferred and non-preferred conclusions. Journal of Personality and Social
Psychology, 63, 568-584.
Dobbins, G. H., Farh, J. L., & Werbel, J. D. (1993). The influence of self-monitoring and
inflation of grade-point averages for research and selection purposes. Journal of Applied
Social Psychology, 23, 321-334.
Dunning, D., Leuenberger, A., & Sherman, D. A. (1995). A new look at motivated inference:
Are self-serving theories of success a product of motivational forces? Journal of
Personality and Social Psychology, 59, 58-68.
Enron: The Smartest Guys in the Room. Dir. Alex Gibney. Jigsaw Productions, 2005.
Gino, F., & Ariely, D. (2012). The dark side of creativity: Original thinkers can be more
dishonest. Journal of Personality and Social Psychology, 102(3), 445-459.
Gino, F. & Pierce, L. (2009). The abundance effect: Unethical behavior in the presence of
wealth. Organizational Behavior and Human Decision Processes, 109(2), 142-155.
Self-Deception 23
Gino, F., Norton, M., & Ariely, D. (2010). The counterfeit self: The deceptive costs of faking it.
Psychological Science, 21(5), 712-720.
Gino, F., Schweitzer, M. E., Mead, N. L., & Ariely, D. (2011). Unable to resist temptation: How
self-control depletion promotes unethical decision making. Organizational Behavior and
Human Decision Processes, 115, 191-203.
Greenwald, A. G. (1980). The totalitarian ego: Fabrication and revision of personal history.
American Psychologist, 35, 603-618.
Gur, R. C., & Sackeim, H. A. (1979). Self-deception: A concept in search of a phenomenon.
Journal of Personality and Social Psychology, 37, 147-169.
HireRight: HireRight 2012 Employment Screening Benchmarking Report (2012). Retrieved
from http://www.hireright.com/benchmarking/.
Hollinger, R. C., & Langton, L. (2007). National Retail Security Survey: Final Report.
Gainesville, Florida: University of Florida.
Josephson Institute Center for Youth Ethics: The Ethics of American Youth: 2010. (2010).
Retrieved from http://charactercounts.org/programs/reportcard/2010/installment02_
report-card_honesty-integrity.html
Kirk, B. A., & Sereda, L. (1969). Accuracy of self-reported college grade averages and
characteristics of non and discrepant reporters. Educational and Psychological
Measurement, 29, 147-155.
Kunda, Z. (1990). The case for motivated reasoning. Psychological Bulletin, 108, 480-498.
Lord, C. G., Ross, L., & Lepper, M. R. (1979). Biased assimilation and attitude polarization: The
effects of prior theories on subsequently considered evidence. Journal of Personality and
Social Psychology, 37(11), 2098-2109.
Self-Deception 24
Mazar, N., Amir, O., & Ariely, D. (2008). The dishonesty of honest people: A theory of self-
concept maintenance. Journal of Marketing Research,45, 633–644.
Mead, N., Baumeister, R. F., Gino, F., Schweitzer, M., & Ariely, D., (2009). Too tired to tell the
truth: Self-control resource depletion and dishonesty. Journal of Experimental Social
Psychology, 45(3), 594-597.
Mele, A. (1997). Real self-deception. Behavioral & Brain Sciences, 20, 91–136
Moore, D. A. & Kim, T. G. (2003). Myopic social prediction and the solo comparison effect.
Journal of Personality and Social Psychology, 85(6), 1121-1135.
Moore, D. A. & Healy, P. J. (2008) The trouble with overconfidence, Psychological Review,
115(2), 502-517.
Moran, S. and Schweitzer, M. (2008). When better is worse: Envy and the use of deception.
Negotiation and Conflict Management Research. 1(1), 3-29.
Norton, M.I., Vandello, J.A., & Darley, J.M. (2004). Casuistry and social category bias. Journal
of Personality and Social Psychology, 87, 817-831.
Paulhus, D. L. (1991). Measurement and control of response bias. In J. P. Robinson, P. R.
Shaver, & L. Wrightsman (Eds.), Measures of personality and social–psychological
attitudes (pp. 17–59). San Diego, CA: Academic Press.
Pyszczynski, T. & Greenberg, J. (1987). Self-regulatory perseveration and the depressive self-
focusing style: A self-awareness theory of the development and maintenance of reactive
depression. Psychological Bulletin, 102, 122-138.
Sackeim, H. A., & Gur, R. C. (1978). Self-deception, self-confrontation, and consciousness. In
G. E. Schwartz & D. Shapiro (Eds.), Consciousness and self-regulation, advances in
research and theory (Vol. 2, pp. 139–197). New York: Plenum Press.
Self-Deception 25
Sackeim, H. A., & Gur, R. C. (1979). Self-deception, other deception, and self-reported
psychopathology. Journal of Consulting and Clinical Psychology, 47, 213–215.
Schaller, M. (1992). In-group favoritism and statistical reasoning in social inference:
Implications for formation and maintenance of group stereotypes. Journal of Personality
and Social Psychology, 63, 61-74.
Schweitzer, M., Ordonez, L., & Douma, B. (2004). The role of goal setting in motivating
unethical behavior. Academy of Management Journal, 47, 422-432.
Sedikides, C., & Strube, M. J. (1997). Self-evaluation: To thine own self be good, to thine own
self be sure, to thine own self be true, and to thine own self be better. In M. P. Zanna
(Ed.), Advances in Experimental Social Psychology, 29, 209-269. New York: Academic
Press.
Shu, L., Gino, F., & Bazerman, M. (2011). Dishonest deed, clear conscience: When cheating
leads to moral disengagement and motivated forgetting. Personality and Social
Psychology Bulletin, 37(3), 330-349.
Shu, L. L., Mazar, N., Gino, F., Ariely, D., & Bazerman, M. H. (2012). When to sign on the
dotted line? Signing first makes ethics salient and decreases dishonest self-reports.
Harvard Business School NOM Unit Working Paper No. 11-117.
Swann, W. B., Jr., Stein-Seroussi, A. & Giesler, B. (1992). Why people self-verify. Journal of
Personality and Social Psychology, 62, 392-401
Taylor, S. E., & Brown, J. (1988). Illusion and well-being: A social psychological perspective on
mental health. Psychological Bulletin, 103, 193–210.
Self-Deception 26
Tenbrunsel, A. E., Diekmann, K. A., Wade-Benzoni, K. A., & Bazerman, M. H. (2010). The
ethical mirage: A temporal explanation as to why we are not as ethical as we think we
are. Research in Organizational Behavior, 30, 153-173.
Tenbrunsel, A.E. & Messick, D.M., Ethical fading: (2004). The role of self deception in
unethical behavior. Social Justice Research, 17(2), 223-236.
Trivers, R. (2000). The elements of a scientific theory of self-deception. Annals of the New York
Academy of Sciences, 907, 114-131.
Weinstein, N. D. (1980). Unrealistic optimism about future life events. Journal of Personality &
Social Psychology, 39, 806-820.
Wright, S. S. (2000). Looking at the self in a rose-colored mirror: Unrealistically positive self-
views and academic performance. Journal of Social and Clinical Psychology, 19, 451-
462.
Zhong, C. B., Bohns, V. K., & Gino, F. (2010). A good lamp is the best police: Darkness
increases self-interested behavior and dishonesty. Psychological Science, 21(3), 311-314.