Learning and Individual Differences 39 (2015) 8995

Contents lists available at ScienceDirect

Learning and Individual Differences

journal homepage: www.elsevier.com/locate/lindif

Dialectical thinking moderates the effect of extrinsic motivation on

intrinsic motivation
Yaoran Li a,, Kennon M. Sheldon b, Rude Liu c
a
Department of Educational, School, and Counseling Psychology, University of Missouri, United States
b
Department of Psychological Sciences, University of Missouri, United States
c
School of Psychology, Beijing Normal University, China

a r t i c l e i n f o a b s t r a c t

Article history: The present research tested the individual difference variable of dialectical thinking style (DT) as a moderator of
Received 16 February 2014 the relationship between extrinsic motivation and intrinsic motivation, in situations where rewards have been
Received in revised form 3 February 2015 introduced for an initially enjoyable activity. In Study 1, participants were asked to imagine gaining rewards
Accepted 29 March 2015
for doing an enjoyable activity and in Study 2, participants were given real rewards for doing an enjoyable activ-
Available online xxxx
ity, after which subsequent extrinsic motivation and intrinsic motivation were rated. Both studies revealed a sig-
Keywords:
nicant interaction effect of DT and self-reported extrinsic motivation in relation to subsequent self-reported
Dialectical thinking intrinsic motivation. The results of Study 3 replicated the general pattern again, using cultural membership
Extrinsic motivation (China vs. the U.S.) as a stand-in for DT. Findings indicate that DT mitigates the detrimental effects of extrinsic
Intrinsic motivation motivation on intrinsic motivation, identifying an important new individual difference moderator of the
Undermining effect undermining effect of extrinsic motivation on intrinsic motivation.
2015 Elsevier Inc. All rights reserved.

1. Introduction construct, however, we rst provide more background information on

Research has shown that peoples' intrinsic motivation to do an activ-
ity often decreases when they perceive external incentives for doing
that activity or for doing it well (Deci, 1971; Kruglanski, Friedman, &
Zeevi, 1971; Lepper, Greene, & Nisbett, 1973; Smith & Pittman, 1978;
Tang & Hall, 1995; Warneken & Tomasello, 2008). Cognitive Evaluation
Theory (CET), a facet of Self-determination Theory (SDT; Deci & Ryan,
1985), explained the undermining effect of external incentives by
considering people's perceived locus of causality for their own behavior
(Deci & Ryan, 1985; Ryan & Deci, 2000). If people perceive that their
personal interests and desires are the cause of their actions, then they
tend to enjoy the activity; however, if they perceive that their actions
are caused by external contingencies or incentives, then they may feel
controlled and thus the intrinsic motivation may be spoiled. Notably, ac-
cording to CET, external rewards or incentives do not necessarily under-
mine intrinsic motivation; what matters are people's perceptions of
those incentives, or the induced extrinsic motivation. Although situa-
tions containing external incentives are more likely, on average, to be
perceived as controlling and coercive, people can differ in their propen-
sity to perceive situations as controlling (Deci & Ryan, 1985), differences
which can affect intrinsic motivation. This article will examine one such
individual difference, namely, dialectical thinking. Before turning to this
Corresponding author at: 16 Hill Hall, University of Missouri, Columbia, MO 65211,
United States. Tel.: +1 602 568 3041.
E-mail address: yl6df@mail.missouri.edu (Y. Li).
SDT.
Although early SDT research relied on the simple distinction be-
tween intrinsic and extrinsic motivation, later versions of the theory
identied a broader spectrum of types of motivation. Today, the theory
species a continuum of motivational internalization, anchored by ex-
trinsic motivation (my behavior is caused by the incentives or pres-
sures in the situation) on the non-internalized end, and intrinsic
motivation (my behavior is caused by my interest in and enjoyment
of the activity) on the internalized end, in between lie introjected mo-
tivation (in which behavior is felt to be caused by non-integrated parts
of the self) and identied motivation (in which behavior is felt to be
caused by the self, but is not necessarily enjoyable; introjected and iden-
tied motivation are not considered in the current article). Ryan and
Connell (1989) supported the motivational internalization concept by
showing that the types of motivation conform to a simplex correlational
pattern, in which any two adjacent dimensions on the continuum corre-
late positively with one another, and the two extreme dimensions cor-
relate negatively with each other. The typically observed negative
correlation between intrinsic and extrinsic motivation is consistent
with the undermining effect discussed above; as perceived extrinsic
motivation increases, perceived intrinsic motivation tends to decrease.
Again, however, this is an on average phenomenon, and in some
cases perceived extrinsic and intrinsic motivation could be uncorrelated
or even positively correlated (Cameron, Pierce, Banko, & Gear, 2005).
However, few studies have examined the moderators of the association.
In this research we examined one potential individual difference

http://dx.doi.org/10.1016/j.lindif.2015.03.019
1041-6080/ 2015 Elsevier Inc. All rights reserved.
90 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995
moderator of the relationship between extrinsic incentives and intrinsic
motivation, namely, dialectical thinking (DT).
1.1. Dialectical thinking and its role in the undermining effect
Cognitive developmental psychologists have found that there are re-
liable individual differences in how adults engage in various types of
complex reasoning. One such difference is relevant in situations involv-
ing potentially contradictory information. In such situations some peo-
ple only use formal thinking, that involves either fully rejecting or
fully accepting one side of a contradiction by polarizing contradictory
propositions, and then choosing only one proposition to believe. In con-
trast, other people have developed dialectical thinking that involves
accepting seeming contradictions by viewing things in a multiple-
dimensional way (Bassett, 2006; Peng & Nisbett, 1999; Yang, 2008).
Dialectical thinkers believe that the world is composed of many con-
tradictions and is in constant change, and they are attentive to the cir-
cumstances of the object and the relationships that involve the object
when making attributions, and rarely categorize objects into a right or
wrong group (Nisbett, Peng, Choi, & Norenzayan, 2001, also see
Spencer-Rodgers, Williams, & Peng, 2010, for a review). Dialectical
thinkers not only perceive other objects and people in a multi-
dimensional way, they also perceive their own personality as containing
sometimes-contradictory factors (e.g., I am sometimes extraverted and I
am sometimes introverted) and their beliefs and behaviors can thus be
more exible and amenable to change when the context has changed
(Boucher, 2011; Spencer-Rodgers, Williams, et al., 2010). With regard
to emotional experience, dialectical thinkers are more comfortable
with mixed emotions than non-dialectical thinkers; though they tend
to perceive more emotional complexity, especially in ambiguous situa-
tions (Leu et al., 2010). Research also showed that DT helps people for-
mulate more exible coping strategies under changing or complex
circumstances (Cheng, 2009). Furthermore, in cultures where dialecti-
cal thinking is prevalent, reports of unpleasant emotions have been
found to be only weakly negatively associated, and sometimes non-
associated, with reports of pleasant emotions (e.g., Kitayama, Markus,
& Kurokawa, 2000; Spencer-Rodgers, Peng, & Wang, 2010).
Regarding the undermining effect, the introduction of external in-
centives into a situation would normally be expected to increase extrin-
sic motivation and thus undermine intrinsic motivation. However, we
hypothesized that DT helps people resist this tendency for several pos-
sible reasons. First, Cognitive Evaluation Theory (Deci & Ryan, 1985)
states that undermined intrinsic motivation results from a change in
the person's perceived locus of causality from internal to external, and
the psychological feeling of pressure that accompanies this change. Peo-
ple high in DT may be able to experience increased external motivation,
but not perceive this as sources of pressure; in their minds, external mo-
tivations and intrinsic motivations can co-exist. Or, even if they do feel
pressure, this may not necessarily impact their positive emotions and
motivations, as suggested by Kitayama et al.'s (2000) research.
Another reason involves the over-justication theory account of the
undermining effect. Over-justication theory (Lepper et al., 1973) pro-
poses that when external motivators are present along with intrinsic
motivations, people tend to adopt the more salient factor (proffered re-
wards) as the explanation for their behavior, discounting the impor-
tance of the original intrinsic motivation. The behavior is over-
justied, and the attribution of intrinsic motivation is not needed. Peo-
ple high in DT may be able to avoid such discounting, because they can
accept multiple reasons for doing things. Finally, cognitive dissonance
theory (Heider, 1958) provides a third reason for expecting DT to
have impact. From this theoretical perspective, people feel dissonance
when their actions (I will get money for doing this) are inconsistent
with their attitudes (I am doing this because I enjoy it). The disso-
nance creates a non-conscious desire to change one of the contradictory
propositions, which causes them to revise the initial attitude. Because
people high in DT are comfortable with contradictions, however, they
may be able to avoid or sidestep such dissonance, thereby maintaining
their intrinsic motivation. Notably, our research does not attempt to dis-
tinguish between these three explanations of undermining, as all three
make similar predictions for our purposes.
In sum, a person low in DT, engaging in one or the other thinking,
may conclude that they are primarily motivated by the salient external
incentive(s), rather than by their own intrinsic interest. This is because
their lay theories (Levy, Ramrez, Rosenthal, & Karafantis, 2013; Plaks,
Levy, & Dweck, 2009) concerning the relationship between intrinsic and
external motivation better ts the typical simplex pattern, in which
external motivation and intrinsic motivation are negatively correlated
(Ryan & Connell, 1989). Thus the introduction of external incentives
may affect their anticipated desire to do the activity. In contrast, a per-
son high in DT may be able to accommodate or compartmentalize the
contradictory information and thus perceive that his/her behavior is
motivated by both intrinsic interest and external rewards. In other
words, intrinsic and external motivation may be uncorrelated in their
lay theories. Hence, their anticipated desire to do an activity may be un-
affected by proffered extrinsic motivation. Furthermore, since dialecti-
cal thinkers also adapt their cognitions and behaviors to new contexts
more exibly than non-dialectical thinkers (Ji, Nisbett, & Su, 2001), it
may be possible for them to reevaluate the situation and reinstate an in-
ternal locus of causality for later behavior after the external motivators
are taken away, such that any suppressed intrinsic interest will rebound.
For example, people high on DT are able to think, I am doing this for re-
wards, but I also can do this because of my interest later or I was doing
this before for rewards, but now I can do it because of my interest.
1.2. Current studies
Our three studies tested the hypothesized role of DT in moderating
the association between external motivation and reported intrinsic mo-
tivation thereafter. We tested this hypothesis using a survey with imag-
inary reward scenarios (Study 1), an experiment which introduced real
rewards (Study 2), and a scenario study with cross-cultural samples
(Study 3). Imaginary scenarios were chosen in Study 1 to measure
people's lay theories about the association between intrinsic motivation
and external motivation. The use of real rewards in Study 2 allowed us
to test whether the lay theory effect extended to doing a real activity.
Study 3 was undertaken because previous research has found cultural
differences in the prevalent thinking style (Peng & Nisbett, 1999;
Spencer-Rodgers & Peng, 2004; also see Spencer-Rodgers, Williams,
et al., 2010, for a review). Easterners were found to use DT to a greater
extent, while Westerners were found to use non-dialectical or formal
logical thinking to a greater extent. Thus in Study 3 we extended the
analysis to cultural differences, focusing on two cultural groups
known to differ in DT (North Americans and East Asians). We hypothe-
sized that the Asian group, presumably higher in DT, would evidence
less undermining by external motivation in imagined scenarios.
2. Study 1
In Study 1 participants were presented with four initially enjoyable
activities and then asked to imagine that rewards had been introduced
for doing the activities. Afterwards they rated their extrinsic motivation
and estimated how their intrinsic interest in the activity would change
after the reward was offered.
2.1. Method
2.1.1. Participants and procedure
Participants were 55 American undergraduate students who identi-
ed themselves as European American (24 males and 31 females, mean
age = 20.8, SD = 1.4) attending the University of Missouri. They were
all from an upper division psychology class, and received extra credit
 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995
in the class in exchange for their participation. The survey was present-
ed on-line.
After logging on to the survey participants rst completed demo-
graphic information and the DT measure (described below). Afterwards
they were asked to type the name of a particular web-site that they
enjoyed browsing. They were then asked to imagine that a market re-
search rm would pay them $10 a day to browse that site and answer
various questions about it. Then were then asked two questions: extrin-
sic motivation (During the market study, I would go to the website be-
cause I want the money, rated on a 1 (strongly disagree) to 5 (strongly
agree) scale), and subsequent intrinsic motivation (After the study was
over, my interest in browsing the website would probably where
the response options were decrease a lot, decrease a little, not
change, increase a little, and increase a lot). We chose to use
these single items in Study 1 because they are prototypical exemplars
of external and intrinsic motivation, respectively, allowing us to exam-
ine the lay theory effects of DT. Again, our hypothesis was that DT
would moderate the association between these two responses such
that those high in DT would evidence a less negative correlation be-
tween extrinsic and intrinsic motivation, compared to those low in DT.
Participants then repeated this procedure for three further cued en-
joyable activities: playing a sport they enjoy, playing a game they like,
and watching a movie they enjoy; in each case, after describing the ac-
tivity they were asked to imagine that an incentive had been intro-
duced, after which they again rated the two items above. The four
extrinsic motivation and the four intrinsic motivation items were each
averaged to create overall extrinsic and intrinsic motivation scores.
2.1.2. Dialectical thinking
Before participants started the imaginary tasks, they completed the
32 item Dialectical Self Scale (DSS, Spencer-Rodgers et al., 2010), using
a 5-point Likert scale with higher scores indicating higher DT. The
scale measures three components of DT: contradiction (e.g., When I
hear two sides of an argument, I often agree with both), cognitive
change (e.g., I often nd that my beliefs and attitudes will change
under different contexts) and behavior change (e.g., I often change
the way I am, depending on who I am with). We averaged the three
subscales together based on a principal components analysis suggesting
that they formed a single factor. The Cronbach's alpha of the aggregate
scale was 0.88.
2.2. Results
2.2.1. Preliminary analyses
There was one participant who did not complete the imaginary
tasks, and another participant who did not complete the DT scale.
Those two participants' data were excluded, leaving 53 participants
for the analyses reported below. Age and gender are ignored below, be-
cause they had no main effects on DT, extrinsic motivation or intrinsic
motivation, and they did not interact with DT or extrinsic motivation
when predicting intrinsic motivation. Descriptive statistics for the
Study 1 variables are shown in Table 1.
2.2.2. Tests of primary hypothesis
To examine the main hypothesis that DT moderates the association
between extrinsic motivation and intrinsic motivation, a multiple
Table 1
Descriptive statistics and variable inter-correlations from Studies 1 and 2.
Study 1 (n = 53) Study 2 (n = 27)
Variables M SD 1 2 M SD
1 Dialectical thinking 2.25 .44 3.39 .59
2 Extrinsic motivation 3.85 .85 .06 6.63 2.75
3 Intrinsic motivation 3.14 .46 .14 .05 4.89 1.41
Note: ** means p b .01.
91
regression model was employed with intrinsic motivation as the depen-
dent variable, and DT, extrinsic motivation, and DT extrinsic motiva-
tion entered simultaneously as predictors. All variables were
standardized before use. Neither main effect was signicant, however
the hypothesized interaction effect was signicant, = .36, t(51) =
2.60, p = 0.01. Post-hoc probing of the interaction (Aiken & West,
1991) indicated that for people whose DT score was one standard devi-
ation above the mean, extrinsic motivation was not signicantly associ-
ated with intrinsic motivation, = .44, t(51) = 1.92, p = .06. In
contrast, for people whose DT rating was one standard deviation
below the mean, extrinsic motivation negatively predicted intrinsic mo-
tivation, = .51, t(51) = 2.29, p = .03. Fig. 1 illustrates the plotted
interaction with the means of intrinsic motivation at different dialectical
thinking levels (1SD and 1SD) and different extrinsic motivation level
(1SD and 1SD).
2.3. Brief discussion
Our central prediction was that DT would moderate the association
between rated extrinsic motivation and rated intrinsic motivation, in a
hypothetical context of how one would feel about doing formerly enjoy-
able activities after external incentives had been introduced. The results
supported this prediction. The extrinsic motivation of people with rela-
tively high DT scores was nearly positively associated with their intrin-
sic motivation, while the association was signicantly negative for
people with relatively low DT scores. This result suggests that DT can
buffer individuals from having their intrinsic motivation undermined
by the extrinsic motivation induced by external incentives.
One limitation of Study 1 is its usage of imaginary scenarios instead
of actual situations involving extrinsic incentives. Participants' real re-
sponse or motivation might not be in line with their lay theories. We
thus conducted a second study that introduced actual monetary re-
wards for doing a particular concrete activity.
3. Study 2
In Study 2, we wanted participants to experience doing an activity
known to be intrinsically enjoyable. We returned to the Soma puzzles
of the early intrinsic motivation research (Deci, 1971, 1972). The Soma
puzzle is a 3-dimensional puzzle that most people enjoy (Deci & Ryan,
1 2
.37
.00 .52**
Fig. 1. Study 1: Plotted interaction in predicting intrinsic motivation.
92 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995

1985). In Study 2, we also expanded our assessment of extrinsic and in-
trinsic motivation, using additional items, and we measured intrinsic
motivation directly rather than asking about the change in interest.
Once again, we hypothesized that the intrinsic motivation of people
high in DT would be less affected by the induced extrinsic motivation.
3.1. Method
3.1.1. Participants
Participants were 27 undergraduates (13 males and 14 females,
mean age = 18.7 SD = 1.1) attending the University of Missouri. They
were introductory psychology students, and received required research
credit in exchange for their participation.
3.1.2. Procedure
Participants were run one at a time. First, all participants completed
a survey containing the DT measure. Then the experimenter showed the
participant the Soma puzzle, which consists of several wooden pieces
which can be arranged into various 3-dimensional shapes. Some 2-
dimensional pictures of shapes were shown. To make the task easier,
the rst important step by which the Soma pieces could be manipulated
to arrive at those shapes was marked on the pictures. Participants were
then asked to try to construct 3 different shapes using the Soma puzzle
pieces. The type of rewards used has the same set-up as Deci's classic
studies (e.g., Deci, 1971), which was classied as completion-
contingent reward (Deci, Koestner, & Ryan, 1999). All participants
were told that they would have 10 min for each shape, and that they
would earn $1 for each shape that they successfully constructed in
time. If a participant was unable to construct a shape in the 10 min, he
or she was instructed to move on to the next shape. After the comple-
tion of the puzzle tasks, participants completed a post-task question-
naire that included the measures of extrinsic motivation and intrinsic
motivation. Finally, participants were de-briefed.
3.1.3. Measures
3.1.3.1. Dialectical thinking. These items were the same as those used
in Study 1, measured with a 7 point Likert scale. The Cronbach's
alpha of the scale was 0.82 in Study 2.
and correlations are shown in Table 1. All measures were standardized
in the analyses below.
3.2.2. Tests of primary hypotheses
3.2.2.1. Intrinsic motivation. A multiple regression with DT, extrinsic mo-
tivation, and their interaction as the predictors revealed a signicant
main effect of extrinsic motivation, = .45, t(23) = 2.46,
p b 0.05, and a signicant interaction effect, = .38, t(23) = 2.22,
p b 0.05. Simple slope tests (Aiken & West, 1991) were conducted to
probe the signicant interaction effect. Results showed that the associ-
ation between extrinsic motivation and the self-reported interest was
signicantly negative when people's DT was one standard deviation
below the mean, = .78, t(23) = 4.16, p b 0.001, while the associ-
ation was not signicant when people's DT was one standard deviation
above the mean, = .11, t(23) = .41, ns. Fig. 2 shows the plotted
interaction with the means of intrinsic motivation at different dialectical
thinking levels ( 1SD and 1SD) and different extrinsic motivation
levels (1SD and 1SD).
3.3. Brief discussion
Study 2 employed a real remunerative activity and tested the mod-
erating role of DT in the associations between extrinsic motivation
and intrinsic motivation. The results of Study 2 replicated the results
of Study 1, that DT moderated the association between extrinsic motiva-
tion and intrinsic motivation. These results suggested that DT buffers
people from the undermining effect of induced extrinsic motivation by
remunerative activity, with regard to both their feelings of enjoyment.
4. Study 3
Although Studies 1 and 2 both demonstrated a moderating role of DT
in the association between extrinsic motivation and intrinsic interest, all
participants were from the same cultural context. Previous research
showed that East Asians are more likely to be dialectical thinkers than
Westerners (see Spencer-Rodgers, Williams, et al., 2010, for a review).
Specically, Chinese participants were found to have higher DT scores
than European Americans (Spencer-Rodgers & Peng, 2004). In Study 3,
we recruited both European Americans and Chinese as participants to

3.1.3.2. Extrinsic motivation. Participants' extrinsic motivation for playing

with the Soma puzzle was measured by two items. Specically, partici-
pants were asked to rate the extent to which they agree with the state-
ments I solved the Soma puzzles only because I wanted to get the
monetary reward and I solved the Soma puzzles only because the ex-
periment required me to, on a 7-point Likert scale, with 1 indicating
strong disagreement and 7 indicating strong agreement. The Cronbach's
alpha of this measure was .71.

3.1.3.3. Intrinsic motivation. The interest/enjoyment subscale of the In-

trinsic Motivation Inventory (IMI, e.g., Ryan, 1982; Ryan, Koestner, &
Deci, 1991) was used to assess participants' intrinsic motivation. This
measure has been used in previous laboratory studies on the
undermining effect of reward on intrinsic interest (e.g., Deci, 1971;
Wiechman & Gurland, 2009). It includes 7 items, one example being I
found the puzzle-solving activity very interesting. The Cronbach's
alpha of this measure was .97.

3.2. Results

3.2.1. Preliminary analyses

Age and gender had no main effects on intrinsic motivation, and did
not interact with DT or extrinsic motivation when predicting intrinsic
motivation. Thus they were ignored below. Every participant completed
at least 1 shape, and thus received at least 1 dollar. Descriptive statistics Fig. 2. Study 2: Plotted interaction in predicting intrinsic motivation.
 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995 93

test the moderating role of DT by investigating cultural differences in
the association between extrinsic motivation and intrinsic motivation.
Instead of assessing self-reported extrinsic motivation, we created
paired scenarios which included a scenario with rewards and a scenario
without rewards. We reasoned that participants should have more ex-
trinsic motivation when imagining the scenario with rewards than
when imagining the scenario without rewards. We conducted mixed
2 (Culture: Chinese vs. European American) 2 (Scenario: reward vs.
no reward) analyses, with repeated measures on the second factor, to
test our interaction hypothesis, namely, that the intrinsic motivation
of the Chinese participants would be less affected by the reward/no re-
ward factor.
4.1. Method
4.1.1. Participants and procedure
Participants were 72 Chinese undergraduate students (15 male and
57 female, mean age = 20.4, SD = 0.7) from the Beijing Normal Univer-
sity and 182 American undergraduates who identied themselves as
European American (62 male and 120 female, mean age = 19.9,
SD = 1.4) from the University of Missouri; they were all taking an intro-
ductory psychology class. The questionnaire was translated into Chinese
by a bilingual person, and then back-translated by a different bilingual
person; discrepancies in meaning were discussed and resolved to
yield an equivalent questionnaire in both languages. The brief question-
naire contained four imaginary scenarios, which were similar to the sce-
narios used in Study 1. In two scenarios, participants were asked to
imagine participating in an interesting activity (e.g., ice-skating) for
which a reward was provided. In the other two paired scenarios, partic-
ipants were asked to imagine participating in the same activity with no
rewards. After they had read the description of a scenario and had imag-
ined the situation, they rated the extent to which their intrinsic motiva-
tion would change after they had done the activity compared with their
initial motivation (After you have participated in this activity, your in-
terest in this activity that you were recently really interested in would
). Participants answered on a 5-point Likert scale (1 = decrease a
lot, 2 = decrease a little, 3 = not change, 4 = increase a little, 5 = in-
crease a lot). The two ratings were averaged for each type of scenario
(reward, no reward). The two ratings were transformed for the results
below by subtracting 3, so that the 0 point indicates no change.
4.2. Results
ns. Fig. 3 shows the mean difference of intrinsic motivation between
scenarios and culture.
4.3. Brief discussion
Study 3 used a very different approach to test the moderating role of
DT in the association between extrinsic motivation and intrinsic motiva-
tion. Results were in line with what we found in Studies 1 and 2, in that
only the European-American sample, showed a negative association be-
tween intrinsic motivation and extrinsic motivation, not the Chinese
sample. Specically, European-Americans reported greater intrinsic
motivation in the no reward scenarios while Chinese reported greater
intrinsic motivation in the reward scenarios. Given the known differ-
ences in DT between these two cultural groups, this supports our hy-
pothesis in a very different way.
5. General discussion
The current study contributes new information concerning the neg-
ative association between extrinsic motivation and intrinsic motivation,
showing that DT moderates this association. The intrinsic motivation of
people higher in DT was reduced less by extrinsic motivation than the
intrinsic motivation of people lower in DT. This occurred when partici-
pants imagined gaining rewards for doing enjoyable activities (Study
1) and when they were given real rewards for doing interesting activi-
ties (Study 2). The results of Study 3 replicated the general pattern,
using cultural membership as a stand-in for DT. DT stems from the belief
that the world is complex and is changing over time, such that particular
situations can be interpreted in multiple ways (Spencer-Rodgers, Wil-
liams & Peng). In a rewarding situation, people who are low on DT
may believe that since they are doing the task in the present for re-
wards, this should continue to be the case in the future. In contrast, peo-
ple who are high on DT may believe that their intrinsic motivation and
extrinsic motivation are not competing, and their motivations can be
quickly adjusted according to a new situation. In the present studies,
participants with high DT did not report less extrinsic motivation in
the rewarding situations than those who are low on DT. This indicates
that people who are high on DT are indeed affected by rewards; howev-
er, what makes them different is their ability to maintain their intrinsic
motivation.

The two samples did not signicantly differ in gender composition.

They differed in age, F(1, 249) = 44.50, p b .01, with the Chinese sample
being slightly younger (Chinese M = 19.56, SD = .77; American M =
20.14, SD = 1.01). Gender had neither main nor interactive effects on
intrinsic motivation, but age was signicantly associated with intrinsic
motivation in the reward condition, so age was entered as a covariate
in the following analysis.
A mixed MANOVA revealed a signicant effect of scenario, F(1,
242) = 7.39, p b 0.01: across the two cultures, people reported more
positive changes in intrinsic motivation in the reward scenario (M =
.43, SD = .05) than in the no reward scenario (M = .29, SD = .04).
The between-group effect for culture was not signicant, F(1, 242) =
2.24, ns. Most importantly, a signicant culture (European American
vs. Chinese) scenario (reward vs. no reward) interaction emerged,
Wilks' Lambda value = .95, Pillai's Trace value = .05, F(1, 242) =
11.99, p = .001. Post-hoc t-tests showed that the two samples did not
differ in their intrinsic motivation in the no reward scenarios,
F(1,247) = 1.27, ns, while there was a signicant group difference in
the reward scenarios, F(1,252) = 19.591, p b 0.01. In the reward
scenarios, the Chinese participants reported more intrinsic motivation
than in the no reward scenarios, t(71) = 5.33, p b 0.001; however, the
European American participants reported less intrinsic motivation in Fig. 3. Study 3: Intrinsic motivation change for the American and Chinese participants in
the reward scenarios than in the no reward scenarios, t(176) = 1.63, the reward versus no reward conditions.
94 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995
One important implication of the results of our research is that
people's perceptions of reward stimuli may be more important than
the reward stimuli themselves. CET (Deci & Ryan, 1985) states that peo-
ple interpret rewards in varying ways, sometimes as conveying
competence information, and sometimes as conveying information
about their ability to choose, and that the nature of these perceptions in-
uences people's intrinsic motivation (Reeve, Nix, & Hamm, 2003;
Vansteenkiste & Deci, 2003). One recent study addressed a relevant in-
dividual difference, showing that individual causality orientations mod-
erated the undermining effect of rewards (Hagger & Chatzisarantis,
2010). These investigators found that people who were dispositionally
inclined to perceive situations as autonomy supportive and informa-
tional showed less of an undermining effect by rewards than people
who are inclined to perceive situations as controlling. The current stud-
ies explored a new type of individual difference moderator, DT. Instead
of helping people feel less externally motivated in rewarding situations,
DT may reduce the undermining effect by helping people segregate ex-
ternal and intrinsic motivation. In our studies, participants who were
high on DT still reported high extrinsic motivation, but this motivation
apparently did not negatively impact their intrinsic motivation.
The present research provided the rst evidence that peoples' dia-
lectical thinking style moderates the effect of external motivation on in-
trinsic motivation, extending previous studies of person by situation
interaction effects upon people's motivation. Our ndings have impor-
tant implications for applied areas such as education, work, and athletic
competition, where tangible rewards and strong competition are prev-
alent. Dialectical thinking apparently buffers people from the potential-
ly undermining impact of external incentives, by giving them the belief
that intrinsic and extrinsic motivation can co-exist. People who are high
on DT are thus enabled to maintain their original interest in an activity,
in circumstances where external motivators become salient. Presum-
ably, DT can be primed or trained (Paletz and Peng, 2009; Cheng,
2009) and such a training may help students, workers and athletes
maintain their internal passion when external motivators (e.g., rewards
or deadlines) are unavoidable.
5.1. Limitations and future research
Compared to research on the effects of external incentives or per-
ceived extrinsic motivation on intrinsic motivation, studies exploring
the moderating roles of individual differences have been limited. Thus,
there are many possible directions for future studies. One avenue
would be to look for mediators of the present moderation effect. Poten-
tial mediators could include some situational perception or feeling mea-
sures, for example, feelings of being controlled or coerced. According to
SDT, when people feel that the situation is controlling, their intrinsic in-
terest will be undermined (Deci & Ryan, 1985). However, people who
are high on DT may feel less controlled in a rewarding situation or
may not persist in their feeling of being controlled after the rewards
have been removed. Cognitive dissonance in the rewarding situation
might also be a potential mediator. It can perhaps be assessed via phys-
iological arousal indicators, self-reported discomfort or brain activity
changes (Glasford, Pratto, & Dovidio, 2008; Jarcho, Berkman, &
Lieberman, 2011; McMillen & Geiselman, 1974; Rodriguez & Morales,
1972; Suinn, 1965). People who are high on DT may feel little arousal
or cognitive dissonance in doing an initially interesting task under a re-
warding situation, which might explain why their intrinsic motivation
was less likely to be undermined by the extrinsic motivation.
Another study limitation is that we only tested the role of DT in the
association between extrinsic motivation and intrinsic motivation,
using situations where tangible rewards (i.e., money) were introduced.
However, there are many other sources of extrinsic motivation, for ex-
ample, desires for promotions in the work place or desires to earn
teachers' praise in class. Future research can consider testing the role
of DT with different sources of the extrinsic motivation or in other po-
tentially detrimental environments.
Additional study limitations include the fact that we only employed
imaginary scenarios in Study 1, although Study 2 rectied this with a
real situation. Study 1 and Study 2 both only focused on rewarding sit-
uations, with no comparison to no reward control situations; this was
because we were only interested in individual differences in responses
to the association between extrinsic motivation and intrinsic motiva-
tion, and a rewarding situation just to induce participants' extrinsic mo-
tivation. Another limitation was that participants' DT was not measured
in Study 3, since we relied on prior ndings of Chinese/US differences.
Future studies can examine whether participants' dispositional DT me-
diates cultural differences in reward effects on intrinsic motivation. Fi-
nally, N for these studies was fairly small, especially in Studies 1 and
2; thus these effects in particular require further replication.
Despite these limitations, the present studies discovered a new
moderator in the association between external incentives and intrinsic
motivation. DT may help people tolerate the seeming contradiction be-
tween salient external incentives and subsequent intrinsic motivation,
and help to restore their intrinsic motivation in the new situation
when rewards are removed. The ndings suggest that individual cogni-
tive thinking styles inuence people's perception of external incentives,
and thus environments which are highly detrimental to some people
are not necessarily harmful for other people.
Acknowledgments
We thank Todd Schachtman for the comments on the rst version of
the manuscript, and David Bergin, Xi Bi, and Marc Halusic for their help
on various aspects of the project.
References
Aiken, L.S., & West, S.G. (1991). Multiple regression: Testing and interpreting interactions.
Thousand Oaks, CA, US: Sage Publications, Inc.
Bassett, C. (2006). Laughing at gilded butteries: Integrating wisdom, development and
learning. In C. Hoare (Ed.), Oxford handbook of adult development and learning
(pp. 281306). New York, NY: Oxford University Press.
Boucher, H.C. (2011). The dialectical self-concept II: Cross-role and within-role consisten-
cy, well-being, self-certainty, and authenticity. Journal of Cross-Cultural Psychology,
42(7), 12511271. http://dx.doi.org/10.1177/0022022110383316.
Cameron, J., Pierce, W.D., Banko, K.M., & Gear, A. (2005). Achievement-based rewards and
intrinsic motivation: A test of cognitive mediators. Journal of Educational Psychology,
97, 641655. http://dx.doi.org/10.1037/0022-0663.97.4.641.
Cheng, C. (2009). Dialectical thinking and coping exibility: A multimethod approach.
Journal of Personality, 77, 471494. http://dx.doi.org/10.1111/j.1467-6494.2008.
00555.x.
Deci, E.L. (1971). Effects of externally mediated rewards on intrinsic motivation. Journal of
Personality and Social Psychology, 18, 105115. http://dx.doi.org/10.1037/h0030644.
Deci, E.L. (1972). Intrinsic motivation, extrinsic motivation, and inequity. Journal of
Personality and Social Psychology, 22, 113120. http://dx.doi.org/10.1037/h0032355.
Deci, E.L., Koestner, R., & Ryan, R.M. (1999). A meta-analytic review of experiments exam-
ining the effects of extrinsic rewards on intrinsic motivation. Psychological Bulletin,
125, 627668. http://dx.doi.org/10.1037/0033-2909.125.6.627.
Deci, E.L., & Ryan, R.M. (1985). Intrinsic motivation and self-determination in human behav-
ior. New York: Plenum.
Glasford, D.E., Pratto, F., & Dovidio, J.F. (2008). Intragroup dissonance: Responses to
ingroup violation of personal values. Journal of Experimental Social Psychology, 44,
10571064. http://dx.doi.org/10.1016/j.jesp.2007.10.004.
Hagger, M.S., & Chatzisarantis, N.L.D. (2010). Causality orientations moderate the
undermining effect of rewards on intrinsic motivation. Journal of Experimental Social
Psychology, 47, 485489. http://dx.doi.org/10.1016/j.jesp.2010.10.010.
Heider, F. (1958). The psychology of interpersonal relations. New York: John Wiley.
Jarcho, J.M., Berkman, E.T., & Lieberman, M.D. (2011). The neural basis of rationalization:
Cognitive dissonance reduction during decision-making. Social Cognitive and Affective
Neuroscience, 6, 460467. http://dx.doi.org/10.1093/scan/nsq054.
Ji, L. -J., Nisbett, R.E., & Su, Y. (2001). Culture, change, and prediction. Psychological Science,
12, 450456. http://dx.doi.org/10.1111/1467-9280.00384.
Kitayama, S., Markus, H.R., & Kurokawa, M. (2000). Culture, emotion, and well-being:
Good feelings in Japan and the United States. Cognition & Emotion, 14, 93124.
http://dx.doi.org/10.1080/026999300379003.
Kruglanski, A.W., Friedman, I., & Zeevi, G. (1971). The effects of extrinsic incentive on
some qualitative aspects of task performance. Journal of Personality, 39, 606617 (Re-
trieved from http://login.ezproxy1.lib.asu.edu/login?url=http://search.proquest.
com/docview/615806207?accountid=4485).
Lepper, M.R., Greene, D., & Nisbett, R.E. (1973). Undermining children's intrinsic interest
with extrinsic reward: A test of the overjustication hypothesis. Journal of
Personality and Social Psychology, 28, 129137. http://dx.doi.org/10.1037/h0035519.
 Y. Li et al. / Learning and Individual Differences 39 (2015) 8995
Leu, J., Mesquita, B., Ellsworth, P. C., ZhiYong, Z., Huijuan, Y., Buchtel, E., & Masuda, T.
(2010). Situational differences in dialectical emotions: Boundary conditions in a cul-
tural comparison of north americans and east asians. Cognition and Emotion, 24(3),
419435. http://dx.doi.org/10.1080/02699930802650911.
Levy, Sheri R., Ramrez, Luisa, Rosenthal, Lisa, & Karafantis, Dina M. (2013). In navigating
the social world: What infants, children, and other species can teach us. In Mahzarin
R. Banaji, & Susan A. Gelman (Eds.), New York, NY, US: Oxford University Press.
McMillen, D.L., & Geiselman, J.H. (1974). Effect of cognitive dissonance on alpha frequen-
cy activity: The search for dissonance. Personality and Social Psychology Bulletin, 1,
150151.
Nisbett, R.E., Peng, K., Choi, I., & Norenzayan, A. (2001). Culture and systems of thought:
Holistic versus analytic cognition. Psychological Review, 108, 291310. http://dx.doi.
org/10.1037//0033-295X.108.2.291.
Paletz, S. B. F., & Peng, K. (2009). Problem nding and contradiction: Examining the rela-
tionship between nave dialectical thinking, ethnicity, and creativity. Creativity
Research Journal, 21, 139151.
Peng, K., & Nisbett, R.E. (1999). Culture, dialectics, and reasoning about contradiction.
American Psychologist, 54, 741754. http://dx.doi.org/10.1037//0003-066X.54.9.741.
Plaks, J.E., Levy, S.R., & Dweck, C.S. (2009). Lay theories of personality: Cornerstones of
meaning in social cognition. Social and Personality Psychology Compass, 3(6),
10691081. http://dx.doi.org/10.1111/j.1751-9004.2009.00222.x.
Reeve, J., Nix, G., & Hamm, D. (2003). Testing models of the experience of self-
determination in intrinsic motivation and the conundrum of choice. Journal of
Educational Psychology, 95(2), 375392. http://dx.doi.org/10.1037/0022-0663.95.2.
375.
Rodriguez, Kant A., & Morales, d. B. (1972). The potential of cognitive dissonance to in-
duce anxiety. Revista Latinoamericana De Psicologa, 4, 103115 (Retrieved from
http://login.ezproxy1.lib.asu.edu/login?url=http://search.proquest.com/docview/
615838848?accountid=4485).
Ryan, R.M. (1982). Control and information in the intrapersonal sphere: An extension of
cognitive evaluation theory. Journal of Personality and Social Psychology, 43, 450461.
http://dx.doi.org/10.1037//0022-3514.43.3.450.
Ryan, R. M., & Connell, J. P. (1989). Perceived locus of causality and internalization: Exam-
ining reasons for acting in two domains. Journal of Personality and Social Psychology,
57, 749761.
Ryan, R. M., & Deci, E. L. (2000). Intrinsic and extrinsic motivations: Classic denitions and
new directions. Contemporary Educational Psychology, 25(1), 5467. http://dx.doi.org/
10.1006/ceps.1999.1020.
95
Ryan, R.M., Koestner, R., & Deci, E.L. (1991). Varied forms of persistence: When free-
choice behavior is not intrinsically motivated. Motivation and Emotion, 15, 185205.
http://dx.doi.org/10.1007/BF00995170.
Smith, T.W., & Pittman, T.S. (1978). Reward, distraction, and the overjustication effect.
Journal of Personality and Social Psychology, 36(5), 565572. http://dx.doi.org/10.
1037/0022-3514.36.5.565.
Spencer-Rodgers, J., Peng, K., & Wang, L. (2010a). Dialecticism and the co-occurrence of
positive and negative emotions. Journal of Cross-Cultural Psychology, 41, 109115.
http://dx.doi.org/10.1177/0022022109349508.
Spencer-Rodgers, J., Williams, M.J., & Peng, K. (2004). The dialectical self: Contradiction,
change, and holism in the East Asian self-concept. In R. M. Sorrentino, D. Cohen, J.
M. Olsen, & M. P. Zanna (Eds.), Culture and social behavior: The Ontario Symposium,
Vol. 10. (pp. 227250). NJ: Lawrence Erlbaum: Mahwah.
Spencer-Rodgers, J., Srivastava, S., Boucher, H. C., English, T., Paletz, S. B., & Peng, K. (2010).
The dialectical self scale. Unpublished data, as cited in Spencer-Rodgers, J., Peng, K.,
Wang, L., & Hou, Y. (2004). Dialectical self-esteem and EastWest differences in psy-
chological well-being. Personality and Social Psychology Bulletin, 30, 1416-1432. doi:
http://dx.doi.org/10.1177/0146167204264243.
Spencer-Rodgers, J., Williams, M.J., & Peng, K. (2010c). Cultural differences in expec-
tations of change and tolerance for contradiction: A decade of empirical re-
search. Personality and Social Psychology Review, 14, 296312. http://dx.doi.org/
10.1177/108886831036298.
Suinn, R.M. (1965). Anxiety and cognitive dissonance. Journal of General Psychology, 73,
113116. http://dx.doi.org/10.1080/00221309.1965.9711258.
Tang, S., & Hall, V.C. (1995). The overjustication effect: A meta-analysis. Applied Cognitive
Psychology, 9(5), 365404. http://dx.doi.org/10.1002/acp.2350090502.
Vansteenkiste, M., & Deci, E.L. (2003). Competitively contingent rewards and intrinsic
motivation: Can losers remain motivated? Motivation and Emotion, 27, 273299.
http://dx.doi.org/10.1023/A:1026259005264.
Warneken, F., & Tomasello, M. (2008). Extrinsic rewards undermine altruistic tendencies
in 20-month-olds. Developmental Psychology, 44(6), 17851788. http://dx.doi.org/10.
1037/a0013860.
Wiechman, B.M., & Gurland, S.T. (2009). What happens during the free-choice period?
Evidence of a polarizing effect of extrinsic rewards on intrinsic motivation. Journal
of Research in Personality, 43, 716719. http://dx.doi.org/10.1016/j.jrp.2009.03.008.
Yang, S. -Y. (2008). A process view of wisdom. Journal of Adult Development, 15, 6275.
http://dx.doi.org/10.1007/s10804-008-9037-8.