Content uploaded by Maria Preißinger
Author content
All content in this area was uploaded by Maria Preißinger on Apr 20, 2018
Content may be subject to copyright.
1
This is the author’s manuscript. Please cite the published version only:
Preißinger, Maria & Harald Schoen. 2018. “Entity and incremental theory of
personality: Revisiting the validity of indicators.” Personality and Individual Differences
130: 21-25. https://doi.org/10.1016/j.paid.2018.03.042
Entity and incremental theory of personality: Revisiting the validity of indicators
Maria Preißinger
Harald Schoen
(University of Mannheim)
Abstract:
This article tests the validity of indicators of entitavist and incremental lay theories about the
malleability of personality (Dweck, Hong, & Chiu, 1993; Levy, Stroessner, & Dweck, 1998) in a two-
wave panel survey over a 12-month period. After controlling for systematic measurement error
stemming from different directions of item wording, the indicators form a single dimension.
Moreover, hypotheses concerning (non-)correlations with socio-demographic characteristics and
psychological dispositions largely receive support from the evidence.
Further, beliefs about the malleability of personality exhibit higher intra-individual stability than
attributions when controlling for measurement error in structural equation modeling. However, beliefs
do not influence these attributions.
1. Introduction
The idea that people hold beliefs about the malleability of personality has considerably
influenced scholarly debates in social psychology (Dweck et al., 1993; Dweck et al., 1995; Chiu et al.,
1997). Accordingly, some people believe that personality consists of fixed dispositions that cannot be
2
changed; these are so-called “entity theorists.” As dispositions are believed to be stable within a
person, it makes sense to attribute other people’s behavior to these dispositions. Core dispositions are
conceived as “the unit of analysis - the fundamental construct in terms of which they understand the
nature and the workings of the social world they live in” (Dweck et al., 1993, pp. 644–645).
“Incremental theorists,” by contrast, think that even the most basic characteristics of a person are
malleable and open for change. For incremental theorists, attributes of a person are not fixed entities
but merely convenient labels they attach to current mental states such as goals, needs, and emotions of
a certain individual in a specific situation (Dweck et al., 1993, p. 645). According to incremental
theorists, these mental states can change and do change in response to situational changes (Dweck et
al., 1993, pp. 651–652). Therefore, incremental theorists should assign more causal weight to the
situation when explaining others’ behavior than entity theorists do (Levy and Dweck, 1999, pp. 1164–
1165; Levy et al., 2001, pp. 160–161).
Despite the influence of this theory, there remain open questions at the conceptual level with
repercussions for its measurement. The aim of this paper is to resolve these issues which we shall
discuss in turn. First, it is unclear whether the existing indicators accurately reflect the dimensionality
of the theoretical concept because the dimensionality itself is theoretically contested. In early
theorizing, beliefs about the malleability of personality are one-dimensional. Recent accounts make a
case for a two-dimensional concept with both belief sets being independent knowledge structures with
separate cognitive representations (Anderson and Lindsay, 1998; Levy et al., 2001, pp. 163–164; Levy
et al., 2006, pp. 9–10; Poon and Koehler, 2006, 2008). The former view appears to rest on the
assumption that lay beliefs about personality can prove useful in guiding social perception only if
people cannot believe simultaneously that personality is fixed and highly malleable. By contrast,
proponents of the latter view argue that entity and incremental beliefs are so often discussed in
everyday life that they are available in everybody’s long-term memory and individuals differ only in
their chronic accessibility (Levy et al., 2001, pp. 163–164; Levy et al., 2006, pp. 9–10; Poon and
Koehler, 2006, 2008). Furthermore, a two-dimensional structure could emerge because people do not
hold a lay theory about personality per se, but about different parts of personality, some of which are
regarded as fixed and some as malleable. Extant evidence does not clearly support one conception or
3
the other, however. Most studies use a Likert-item battery comprising indicators worded in a way that
approval of these items means approval of entity-beliefs of personality and therefore cannot examine
the dimensionality (Erdley and Dweck, 1993; Dweck et al., 1995; Hong et al., 1997; Poon and
Koehler, 2006). Others use an extended item battery including indicators worded in both directions
(developed by Levy et al., 1998) but do not properly examine the dimensionality. When tackling this
issue, one has to consider the methodological challenges arising from survey response sets. In a
confirmatory factor analysis, a two-dimensional factorial structure could do a better job than a one-
dimensional model in reproducing inter-item-correlations not because the underlying concept is indeed
two-dimensional but because survey respondents tend to approve rather than to disapprove of survey
items (acquiescence bias) – irrespective of their content.1 Others use bipolar semantic differential
scales that may avoid acquiescence bias but are not helpful in empirically investigating the
dimensionality of the underlying construct (e.g. Dickhäuser et al., 2016; see also Lüftenegger and
Chen, 2017). In this paper, we examine the unresolved question of the concept’s dimensionality by
using the two-sided Likert-item battery and comparing different measurement models in a
confirmatory factor analysis approach while controlling for systematic measurement error caused by
acquiescence bias.
Second, it is not clear whether the indicators possess sufficient reliability because a reasonable
benchmark to judge results against is missing. Using data from Likert-batteries measured in panel
surveys, Poon and Koehler (2008) report an over-time Pearson’s correlation coefficient r=0.57 over an
eight-week period; Dweck et al. (1995) find a test-retest correlation of r=0.82 over two weeks,
whereas Levy et al. (1998) report r=0.71 over four weeks. While measuring over-time correlations is
important, their significance depends on whether they meet theory-based criteria. Although Poon and
Koehler (2008) compare their stability estimate of lay theories to some “widely studied individual-
difference variables” (p. 973), they do not specify expectations for this comparison. Building on the
idea that beliefs about the malleability of personality provide individuals with a framework for causal
attributions, we argue that these beliefs should prove more stable than causal attributions individuals
1 Examining related lay beliefs about the malleability of intelligence, Tempelaar et al. (2015) conduct this kind
of analysis but do not take acquiescence bias into account. Similarly, the fact that an exploratory factor analysis
results in two factors is not sufficient evidence for a two-dimensional structure (e.g. Stipek and Gralinski, 1996;
Dupeyrat and Mariné, 2005).
4
make in identical scenarios measured at multiple occasions. Previous reliability-studies did not report
this benchmark statistic, however. What is more, previous studies did not check whether the
measurement model is invariant over time and consequently whether the computation of stability
estimates is reasonable to begin with. Therefore, in this paper we close this gap and examine the
stability of malleability-beliefs about personality over a 12-months period in a confirmatory factor
analysis approach and provide analysis of the stability in attributions as a benchmark.
Third, prior research on the discriminant and convergent validity of indicators was conducted
with small convenience samples only (Dweck et al., 1995; Levy et al., 1998). In this paper, we employ
a large sample of the general population in order to include enough variation in the respective
variables to test these validation hypotheses appropriately. A first set of expectations derives from the
argument that incremental theory is not more correct than entity theory or vice versa, instead both
theories are conceived as “alternative ways of constructing reality, each with its potential costs and
benefits” (Dweck et al., 1995, p. 268; see also Levy et al., 2001, p. 157). Gender, education, and life
experience therefore should not provide superior insight and should be unrelated to indicators of
malleability beliefs if the latter capture the concept validly (Dweck et al., 1995; Levy et al., 1998). In
contrast, malleability beliefs should be related to cognitive needs. Because global personality traits can
be judged as good or bad in a straightforward manner, holding an entity view about personality may
fulfill the need to evaluate – at least as far as people as attitude objects are concerned (Dweck et al.,
1993; Hong et al., 1997). By contrast, incremental theorists’ tendency to understand dispositions as
mere labels of dynamic psychological states might fulfill a need to engage in effortful cognitive
activity (need for cognition).
As the above discussion suggests, prior research has not been completely successful in
demonstrating that the proposed indicators have theoretically desirable properties. We will thus take a
fresh look at the measurement. Relying on data from a two-wave panel survey, we examine the
validity of indicators by investigating their dimensionality, whether they exhibit sufficiently high over-
time stability and correlate with other concepts in the expected way. In terms of methodology, we
improve on prior research by avoiding problems arising from random and structural measurement
error by employing structural equation modeling.
5
2. Method
2.1. Procedure
We collected survey data from 1,098 adult Germans quota-sampled from an online-access
panel administered by Respondi AG leading to a three-way uniform distribution in gender, three
education groups (low, middle and high education) and five age groups (18-29, 30-39, 40-49, 50-59,
60 years or older). This sample size provides sufficient variation in sociodemographic characteristics
to test Dweck et al.’s (1995) claim that malleability beliefs are independent of these characteristics.
Respondents were awarded 1.50€ in exchange for their participation in the 15-minute interview and
answered the survey on their personal device. Respondents were invited to take part in a follow-up
interview about 12 months later in February 2017. 548 respondents completed the second wave. We
considered only participants that were able to verify their identity by stating identical information
regarding gender, year of birth, month of birth and first letter of birthplace in wave 2 as previously
stated in wave 1. Respondents who completed the re-interview and those failing to do so do not differ
substantially in beliefs about the malleability of personality as measured in wave 1 (results not
reported in tabular form).
2.2. Lay theory measures
An item battery in German was modeled after the 8-item battery by Levy et al. (1998) via a
translation and back translation procedure. This battery includes Likert items taking a fixed view on
personality as well as items taking an incremental view. One item in Levy et al.’s original battery
quotes a proverb (“you can’t teach an old dog new tricks”). This item was not included in the German
battery because proverbs sound familiar and thus are more strongly endorsed than other items (mere
exposure effect: Zajonc, 1968); instead a new entity-worded item was created (item 3). Incremental
items had to be adapted as well because very similarly worded items in English turned out to be word-
for-word identical in the German translation. As a result, the German battery comprises seven items,
four worded in favor of an entity view, three worded in favor of an incremental view (Table 1).
6
Participants were asked to rate each item on a single screen on a 5-point scale with options
labeled “strongly agree”, “agree”, “partly agree/partly disagree”, “disagree”, and “strongly disagree”.
There was no “don’t know” option. Respondents could, however, continue to the next screen without
providing an answer. The scale of the original item battery by Levy et al. (1998) did not include a
“don’t know” option either but – in contrast to our measure – also lacked an exact mid-point. For
someone who has not thought about the malleability of personality before and is looking for a way to
express this lack of opinion, refusing to give an answer or clicking the middle category might
represent a viable strategy (Krosnick, 1991). If no middle category had been provided, respondents
might scatter between the adjacent scale points, falling – in the worst case, randomly – either into the
entity or the incremental half of the scale.
2.3. Other measures
In order to measure if someone attributes the cause of a behavior to personal dispositions or
the situation, respondents were asked to search their memory for someone jumping a queue and
someone offering their seat on a bus and provide what they perceive as the cause of this behavior in
their own words. Employing the approach used by Russell (1982), McAuley et al. (1984) and
Schaufeli (1988), respondents who gave an answer to the open-ended question were asked on the next
page, whether the observed behavior was caused by the character of the person or by the
circumstances of the situation. The response options were: “definitely character of the person”, “rather
character of the person”, “part character/part situation”, “rather situation”, “definitely situation”. The
open-ended answers of the respondents were checked for obvious junk (e.g., random letter
combinations) and remarks indicating the inability or refusal to state a cause for the specific behavioral
scenario. When we identified this kind of invalid response, we also set the answers to the closed-ended
follow-up question as missing. We also included indicators of concepts for which prior research
suggests (non-)correlations with person-malleability beliefs. Age and gender are measured via self-
reports. Education is measured by respondents’ highest school diploma. We differentiate between low
education (diploma after 9 years of schooling or no school diploma at all), middle education (diploma
after 10 years of schooling) and high education (university entrance certificate). For measuring the
7
need to evaluate and the need for cognition, four indicators each were included in the survey (see the
online supplement for the wording of these indicators).
3. Results & Discussion
3.1. Dimensionality
Table 1 includes summary information about all items. To make sure that higher values
indicate more support for entity theory, we recoded all items that were originally worded in a pro-
incrementalism way. All items show very similar mean values located in the middle of the scale and
similar standard deviations. For all items, the middle category is the mode value.
Table 1
Descriptive information about lay theory items
Wording
N
Mean (SD)
Middle
item1
Our personality is a part of us that we
cannot change.
1097
2.97 (1.12)
36%
item2*
If people want to, they can change
even their most basic characteristics.
1091
2.89 (1.02)
38%
item3
People can change their circumstances
of life but not their personality.
1096
3.24 (1.05)
34%
item4*
People often change their basic
characteristics in the course of a
lifetime.
1093
3.00 (0.93)
41%
item5*
People can change so much that you
don’t recognize them anymore.
1097
2.66 (1.09)
32%
item6
Everyone is a certain type of person,
and there is not much that can be done
to really change that.
1091
2.91 (0.92)
40%
item7
People can do things differently, but
their character can’t really be changed.
1094
3.22 (1.02)
35%
Note. Items used in study are in German. Items run from 1 to 5, higher values indicate higher
agreement with entity conceptions of personality. * = items are reversed. “Middle” = share of answers
in the middle category.
In order to test the dimensionality, we used confirmatory factor analysis. The fit of a single-
factor model in which a single latent variable drives the responses to all seven items and measurement
error is assumed to be entirely random is poor (Table 2). A two-factorial model with an entity theory
factor which is measured by the four pro-entity indicators and an incrementalism theory factor which
is measured by the three pro-incrementalism indicators fits the data much better (see χ²-difference test)
but yields a rather strong correlation between latent factors. The two factors appear to be not as
8
independent from each other as a strong reading of two-dimensionality claims suggests. In addition,
the two-factorial solution may do a better job in reproducing the correlations among the reverse-coded
items because of respondents’ tendency to approve rather than disapprove of items in a survey
(acquiescence bias). As the data to not allow us to estimate a model in which the concept is two-
dimensional and control for systematic measurement error at the same time2, we fit a single-factor
model in which we control for acquiescence bias by specifying correlated measurement errors between
all negative items in the single-factor model (item 2, 4 and 5; see Brown, 2006). The correlations pass
conventional levels of statistical significance (Table 3) and the model has a superior fit compared to
the first single-factor model (see χ²-difference test). Because the two-factor and the error model are not
nested, a χ²-difference test between the two is not appropriate. Comparing AIC, however, the error
model fits the data better than the two-factor model. It also possesses favorable values on approximate
fit indices. We therefore conclude that the evidence lends more support to a one-dimensional
conception with systematic measurement error for than to a two-dimensional conception. For the
remainder of the analyses, we employed the single-factor measurement model with correlated errors.
2 One could do so, however, by crafting two sets of items with each item of the first set having a “mirror item” in
the second set in its negated form while holding everything else about the wording constant. By estimating a
two-factorial model with correlated measurement errors between each item and its “mirror item”, the resulting
correlation between the two latent factors would be free from acquiescence bias and indicative of the
dimensionality of the concept. Answering items and mirror items, however, can be tedious for respondents.
9
Table 2
Comparison of fit of different measurement models of lay theory
1) Single
factor
2) Two
factors
3) Error
model
Correlation
of factors
.64***
N
1073
1073
1073
χ²(df)
266(14)***
103(13)***
48(11)***
χ²-diff(df)
163(1)***
218(3)***
RMSEA
.13
.08
.06
CFI
.87
.95
.98
SRMR
.08
.05
.03
AIC
19947
19786
19736
Note. Confirmatory factor analysis, maximum likelihood. All items are coded in a way that higher
values indicate higher agreement with entity conceptions of personality.
*** p < 0.001
Table 3
Results of the error model of lay theory
Error model
Loadings
item1
.51 ***
item2
.51 ***
item3
.67 ***
item4
.27 ***
item5
.32 ***
item6
.73 ***
item7
.82 ***
Variances
error.item1
.74 ***
error.item2
.74 ***
error.item3
.55 ***
error.item4
.93 ***
error.item5
.90 ***
error.item6
.47 ***
error.item7
.34 ***
Lay Theory
1.00 (fixed)
Correlations
errors item2&item4
.29 ***
errors item2&item5
.21 ***
errors item4&item5
.35 ***
Note. Confirmatory factor analysis, maximum likelihood. All items are coded in a way that higher
values indicate higher agreement with entity conceptions of personality. Fully standardized estimates
reported.
*** p < 0.001
3.2. Over-time stability
It is appropriate to compute a stability coefficient between two latent factors (here:
malleability-person theory measured in wave 1 and wave 2), if the measurement model is metric
10
invariant over-time. As can be seen in Table 4, this is the case: A χ²-difference test between a model
with factor loadings constrained to be equal across time and auto-correlated errors (model 2) and a
model with free loadings and auto-correlated errors (model 3) does not pass conventional levels of
statistical significance. What is more, RMSEA and AIC point to a worse fit of model 3 than model 2.
A structural path from wave 1 to wave 2 yields a standardized coefficient beta = .63 (95% CI: [.55,
.70], p=.000). This estimate is considerably higher than the over-time stability of causal attributions
from the first to the second wave (offer a seat: Pearson’s r=.36, p=.000; jump a queue: r=.35, p=.000)
although one has to bear in mind that for attributions we have only one item per scenario and therefore
cannot control for measurement error. Taken together, the evidence indicates a fair degree of stability
in comparison with attribution measures.
Table 4
Measurement Invariance Testing
Model 1:
same loadings,
same errors
Model 2:
same loadings,
autocorrelated errors
Model 3:
free loadings,
autocorrelated errors
N
524
524
524
χ²(df)
337(83) ***
171(69) ***
167(63) ***
χ²-diff(df)
166(14)***
4(6)
RMSEA
.08
.05
.06
CFI
.89
.95
.95
SRMR
.08
.06
.06
AIC
18578
18440
18448
Note. Confirmatory factor analysis, maximum likelihood.
*** p < 0.001
Table 5
Effects of lay theory on causal attribution
Wave 1
Wave 2
Offer a seat
Jump a
queue
Offer a seat
Jump a
queue
Lay Theory
-.03
-.03
-.13 **
-.01
[-.10, .04]
[-.10, .04]
[-.23, -.04]
[-.11, .08]
N
976
1012
500
510
χ²(df)
47(17) ***
58(17) ***
41(17)***
45(17)***
RMSEA
.04
.05
.05
.06
CFI
.98
.98
.97
.97
SRMR
.03
.03
.03
.04
Note. Structural equation model, maximum likelihood. Only structural paths are reported. Fully
standardized estimates. In brackets: 95% confidence intervals.
11
** p < 0.01, *** p < 0.001
Using the stability of attribution as a yardstick to evaluate the stability of person-malleability
beliefs builds on the premise that these beliefs influence these attributions. In examining this, we
expected negative coefficients indicating that the higher agreement with entity conceptions of
personality the less impact people should attribute to situational factors.3 As Table 5 demonstrates, the
estimated coefficients, however, are close to zero and fail to reach statistical significance. In order to
check the robustness of our findings, we re-ran the analyses concerning attributions in wave 2. Despite
the smaller number of observations in wave 2, the coefficient of lay theory in the offering-a-seat
scenario reaches statistical significance and points into the expected direction. Results from wave 1 as
presented in Table 5 do not change substantially if we restrict the sample to respondents who
participated in both waves; therefore, individual differences between panelists and non-panelists
cannot account for the different findings. Because prior research did not only employ continuous
measures of lay theory, but also dichotomous variables that were created by splitting the samples at
different cut-off points (Hong, 1994; Chiu et al., 1997; Hong et al., 1997; Levy et al., 1998; Levy and
Dweck, 1999), we explored the robustness of our findings by rerunning these analyses with eight
dichotomous versions of the lay theory measure. However, across all operationalizations we did not
find systematic evidence for effects of lay beliefs on attributions (see online supplement).4
3.3. Correlations with other concepts
The results reported in Table 6 demonstrate that the indicators used in prior research, by and
large, are related to the other concepts in expected ways. There are some minor exceptions, however.
The correlation with need to evaluate is positive as expected, but does not pass conventional levels of
statistical significance. However, the relationship was assumed to be weak because not all people
holding an entity theory might feel the need to evaluate other objects besides people. In contrast to our
expectation, highly educated people appear to be more likely to hold incremental beliefs about
personality. Taken together, the majority of expectations are met.
3 For descriptive information about attribution measures see online supplement.
4 We also examined whether intra-individual change in malleability-beliefs over time explains intra-individual
change in attributions. Again, this is not the case (results not presented).
12
Table 6
Correlations of lay theory with other concepts
Expected
Actual results
Need for Cognition
-
-.22 ***
Need to Evaluate
+
.06
Age
0
.02
Low Education
0
.05
High Education
0
-.08 *
Female
0
-.02
N
1027
χ²(df)
432(115)***
RMSEA
.05
CFI
.90
SRMR
.05
Note. Structural equation model, maximum likelihood. Fully standardized results. Left out reference
category: medium education. Data from wave 1.
* p < 0.05, *** p < 0.001
4. Conclusion
In this paper, we used a large-scale sample from the general population in Germany to study
issues in the measurement of beliefs about the malleability of personality. The evidence supported a
one-dimensional conceptualization of personality-malleability beliefs. At the same time, acquiescence
bias was far from absent in response behavior. Therefore, instruments avoiding such biases in the first
place may be desirable alternatives for future research (see Lüftenegger and Chen, 2017). In terms of
discriminant and convergent validity, indicators were correlated with other concepts in expected ways.
In addition, measurement models proved metric-invariant over time and malleability beliefs were more
stable than the attribution measures. However, lay beliefs and attributions were virtually unrelated to
each other. This suggests that comparing stability estimates of the two concepts does not make much
sense. In addition, it raises concerns about the validity of the belief measures. The effect of lay beliefs
on attributions is of key importance to theorizing about lay beliefs. Because entity and incremental
theorists differ in their conceptions of personality, they should differ in their inclination to assign
causal relevance to people’s dispositions or the situation. A non-finding on this matter thus casts
severe doubts on the validity of the selected indicators. Our methodology differs from that employed
in prior studies. Previous research mainly compared the strength of dispositional inference between
individuals with diverging theories about the malleability of personality (e.g. Erdley and Dweck, 1993;
13
Chiu et al., 1997; Gervey et al., 1999) which is arguably not the same thing as the type of causal
attribution. In this analysis we deviated from this approach but focused on the type of attribution
individuals infer as did Hong (1994) and Levy and Dweck (1999). In contrast to them, however, we
refrained from relying on content-coding of responses to open-ended questions to identify the type of
attribution. As respondents and coders may hold different beliefs about the malleability of personality,
content-coding of attributional responses is subject to the “fundamental attribution researcher error”
(Russell, 1982). In order to avoid this problem, we relied on respondents’ self-reports about whether
their open-ended attributional answers have something to do with the character of the person or with
the situation (Russell, 1982; McAuley et al., 1984; Schaufeli, 1988). We are thus confident that our
findings differ from previous research not because of poor methodology employed in our analysis.
Rather, the findings reported by earlier studies could be due to biased measurement of attributions.
However, as we included only two different attributional scenarios and run only one panel survey in
one country, such a conclusion would be premature. Further research should look into these issues in
more detail by surveying attributional answers to many more scenarios as in Hong (1994) and
systematically comparing content-coding and respondent-coding approaches as well as measures of
dispositional inference strength. We thus conclude that lay beliefs about the malleability of personality
deserve more scholarly attention and more fine-grained analyses, in order to address key conceptual
and methodological issues.
14
References
Anderson, C.A., Lindsay, J.J., 1998. The development, perseverance, and change of naive theories.
In: Wegener, D.T., Petty, R.E. (Eds.), Naive theories and social judgement, pp. 8–30.
Brown, T.A., 2006. Confirmatory factor analysis for applied research. Guilford Press, New York.
Chiu, C.y., Hong, Y.y., Dweck, C.S., 1997. Lay dispositionism and implicit theories of personality.
Journal of Personality and Social Psychology 73 (1), 19–30.
Dickhäuser, O., Dinger, F.C., Janke, S., Spinath, B., Steinmayr, R., 2016. A prospective correlational
analysis of achievement goals as mediating constructs linking distal motivational dispositions to
intrinsic motivation and academic achievement. Learning and Individual Differences 50, 30–41.
10.1016/j.lindif.2016.06.020.
Dupeyrat, C., Mariné, C., 2005. Implicit theories of intelligence, goal orientation, cognitive
engagement, and achievement: A test of Dweck’s model with returning to school adults.
Contemporary Educational Psychology 30 (1), 43–59. 10.1016/j.cedpsych.2004.01.007.
Dweck, C.S., Chiu, C.y., Hong, Y.y., 1995. Implicit theories and their role in judgments and reactions:
A word from two perspectives. Psychological Inquiry 6 (4), 267–285.
10.1207/s15327965pli0604_1.
Dweck, C.S., Hong, Y.y., Chiu, C.y., 1993. Implicit theories individual differences in the likelihood
and meaning of dispositional inference. Personality and Social Psychology Bulletin 19 (5), 644–
656. 10.1177/0146167293195015.
Erdley, C.A., Dweck, C.S., 1993. Children's implicit personality theories as predictors of their social
judgments. Child Development 64 (3), 863–878.
Gervey, B., Chiu, C.y., Hong, Y.y., Dweck, C.S., 1999. Differential use of person information in
decisions about guilt versus innocence: The role of implicit theories. Personality and Social
Psychology Bulletin 25 (1), 17–27.
Hong, Y.y., 1994. Predicting trait versus process inferences: The role of implicit theories. Columbia
University, New York.
Hong, Y.y., Chiu, C.y., Dweck, C.S., Sacks, R., 1997. Implicit theories and evaluative processes in
person cognition. Journal of Experimental Social Psychology 33 (3), 296–323.
10.1006/jesp.1996.1324.
Krosnick, J.A., 1991. Response strategies for coping with the cognitive demands of attitude measures
in surveys. Applied Cognitive Psychology 5 (3), 213–236. 10.1002/acp.2350050305.
Levy, S.R., Chiu, C.y., Hong, Y.y., 2006. Lay theories and intergroup relations. Group Processes &
Intergroup Relations 9 (1), 5–24. 10.1177/1368430206059855.
Levy, S.R., Dweck, C.S., 1999. The impact of children's static versus dynamic conceptions of people
on stereotype formation. Child Development 70 (5), 1163–1180. 10.1111/1467-8624.00085.
Levy, S.R., Plaks, J.E., Hong, Y.y., Chiu, C.y., Dweck, C.S., 2001. Static versus dynamic theories and
the perception of groups: Different routes to different destinations. Personality and Social
Psychology Review 5 (2), 156–168. 10.1207/S15327957PSPR0502_6.
Levy, S.R., Stroessner, S.J., Dweck, C.S., 1998. Stereotype formation and endorsement: The role of
implicit theories. Journal of Personality and Social Psychology 74 (6), 1421–1436. 10.1037/0022-
3514.74.6.1421.
Lüftenegger, M., Chen, J.A., 2017. Conceptual Issues and Assessment of Implicit Theories. Zeitschrift
für Psychologie 225 (2), 99–106. 10.1027/2151-2604/a000286.
McAuley, E., Duncan, T.E., Russell, D.W., 1984. Measuring causal attributions: The revised causal
dimension scale (CDSII). Personality and Social Psychology Bulletin 18 (5), 566–573.
10.1177/0146167292185006.
Molden, D.C., Plaks, J.E., Dweck, C.S., 2006. "Meaningful" social inferences: Effects of implicit
theories on inferential processes. Journal of Experimental Social Psychology 42 (6), 738–752.
Plaks, J.E., Stroessner, S.J., Dweck, C.S., Sherman, J.W., 2001. Person theories and attention
allocation: Preferences for stereotypic versus counterstereotypic information. Journal of
Personality and Social Psychology 80 (6), 876–893. 10.1037//0022-3514.80.6.876.
Poon, C.S.K., Koehler, D.J., 2006. Lay personality knowledge and dispositionist thinking: A
knowledge-activation framework. Journal of Experimental Social Psychology 42, 177–191.
Poon, C.S.K., Koehler, D.J., 2008. Person theories: Their temporal stability and relation to intertrait
inferences. Personality and Social Psychology Bulletin 34 (7), 965–977.
10.1177/0146167208316690.
15
Russell, D., 1982. The causal dimension scale: A measure of how individuals perceive causes. Journal
of Personality and Social Psychology 42 (6), 1137–1145. 10.1037/0022-3514.42.6.1137.
Schaufeli, W.B., 1988. Perceiving the causes of unemployment: An evaluation of the causal
dimensions scale in a real-life situation. Journal of Personality and Social Psychology 54 (2), 347–
356. 10.1037/0022-3514.54.2.347.
Stipek, D., Gralinski, J.H., 1996. Children's Beliefs About Intelligence and School Performance.
Journal of Educational Psychology 88 (3), 397–407.
Tempelaar, D.T., Rienties, B., Giesbers, B., Gijselaers, W.H., 2015. The Pivotal Role of Effort Beliefs
in Mediating Implicit Theories of Intelligence and Achievement Goals and Academic Motivations.
Social Psychology of Education 18 (1), 101–120. 10.1007/s11218-014-9281-7.
Zajonc, R.B., 1968. Attitudinal effects of mere exposure. Journal of Personality and Social Psychology
9 (2), 1–27. 10.1037/h0025848.
