807539

research-article2018
EER0010.1177/1474904118807539European Educational Research JournalFroehlich et al.

Original Research Paper

European Educational Research Journal

Stereotype threat-effects for

1­–25
© The Author(s) 2018
Article reuse guidelines:
Turkish-origin migrants in sagepub.com/journals-permissions
DOI: 10.1177/1474904118807539
https://doi.org/10.1177/1474904118807539
Germany: Taking stock of journals.sagepub.com/home/eer

cumulative research evidence

Laura Froehlich
FernUniversität in Hagen, Germany

Sog Yee Mok

University of Zurich, Switzerland

Sarah E. Martiny
UiT, The Arctic University of Norway, Norway

Kay Deaux
New York University, USA

Abstract
Turkish-origin migrants on average show lower academic performance than Germans. This
achievement gap cannot be fully explained by socio-economic differences between the groups.
Negative competence stereotypes about Turkish-origin students predict the causal attributions
that German preservice teachers make for migrants’ academic underperformance. Specifically,
the more strongly preservice teachers endorse negative competence stereotypes, the more likely
they are to attribute academic underperformance of Turkish-origin migrants to the migrants
themselves and less to the educational system. Stereotype threat theory posits that the activation
of stereotypes in test situations can reduce the performance of members of the negatively
stereotyped group. Based on this theory, we propose that negative stereotypes provide a
social-psychological explanation for the academic underperformance of Turkish-origin migrants
compared to Germans. A series of six experiments conducted within a research project funded by
the German Ministry of Education and Research investigated stereotype threat effects for Turkish-
origin migrants. Two new moderator variables were identified: implicit theory of intelligence and

Corresponding author:
Laura Froehlich, Department of Social Psychology, Faculty of Psychology, FernUniversität in Hagen, Universitätsstr. 33,
Hagen, D - 58097, Germany.
Email: [email protected]
2 European Educational Research Journal 00(0)

vertical collectivism. A meta-analysis of the six studies showed a small, non-significant mean effect
for stereotype threat main effects, but a significant medium-sized mean effect for moderated
stereotype threat effects. Limitations and practical implications of stereotype threat effects in
educational settings are discussed.

Keywords
Stereotype threat, Turkish-origin migrants, stereotypes, education, social identity

Children and adolescents living in Germany have heterogeneous ethnic backgrounds. For example,
36% of elementary-school students have either immigrated to Germany themselves or have parents
or grandparents who have immigrated (Statistisches Bundesamt, 2016). Students’ ethnic back-
ground is one factor that predicts their educational performance: on average, German students
show higher performance than students from other ethnic groups. In other words, migrant students
are disadvantaged in the German educational system compared to their native peers (e.g. Klieme
et al., 2010; Prenzel, 2013; Rauch et al., 2016). Migrants have shown lower performance in various
academic domains (e.g. mathematics, science, and reading literacy) across a variety of age groups
(e.g. elementary- and secondary-school students). Research has identified several variables related
to students’ economic and socio-cultural context that explain the performance differences between
migrant and non-migrant students. If, for example, students’ socio-economic status, their parents’
education, and the language spoken within the family are statistically controlled for, the perfor-
mance differences are considerably reduced or even non-significant (e.g. Mok et al., 2016; Prenzel,
2013; Rauch et al., 2016; Schwippert, Wendt, and Tarelli, 2012; Stanat, Rauch, and Segeritz, 2010;
Wendt, Schwippert, and Stubbe, 2016). However, the picture is more complex, as the patterns of
educational disadvantage are not uniform for all ethnic groups of migrant students. Specifically, for
students of Turkish origin the performance differences compared to Germans remain significant
even when economic and socio-cultural context variables are statistically controlled for (e.g.
Autorengruppe Bildungsberichterstattung, 2016). This means that in contrast to other migrant
groups in Germany, some of the causes for the underperformance of Turkish-origin students are
still unaccounted for.
Turkish-origin migrants are the largest migrant group in Germany (2.80 million persons in
2016) and constitute 15.1% of all migrants living in Germany (Statistisches Bundesamt, 2017).
Almost half of them (47%) were actual migrants to Germany, many of whom arrived in Germany
as so-called “guest workers” mainly to be employed in the industrial sector during the 1950s to
1970s (Statistisches Bundesamt, 2017). The remaining 53% were born in Germany as second-
generation (i.e. parents have migrated) or third-generation migrants (i.e. grandparents have
migrated). Turkish-origin migrants are on average younger than the German population without
migration background: whereas the mean age of Germans without migration background is 46.9
years (20.4% are younger than 26 years), Turkish-origin migrants are on average 33.2 years old
(38.9% are younger than 26 years; Statistisches Bundesamt, 2017). Because of this relatively
young demographic, a substantial proportion of Turkish-origin migrants are likely to be enrolled in
educational programs. Therefore, young Turkish-origin migrants have a high potential to be inte-
grated into German society via the route of education. Education is a basis for integration in terms
of further participation in central societal domains (e.g. economy and the labor market, e.g. De
Paola and Brunello, 2016; Die Beauftragte der Bundesregierung für Migration, Flüchtlinge und
Integration, 2016). However, Turkish-origin migrants have more difficulties in performing well in
Froehlich et al. 3

Germany’s academic system than do Germans without migration background or other migrant
groups (e.g. Autorengruppe Bildungsberichterstattung, 2016), which can act as a barrier to their
societal integration.
In order to overcome the inequalities related to ethnicity in the German educational system,
research needs to identify further variables that explain the remaining performance differences
between Turkish-origin and German students. In the present work, we focus on a psychological
variable, namely stereotypes, that is, how Turkish-origin migrants are viewed in German society.
Stereotypes are important in this context because research has shown that stereotypes not only
reflect the general social climate within a society, but can also influence group members’ academic
performance (e.g. Steele and Aronson, 1995). Academic performance is in turn related to migrants’
societal and economic integration (e.g. Länderoffene Arbeitsgruppe “Indikatorenentwicklung und
Monitoring”, 2017). Therefore, negative stereotypes can create a vicious cycle perpetuating the
unfavorable situation for Turkish-origin migrants in Germany in that negative stereotypes predict
lower academic performance and thus lower integration, which is in turn related to a more negative
view of Turkish-origin migrants by the German majority.
In this manuscript, we will first focus on the socio-cultural climate for Turkish-origin migrants
in Germany by reviewing social-psychological research on stereotypes about this group in the
educational domain. We will then summarize how stereotypes can reduce the academic perfor-
mance of negatively stereotyped group members – a phenomenon called stereotype threat (e.g.
Steele, 1997) and review studies investigating stereotype threat effects for migrants in the European
context. In the next section of the manuscript, we review and meta-analyze evidence showing that
stereotype threat can contribute to an explanation of the academic underperformance of Turkish-
origin students in the German educational system. The review and meta-analysis is based on six
empirical studies that were conducted within a research project funded by the German Federal
Ministry of Education and Research (BMBF). In this project, we also identified two new modera-
tor variables of stereotype threat for Turkish-origin migrants: vertical collectivism and implicit
theory of intelligence. These variables help to explain individual differences in Turkish-origin
students’ susceptibility to stereotype threat (i.e. the extent to which their performance is reduced
due to the activation of stereotypes). The current research focuses on Turkish-origin adolescents
(9th/10th grade), because we consider the influence of stereotype threat most relevant during pri-
mary and secondary schooling. Education in these years is the basis for university education, voca-
tional training and participation in the labor market.

Negative stereotypes about Turkish-origin migrants in Germany

We argue that the perception of Turkish-origin migrants in German society (i.e. the socio-cultural
climate) can have important consequences for their educational success and subsequent societal
integration. In order to investigate this socio-cultural climate, we focus on stereotypes, which are
“beliefs and associations that link a whole group of people with certain traits or characteristics”
(Kassin, Fein, and Markus, 2011: 148), and are socially shared in a given context (e.g. the national
context of Germany). Stereotypes are a central social-psychological variable to explain academic
performance differences between Germans and Turkish-origin migrants. Based on the stereotype
content model (Fiske et al., 2002), stereotypes can be described on two basic dimensions: warmth
and competence. The warmth dimension reflects the basic question “What is the outgroup’s intent
toward the ingroup?” and encompasses perceptions of friendliness and likability of a given social
group. Outgroups are rated low on warmth if they are perceived to be in competition with the
ingroup or if they are perceived to be a threat to the ingroup’s resources and values (e.g. Kervyn,
Fiske, and Yzerbyt, 2015). The competence dimension reflects the basic question “How capable is
4 European Educational Research Journal 00(0)

the outgroup of realizing this intent?” and encompasses perceptions of respect and competence.
Outgroups are rated low in competence if they are perceived to have low societal status (e.g.
Kervyn et al., 2015). Stereotypical perceptions of different social groups usually vary and can be
mapped on the two-dimensional stereotype content space, with each group receiving high or low
ratings on either dimension (Cuddy, Fiske, and Glick, 2007). As the socio-cultural climate for dif-
ferent migrant groups usually varies within a given society, stereotypes about migrant groups also
vary along warmth and competence (Lee and Fiske, 2006).
In the context of educational performance, the competence dimension is of particular relevance,
because it can shape majority members’ expectations and attitudes toward migrants, and can also
influence migrants’ academic achievements. Therefore, we summarize the results of a series of
three correlational studies we conducted with German preservice teachers as participants (Froehlich
et al., 2016b) to investigate the stereotypes these preservice teachers have about two different
migrant groups in Germany: Turkish-origin and Italian-origin migrants. Preservice teachers are
students studying to become teachers who already have some teaching experience. Investigating
stereotypes among preservice teachers is particularly relevant, because they will teach students
from different ethnic backgrounds in the future. Preservice teachers’ stereotypes affect how they
view different migrant groups and might shape their behavior toward migrant students. We selected
the groups of Turkish-origin and Italian-origin migrants because, in the German context, Turkish-
origin migrants are stereotyped most negatively, namely as cold and incompetent, whereas Italian-
origin migrants show comparably low educational performance but are not stereotyped as
negatively (Asbrock, 2010; Froehlich et al., 2016b; Schmid, 2014). Results of the three studies
showed that Turkish-origin migrants were stereotyped more negatively than Italian-origin migrants
and Germans (Froehlich et al., 2016b). We further investigated how competence stereotypes about
the two migrant groups predicted the extent to which German preservice teachers perceived the
migrant groups to be responsible for their low academic performance. Results showed that the
more negative the competence stereotypes, the more migrants themselves were seen as responsible
for their low performance. This relationship was stronger for Turkish-origin than for Italian-origin
migrants. Negative competence stereotypes also predicted that participants perceived the educa-
tional system as less responsible for migrants’ low performance. However, this pattern emerged
only for Turkish-origin, but not for Italian-origin migrants. In sum, these results substantiate that
the stereotypes German preservice teachers endorse are important for how they perceive the aca-
demic situation of different migrant groups. When preservice teachers think that migrants are per-
forming poorly because they lack motivation or ability rather than because they are disadvantaged
in the educational system, then they might perceive their own pedagogical opportunities as limited
and might be less motivated to support and foster these migrant students. Thus, they might not
actively contribute to the educational integration of those migrants but rather perpetuate the ethnic
inequalities in the German educational system.
Importantly, results also showed that not all migrant groups with low academic performance were
perceived in the same way. Both Italian- and Turkish-origin migrants are perceived to have low status
in German society, underperform in the educational system compared to Germans (e.g. Baier et al.,
2010; Schmid, 2014), and are stereotyped as incompetent. However, Italian-origin migrants’ intent
toward Germans is perceived to be more positive and friendly and Italian-origin migrants are conse-
quently stereotyped as higher on warmth than are Turkish-origin migrants (Asbrock, 2010; Froehlich
et al., 2016b; Froehlich and Schulte, 2018; Schmid, 2014). These differences in the social perception
of the ethnic groups predicted different perceptions of responsibility for the low performance. In
general, negative competence stereotypes have various consequences in the educational context: stu-
dents belonging to negatively stereotyped ethnic groups can be discriminated against (e.g. Whitley
and Kite, 2016) by their classmates and teachers, and they can also be hindered from fulfilling their
Froehlich et al. 5

academic potential (e.g. Steele and Aronson, 1995). Taken together, the socio-cultural climate for the
largest group of migrants in Germany, namely, Turkish-origin migrants, is quite negative, which
makes it especially relevant to investigate the consequences of stereotypes for this particular ethnic
group. Therefore, we argue that the influence of stereotypes can account for some of the yet unex-
plained variance in performance differences between Germans and Turkish-origin migrants.

Stereotype threat
Stereotype threat theory was originally developed to provide a social-psychological account for the
achievement gap between African-Americans and white people in the United States (Steele and
Aronson, 1995). The theory assumes that group differences in performance are not necessarily
based on differences in underlying abilities. Instead, stereotype threat theory proposes that charac-
teristics of the test situation itself can create performance differences, which are in turn related to
stereotypes about the intellectual ability of the involved groups (for overviews, see Inzlicht and
Schmader, 2012; Spencer, Logel, and Davies, 2016). The activation of stereotypes in test situations
can reduce the performance of negatively stereotyped group members (Steele and Aronson, 1995).
Specifically, when competence-related stereotypes are activated before a standardized test, mem-
bers of negatively stereotyped groups show lower performance than when stereotypes are not
activated (i.e. stereotype threat effect). The stereotypes are assumed to act as “a threat in the air” to
the social identity of stereotyped group members (Steele, 1997). Based on social identity theory
(Tajfel and Turner, 1979), the activation of negative stereotypes threatens group members’ need to
feel good about belonging to the respective group (i.e. a positive social identity associated with the
group). The activation of negative stereotypes can be considered as one type of social identity
threat (Steele, Spencer, and Aronson, 2002). Individuals need to cope with stereotype threat and are
highly motivated to disconfirm the stereotype, which paradoxically leads to reduced test perfor-
mance (e.g. Schmader, Johns, and Forbes, 2008). Coping strategies to restore a positive social
identity include distancing oneself from the negatively stereotyped group (i.e. individual mobility;
Tajfel and Turner, 1979, 1986), as well as distancing oneself from the negatively stereotyped
domain (e.g. refusing to believe that the stereotype is true; Jones, Ruff, and Paretti, 2013). The
mechanisms explaining the resulting performance reduction are complex and can involve increased
anxiety, negative thinking, worrying, and mind-wandering (e.g. Pennington et al., 2016). These
processes are assumed to reduce the capacity of working memory needed to work on the test,
which in turn leads to lower performance (for a meta-analytic overview of mechanisms, see
Pennington et al., 2016). The concept of stereotype threat has been extended to various social
groups as well as different cognitive domains (Inzlicht and Schmader, 2012; Spencer et al., 2016).

Stereotype threat for migrants in Europe

Because Turkish-origin migrants are negatively stereotyped, especially in the domain of compe-
tence, we hypothesized that stereotype threat provides a further explanation for performance dif-
ferences between Turkish-origin migrants and Germans in the German educational context, and
might thus reduce the variance in ethnic performance differences that is still unaccounted for (e.g.
Martiny and Götz, 2011; Schofield and Alexander, 2012; Strasser, 2012). To substantiate this
hypothesis, we consider the European context more generally. We summarize empirical evidence
that stereotype threat effects contribute to the explanation of performance differences between
migrants and non-migrants in different European countries, especially in countries in which the
socio-cultural climate for Turkish-origin migrants is assumed to be comparably negative as in
Germany (i.e. the Netherlands, Belgium, Austria).
6 European Educational Research Journal 00(0)

Two studies with adolescents with Turkish and Moroccan background in Belgium considered the
influence of stereotype threat on educational success and academic performance. Baysu, Phalet, and
Brown (2011) conducted a study with Turkish-Belgian students and investigated perceived ethnic
discrimination as a proxy for stereotype threat in combination with acculturation strategies. Results
showed that students who predominantly identified with their ethnic group (i.e. separation) as well as
those who predominantly identified with the national group (i.e. assimilation) showed higher educa-
tional success when they reported high threat (i.e. frequent discrimination experiences). In contrast,
students who identified with both the ethnic and the national group (i.e. dual identity) showed lower
educational success when they reported high threat. The authors concluded that dual identity (i.e. inte-
gration) is only beneficial in the absence of stereotype threat. However, it should be noted that the
study did not contain an experimental design to activate stereotype threat. Another study conducted in
Belgium by Baysu et al. (2016) investigated stereotype threat effects for a diverse sample of migrant
adolescents predominantly with Turkish and Moroccan backgrounds in ethnically diverse schools.
Stereotype threat was activated by making ethnicity salient before (versus after) a performance test.
Results of the multilevel analysis in this study showed that the influence of stereotype threat on perfor-
mance was mediated by disengagement from the performance task. In sum, these two studies show
that perceived discrimination and stereotype threat can reduce the educational success and test perfor-
mance of migrants from Turkish and Moroccan background in Belgium. We therefore argue that ste-
reotype threat effects can also reduce Turkish-origin migrants’ academic performance in Germany.
There is first evidence for stereotype threat effects for migrants with diverse ethnic backgrounds
in the German-speaking countries of Germany and Austria, which further corroborates our hypoth-
esized stereotype threat effects for Turkish-origin migrants. Weber, Appel, and Kronberger (2015)
conducted two experiments in Austria with migrant students of different origins, but predomi-
nantly from the Balkans, and investigated stereotype threat in combination with ethnic and national
identity strength. Results revealed that higher national identity was related to higher performance
under explicit stereotype threat, whereas ethnic identity was unrelated to performance. In a recent
study with elementary-school migrant students of different origins in Germany, stereotype threat
effects were found in children’s verbal learning: When stereotype threat was activated by making
salient that the children’s first language was non-German, they showed a lower vocabulary increase
compared to a control group. Ethnic identity was again unrelated to learning under stereotype
threat (Sander et al., 2018).
Finally, a meta-analysis by Appel, Weber, and Kronberger (2015) investigated stereotype threat
effects for migrants in Europe and the United States. Overall, results showed a medium-sized ste-
reotype threat effect across 19 independent samples. These studies included Latinos/as in the
United States, different migrant groups from African and Arabic countries in France, migrants
from diverse backgrounds in Austria and the Netherlands, and migrants from Turkey in Germany
(eight studies from Europe, including Studies 1 and 4 from the current meta-analysis). In sum,
studies from various European countries showed that negative stereotypes can decrease the perfor-
mance and learning of migrant students. However, most studies either combined students with very
diverse ethnic backgrounds into one sample or did not include a systematic experimental activation
of stereotypes. As stereotypes about different migrant groups vary considerably within and across
national contexts, it is still an empirical question how strongly the group of Turkish-origin migrants
is affected by stereotype threat in the German context. The current review and meta-analysis will
focus on the specific group of Turkish-origin migrants in Germany and take stock of the empirical
database on stereotype threat currently available for this group and national context. A second aim
of this review is to summarize empirical evidence for two newly identified moderator variables of
stereotype threat effects for Turkish-origin migrants, which can be important as they attenuate or
intensify the experience of stereotype threat for Turkish-origin students.
Froehlich et al. 7

Moderators of stereotype threat effects

Not all individuals experience the performance-reducing consequences of stereotype threat to a
similar extent. Research has identified several moderator variables predicting individuals’ suscep-
tibility to stereotype threat. These moderator variables predict the degree to which individuals
perceive the performance situation and its outcome as important for their social identity and thus
for their self-concept, which in turn determines the extent of the performance-reducing effect of
stereotype threat (i.e. boundary conditions for stereotype threat effects). Previous studies showed
that performance decreases were greater for individuals who were highly identified with the nega-
tively stereotyped group (e.g. Schmader, 2002) and the negatively stereotyped domain (Keller,
2007), who believed the stereotype to be true (Elizaga and Markman, 2008; Schmader, Johns, and
Barquissau, 2004), and who had low self-esteem (Rydell and Boucher, 2010) or an internal locus
of control (Cadinu et al., 2006). Furthermore, the meta-analysis of Appel et al. (2015) revealed that
for migrants in Europe and the United States significant stereotype threat effects emerged in both
published and unpublished studies, for different age groups (although effect sizes were larger for
adults than for adolescents and children), and for all treatments to activate stereotype threat, as well
as for non-verbal and knowledge-based performance tests. When migrant status was determined
according to self-identification with an ethnic group, effect sizes were larger compared to when it
was determined by demographic variables (Appel et al., 2015). Knowledge about the extent to
which members of negatively stereotyped groups are affected by stereotype threat can create path-
ways to the development of interventions designed to alleviate the consequences of stereotype
threat.
However, different target groups of stereotype threat can also be affected by different moderator
variables. Based on the hypothesis that stereotype threat effects occur for Turkish-origin migrants
in Germany, a next step is the identification of further moderators of stereotype threat for this
group. To do so, we aimed at identifying belief/cultural orientation variables that indicate the extent
to which Turkish-origin migrants believe the performance situation is important for their social
identity, which in turn could predict the degree to which performance is reduced for Turkish-origin
migrants under stereotype threat. One such variable is the attitude toward collectivism. Turkish
culture can be characterized as more collectivistic than German culture (e.g. Hofstede, Hofstede,
and Minkov, 2010; Varnum et al., 2010), which means that belonging to social groups and keeping
positive relations with one’s fellow group members is important to individuals socialized in the
Turkish culture. We believe that this collectivistic cultural orientation of Turkish-origin migrants is
important to consider in performance situations because Turkish-origin families have high perfor-
mance expectations to which Turkish-origin students strive to live upto, especially when negative
performance-related stereotypes are activated. Therefore, we investigated two individual-level
variables related to Turkish-origin migrants’ cultural orientation as further moderators of stereo-
type threat effects.

Moderator variable: Vertical collectivism. A variable linking Turkish-origin migrants’ cultural orienta-
tion of collectivism with the perceived importance of the performance situation to their social iden-
tity is vertical collectivism (Mok et al., 2017). The Turkish culture endorses values of interpersonal
connectedness and social belonging with a strong emphasis on relatedness (e.g. Güngör et al., 2014).
This relatedness and behavioral adaptation to group norms is reflected by the concept of vertical
collectivism (Singelis et al., 1995). Individuals high in vertical collectivism adhere to group norms
and prioritize ingroup members’ interests over their own interest (Komarraju and Cokley, 2008).
Turkish-origin families expect high achievement motivation and performance from their children
(e.g. Phalet and Claeys, 1993; Verkuyten, Thijs, and Canatan, 2001). Thus, Turkish-origin students
8 European Educational Research Journal 00(0)

highly endorsing vertical collectivism show high achievement motivation because this is expected
by their ingroup members.
One mechanism leading to stereotype threat effects is worrying about one’s performance, which
occupies the capacity of working memory needed to complete the task (for a review, see Pennington
et al., 2016). This worry component of stereotype threat is likely to be exacerbated for individuals
high in vertical collectivism, as they might worry that confirming the stereotype by underperform-
ing not only reflects on their individual ability, but also on the ability of their ingroup as a whole.
Further, they might be concerned that the negative stereotype implies that they cannot fulfill their
ingroup members’ high performance expectations. Consequently, we proposed that when negative
competence-related stereotypes are activated, Turkish-origin migrants should perform worse, the
more strongly they endorse vertical collectivism (Mok et al., 2017).

Moderator variable: Implicit theory of intelligence. Another variable related to Turkish-origin migrants’
cultural orientation and the perceived importance of the performance situation for their social iden-
tity is one’s implicit theory of intelligence. In general, individuals vary in their endorsement of
basic conceptions of the nature of intelligence. Some people believe that they have a fixed amount
of intelligence that cannot be changed (i.e. an entity theory), whereas others believe that they can
increase their intelligence by effort and practice (i.e. an incremental theory; e.g. Dweck, 1999).
These lay conceptions of the fixedness or malleability of intelligence have consequences for per-
formance on difficult tasks or after failure. When entity theorists encounter difficulties, their per-
formance is reduced; in contrast, incremental theorists do not experience performance losses in the
face of difficulties (e.g. Grant and Dweck, 2001, 2003; Levy, Stroessner, and Dweck, 1998). One
mechanism to explain these different reactions to difficulties and related performance differences
is that entity theorists endorse performance goals and thus are constantly seeking to verify whether
their (fixed) level of intelligence is sufficient to perform successfully on a given task. Difficulties
or failure are perceived as indications of lack of ability, which they believe is unchangeable and
therefore an undesired outcome. Incremental theorists, however, endorse learning goals and react
to difficulties or failure with increased effort and motivation, because they believe that they can
thereby increase their ability (e.g. Dweck, Chiu, and Hong, 1995).
We proposed theory of intelligence as a moderator of stereotype threat effects for Turkish-origin
migrants (Froehlich et al., 2016a). Competence stereotypes imply that some groups have higher
ability than other groups, thus linking social identity to implicit theories of intelligence. Individuals
who belong to a group that is negatively stereotyped in the competence domain might infer that
because of their group membership their ability is not sufficient to perform well in the respective
domain. This should be perceived as more threatening to their social identity by entity theorists
than by incremental theorists. In other words, stereotype threat should be exacerbated for entity
theorists. The collectivistic cultural orientation of Turkish-origin migrants should make them sen-
sitive to the relationship between negative stereotypes about their group’s ability and the perfor-
mance situation, as their relatedness and subjective importance of group membership is highly
important to their identity. Consequently, we proposed that endorsement of an entity theory of
intelligence would moderate Turkish-origin migrants’ susceptibility to stereotype threat: when
negative competence-related stereotypes are activated, Turkish-origin migrants will perform worse
the more strongly they endorse an entity theory of intelligence (Froehlich et al., 2016a).

The research project

Based on the social-psychological evidence for the prevalence of negative competence-related
stereotypes about Turkish-origin migrants in Germany as well as evidence for stereotype threat
Froehlich et al. 9

effects for various migrant groups in Europe, we investigated whether stereotype threat plays a role
in the academic underperformance of Turkish-origin compared to German students. In a research
project funded by the German Federal Ministry of Education and Research (BMBF), we conducted
a series of six experimental studies in classrooms in order to investigate stereotype threat effects
for Turkish-origin migrants (project duration 2012–2015; grant number: 01JC1104).1 In the current
paper, we take stock of the combined empirical findings on stereotype threat for Turkish-origin
adolescents in Germany. Thus, we evaluate how strongly the cumulative evidence supports the
following hypotheses: (I) stereotype threat reduces Turkish-origin migrants’ test performance; and
(II) vertical collectivism and implicit theory of intelligence moderate the extent to which Turkish-
origin migrants are susceptible to stereotype threat effects. In a first step, we summarize the empiri-
cal evidence from the separate studies, and in a second step, we conduct a meta-analysis to
investigate how well the mean effect sizes of the studies support Hypotheses I and II.

Overview of the research design

We conducted six experiments with 9th- and 10th-grade high-school students on different educa-
tional tracks in multicultural urban regions of Germany. Each of the experiments had a two-facto-
rial between-participants design: the first factor was participants’ ethnicity (German vs. Turkish
migration background). In Study 1, ethnicity was coded according to students’ self-identification
with ethnic groups, whereas in all other studies demographic information (students’ and parents’
country of birth) was used. The second factor was experimental condition (stereotype activation vs.
no stereotype activation; random assignment of students to conditions in classrooms). Students
participated in their classrooms in the presence of an experimenter and a teacher. Prior to working
on a standardized test assessing either mathematical or verbal performance, stereotype activation
was manipulated in the test instructions. In each study, one of three kinds of experimental manipu-
lations for stereotype activation was used. In the diagnosticity manipulation (e.g. Steele and
Aronson, 1995), students either read that the upcoming test was diagnostic of their ability (stereo-
type activation condition) or that it was a non-diagnostic practice test (control condition). In the
fairness manipulation (e.g. Spencer, Steele, and Quinn, 1999) they read that the test had previously
produced group differences in performance between Germans and Turkish-origin migrants (stereo-
type activation condition) or that there were no group differences in performance (control condi-
tion). In the identity salience manipulation (Shih, Pittinsky, and Trahan, 2006), either the negatively
stereotyped identity (i.e. ethnic group membership) was made salient (stereotype activation condi-
tion) or a positive identity (i.e. attendance at private school) was made salient (control condition).
Students summarized the instructions (including the experimental manipulation) in writing as a
manipulation check (except when the identity salience manipulation was used because identity was
made salient by answering identity-related questions) and subsequently worked on the perfor-
mance test. Finally, they completed a questionnaire assessing psychological variables and demo-
graphics, and were fully debriefed. Moderator variables were assessed in separate sessions some
time prior to the main sessions. Four of the six studies are published in peer-reviewed journals and
the remaining two studies are unpublished. Table 1 presents an overview of the characteristics of
the included studies (chronologically ordered according to time of study conduction).

Summary of study results

Hypothesis I: Stereotype threat effects for Turkish-origin migrants
Hypothesis I was tested in three experiments (numbered according to Table 1). Study 1 (Martiny
et al., 2014a) investigated stereotype threat effects for Turkish-origin migrants in the math domain
10 European Educational Research Journal 00(0)

Table 1. Overview of characteristics of included studies.

Study Reference/ Status Sample Size Manipulation Domain Grade level Moderator
(IQB no.) Authors
1 (1b) Martiny Published N = 148 Fairness Math 9th No
et al., (75 Turkish-origin
2014a migrants,
73 Germans)
2 (2a/3a) Mok et al., Unpublished N = 200 Diagnosticity Math 9th, 10th No
2014 (79 Turkish-origin
migrants, 121
Germans)
3 (n.a.) Martiny Unpublished N = 179 Fairness Verbal 9th, 10th No
et al., (92 Turkish-origin
2014b migrants, 87
Germans)
4 (1a) Froehlich Published N = 174 Diagnosticity Verbal 9th Implicit
et al., (127 Turkish- theory of
2016a origin migrants, intelligence
47 Germans)
5 (1cDE) Froehlich Published N = 186 Diagnosticity Verbal 9th Implicit
et al., (62 Turkish-origin theory of
2016a migrants, 124 intelligence
Germans)
6 (n.a.) Mok et al., Published N = 94 Identity Verbal 9th, 10th Vertical
2017 (94 Turkish-origin salience collectivism
migrants, no
Germans)

using a fairness manipulation. The sample consisted of 148 students (75 Turkish-origin migrants,
73 Germans). Age ranged from 13 to 17 years (M = 14.52; SD = 0.69), 112 students were female.
Results showed an interaction of ethnicity and stereotype activation to predict math performance;
F(1,144) = 4.20, p < .040 η²p = .03. In the control condition, the test performance of Turkish-
origin migrants and Germans did not differ significantly. However, in the stereotype activation
condition, Turkish-origin migrants showed lower performance than Germans; MDiff = 2.12, 95%
CI = [0.89, 3.36], p < .001, d = 0.86. Furthermore, Turkish-origin migrants in the stereotype
activation condition also showed lower performance than Turkish-origin migrants in the control
condition; MDiff = 1.19, 95% CI = [-0.13, 2.40], p = .053, d = 0.46. These results reflect a classi-
cal stereotype threat effect.
Furthermore, Study 1 investigated two strategies to cope with stereotype threat: increased iden-
tification with the ethnic group and decreased belief that the shared stereotype is true. The rationale
behind this was that individuals can apply different strategies to cope with stereotype threat and
restore a negative social identity (e.g. Tajfel and Turner, 1986). Stereotype threat implies a negative
association between the stereotyped group and the performance domain, which threatens negatively
stereotyped group members’ positive social identity. As a reaction to this threat, individuals can
increase their identification with the group in order to reassure and protect their social identity (e.g.
Jetten et al., 2016; Tajfel and Turner, 1986). Alternatively, they can refute the socially shared meta-
stereotype (i.e. reject the negative association of the group and the domain; Jones et al., 2013). We
investigated whether Turkish-origin migrants used these strategies to cope with stereotype threat.
Froehlich et al. 11

Results showed a non-significant trend in the predicted direction: interaction between ethnicity and
experimental condition on group identification; F(1,143) = 3.05, p = .083, η²p = .02. The activa-
tion of stereotypes did not affect Germans’ group identification, but Turkish-origin migrants identi-
fied more strongly with their ethnic group after stereotype activation compared to the control
condition; MDiff = 0.86, 95% CI = [0.14, 1.59], p < .001, d = 0.55. Furthermore, there was a two-
way interaction on belief in the meta-stereotype; F(1,138) = 4.11, p = .045, η²p = .03. Germans’
perception of how most people would evaluate their groups’ math ability did not differ between
experimental conditions. However, when stereotypes were activated, Turkish-origin migrants per-
ceived that others evaluate Germans’ math ability as lower than when no stereotypes were activated;
MDiff = 0.93, 95% CI = [0.21, 1.65], p = .012, d = 0.55. In sum, Study 1 showed a significant
stereotype threat effect for Turkish-origin migrants in the math domain, as well as the use of two
different coping strategies to manage their social identity under threat.
Study 2 (Mok et al., 2014) investigated stereotype threat in the math domain with a diagnosticity
manipulation. The sample consisted of 200 participants (79 Turkish-origin migrants, 121 Germans).
Age ranged from 14 to 17 years (M = 14.76 SD = 0.81), 99 participants were female. Results
showed a main effect of ethnicity on test performance (i.e. Turkish-origin migrants showed lower
performance than Germans irrespective of stereotype activation; F(1, 196) = 7.23, p = .008, η²p
=.04, but no stereotype threat effect for Turkish-origin migrants. The internal consistencies of the
employed performance measures were in part insufficient, which is a severe limitation of the study.
Study 3 (Martiny et al., 2014b) investigated stereotype threat effects in the verbal domain using
the fairness manipulation. The sample consisted of 179 participants (92 Turkish-origin migrants,
87 Germans). Age ranged from 14 to 17 years (M = 15.12, SD = .57), 88 participants were female.
Results again showed no significant interaction of ethnicity and experimental condition on perfor-
mance (i.e. no stereotype threat effect). Similar to Study 2, the low internal consistency of the
performance measure restricted the reliability of results.
In sum, one study (Martiny et al., 2014a) showed that stereotype activation affected Turkish-
origin students’ math performance and use of strategies to cope with stereotype threat, whereas two
further studies were not able to show stereotype threat effects. However, the results of these studies
should be interpreted with caution due to psychometric limitations.

Hypothesis II: Moderated stereotype threat effects

Three additional experiments explored new moderator variables of stereotype threat effects for
Turkish-origin migrants in Germany. Studies 4 and 5 (Froehlich et al., 2016a) investigated endorse-
ment of an entity theory of intelligence (i.e. the lay belief that intelligence is fixed) as a moderator
of stereotype activation effects. In addition to stereotype threat effects for Turkish-origin migrants
(i.e. performance decreases for the negatively stereotyped group), these studies also investigated
stereotype lift effects for Germans (i.e. performance increases for the favorably stereotyped group;
e.g. Walton and Cohen, 2003). Study 4 consisted of 174 participants (127 Turkish-origin migrants,
47 Germans). Age ranged from 14 to 21 years (M = 14.69, SD = 0.82), 69 participants were
female. Study 5 consisted of 186 participants (62 Turkish-origin migrants, 124 Germans). Age
ranged from 14 to 18 years (M = 15.21, SD = 0.62), 84 participants were female. Both studies
investigated stereotype activation effects in the verbal domain using a diagnosticity manipulation.
Results of Study 4 showed that entity theory endorsement moderated stereotype threat effects for
Turkish-origin migrants; three-way interaction of entity theory, ethnicity, and condition: b = -4.02,
95% CI [-7.12, -0.93], β = -.61, t(145) = -2.57, SE = 1.56, p = .001. Only when stereotypes were
activated did higher entity theory endorsement predict lower performance of Turkish-origin
migrants; simple slope: b = -0.99, 95% CI [-1.96, -0.02], t(145) = -2.03, SE = 0.49, p = .045.
12 European Educational Research Journal 00(0)

Study 5 replicated the moderated stereotype threat effect, but in addition showed that entity theory
endorsement positively predicted Germans’ performance; three-way interaction: b = -4.28, 95%
CI [-7.80, -0.76], β = -.29, SE = 1.78, t(149) = -2.40, p = .018. When stereotypes were activated,
entity theory endorsement predicted lower performance of Turkish-origin migrants; simple slope:
b = -2.11, 95% CI [-4.24, 0.02], t(149) = -1.96, p = .052; while being associated with higher
performance of Germans (i.e. stereotype lift effect; simple slope: b = 1.89, 95% CI [0.46, 3.31],
t(149) = 2.61, SE = 0.72, p = .010). Thus, the belief in a fixed view of intelligence predicted the
extent to which both members of the negatively stereotyped group and members of the favorably
stereotyped group are susceptible to stereotype activation effects.
Lastly, Study 6 (Mok et al., 2017) investigated vertical collectivism as a moderator of stereotype
threat effects for Turkish-origin migrants. This study was conducted with 94 Turkish-origin stu-
dents from two private schools that offered Turkish language classes. Age ranged from 13 to 18
years (M = 15.41, SD = 1.03), 48 participants were female. Study 6 investigated performance in
the verbal domain and used the identity salience manipulation. Results showed that vertical col-
lectivism moderated the effects of stereotype activation on Turkish-origin migrants’ performance;
two-way interaction: b = -0.66, 95% CI [-1.24, -0.08], t(89) = -2.30, SE = 0.29, p = .024. Only
when negative stereotypes were activated, higher vertical collectivism predicted lower perfor-
mance; simple slope: b = 0.69, 95% CI [-1.13, -0.25], t(89) = -3.20, SE = 0.22, p = .002. In
addition, Study 6 investigated individual mobility motivation as a further strategy to cope with
stereotype threat (Tajfel and Turner, 1979, 1986). Individual mobility describes the tendency to
leave the negatively stereotyped group in order to join a higher-status outgroup as a means to
restore one’s positive social identity (Ellemers, Knippenberg, and Wilke, 1990). In line with verti-
cal collectivism, joining a higher-status (e.g. better performing) outgroup would be a strategy to
fulfill the high achievement motivation expected by one’s ingroup members. Results supported this
reasoning. When negative stereotypes were activated (but not when positive stereotypes were acti-
vated), Turkish-origin migrants highly endorsing vertical collectivism were motivated to employ
individual mobility strategies (i.e. the desire to represent the outgroup of Germans in a future per-
formance situation and the preference for a higher ratio of Germans than Turkish-origin migrants
in a team; two-way interaction: b = 0.81, 95% CI [0.23, 1.39], t(90) = 2.81, SE = 0.29, p = .006;
simple slope: b = 0.68, 95% CI [0.24, 1.12], t(90) = 3.06, SE = 0.22, p = .003). Thus, Study 6
showed that the cultural orientation of vertical collectivism – along with a high achievement moti-
vation – determined Turkish-origin migrants’ susceptibility to stereotype threat effects, as well as
individual mobility as an identity management strategy after the activation of negative compe-
tence-related stereotypes. Because Appel et al. (2015) showed that stereotype threat effects for
migrants were stronger when their self-identification was used compared to their demographic
identification, we repeated the analyses of Studies 2–6 with self-identification measures instead of
demographic measures and the main results remained statistically significant.
In sum, the current research project has identified two new moderator variables that explain the
extent to which Turkish-origin migrants are affected by stereotype threat effects. A belief in the
fixedness of intelligence and endorsement of vertical collectivism seem to be especially unfavora-
ble for Turkish-origin students’ academic performance in the face of negative stereotypes about
their ethnic group’s competence.

Meta-analysis
Aim of the meta-analysis and inclusion criteria
The aim of the meta-analysis was to investigate whether there is cumulative evidence for (moder-
ated) stereotype threat-effects for Turkish-origin high-school students in Germany. Therefore, the
Froehlich et al. 13

inclusion criteria for studies were the following: (1) the studies included an experimental manipu-
lation of stereotype threat, (2) a comparison of Germans with Turkish-origin migrants, (3) a sample
of high-school students, as well as (4) they were written/published in English or German language.
To investigate whether there were any further studies to be considered for the meta-analysis aside
from the studies conducted in the research project described above, we conducted a literature
search in March 2018 using all EBSCO Host databases. We used the key words “stereotype threat,”
“social identity threat,” and the German translation “Stereotypenbedrohung” in combination with
at least one of the terms “Turk*,” “immigra*,” “migra*,” as well as “Migrationshintergrund”
(asterisk as a placeholder for different word endings). In addition, we requested unpublished stud-
ies via emailing lists. No further studies fitted the inclusion criteria.

Sample and computation of effect sizes

The total sample of the six experimental studies consisted of 981 participants (529 Turkish-origin
migrants and 452 Germans). The overall mean effect for the stereotype threat effects was analyzed
in all six studies, whereas the overall mean effect for the moderated stereotype threat effects was
analyzed in a subsample of three studies that included moderator variables in the study design.
To compute the effect sizes of the studies, we used the Comprehensive Meta-Analysis software
(Borenstein et al., 2009). We used a random effects model because the participants in the studies
attended different school types and were from different federal states in Germany, thus making it
likely that the effect sizes varied (Borenstein et al., 2009). To investigate Hypothesis I, we com-
puted the overall mean effect for the stereotype threat effects by using the standardized mean dif-
ference (Cohen’s d; Cohen, 1988) as an effect size measure for performance difference between
Turkish-origin migrants in the stereotype activation condition and Turkish-origin migrants in the
control condition (Borenstein et al., 2009). To investigate Hypothesis II, we computed the overall
mean effect for the moderated stereotype threat effects. In doing so, we followed two steps: first,
we entered the t-value of the significant simple slope of each moderator variable (i.e. entity theory
of intelligence and vertical collectivism) for Turkish-origins migrants in the stereotype activation
condition and the sample size of Turkish-origin students of each study in order to compute the
coefficient r, which indicates a relationship between two variables such as the moderator variable
and the stereotype threat manipulation. Second, we transformed the coefficient r of each study into
Cohen’s d (i.e. standardized mean difference) to make the effect sizes comparable.

Meta-analytic results
Hypothesis I: Overall mean effect for stereotype threat effects. The overall mean effect for all six stud-
ies was small (d = .098, SE = 0.087, 95% CI [-0.073, 0.269], p = .263). However, in Study 1
(Martiny et al., 2014a), we found a medium-sized stereotype threat effect for Turkish-origin
migrants in the mathematical domain of d = 0.461 (SE = 0.234, 95% CI [0.002, 0.920], p = .049)
according to the classification of Cohen (1988). All other studies yielded very small to small effect
sizes (see Figure 1). The heterogeneity test was not significant (Q(5) = 3.646, p = .601), which
indicated that the effect sizes in our sample did not vary significantly between studies. However,
this result should be interpreted with caution because the number of studies included in the meta-
analysis is rather small: statistical non-significance could be due to low test power (Higgins et al.,
2003).

Hypothesis II: Overall mean effect for moderated stereotype threat effects. To calculate the overall
mean effect for the studies investigating moderated stereotype threat effects, we included Study 4
14 European Educational Research Journal 00(0)

Figure 1. Forest plot of mean effect sizes of all studies.

Figure 2. Forest plot of mean effect sizes of studies with moderated stereotype threat effects.

and Study 5 (Froehlich et al., 2016) as well as Study 6 (Mok et al., 2017). We found an overall
mean effect of medium size (d = 0.492, SE = 0.125, CI [0.737, 0.247], p < .001), and the hetero-
geneity test for this subsample was also non-significant (Q(2) = 1.126, p = .57). All moderated
stereotype threat effect studies showed small to medium effect sizes (d ranging from .365 to .667;
see Figure 2).

Discussion
Turkish-origin migrants underperform compared to Germans in the educational system, and
previous research showed that these performance differences are not fully explained by variables
related to students’ economic and socio-cultural background (e.g. Autorengruppe
Bildungsberichterstattung, 2016). Based on social-psychological research showing widespread
negative competence stereotypes about Turkish-origin migrants in Germany (Froehlich et al.,
2016), we investigated whether stereotype threat (e.g. Inzlicht and Schmader, 2012) contributes
to the explanation of performance differences between Germans and Turkish-origin migrants,
focusing on six experimental studies done with 9th/10th-grade high school students. Here we
take stock of the cumulative research evidence when a common paradigm is used. We
Froehlich et al. 15

investigated whether the study results showed stereotype threat effects (i.e. lower performance
of Turkish-origin migrants after stereotype activation compared to a control condition; Hypothesis
I), and whether Turkish-origin migrants’ susceptibility to stereotype threat is explained by two
newly proposed moderator variables related to the cultural orientation of collectivism (i.e. the-
ory of intelligence and vertical collectivism; Hypothesis II).

Evidence for Hypothesis I: Stereotype threat effects

We found strong empirical support for stereotype threat effects for negatively stereotyped migrants
in Europe and first evidence pointing in the direction of stereotype threat effects for Turkish-origin
migrants in Germany, thus supporting Hypothesis I. The summary of empirical evidence revealed
that the activation of negative stereotypes in test situations can reduce Turkish-origin migrants’
performance, but this evidence was somewhat mixed. A meta-analysis of the six project studies
showed that the overall effect size was for the most part directionally consistent, but small and non-
significant. Further, effect sizes did not significantly vary between studies, which might be
explained by the low number of studies included in the meta-analysis and the corresponding low
test power (Higgins et al., 2003).
Meta-analyses investigating stereotype threat effects for migrants and minorities in different
countries and cognitive domains generally found small to medium effect sizes (e.g. Appel et al.,
2015; Nadler and Clark, 2011; Nguyen and Ryan, 2008; Walton and Spencer, 2009). Nguyen and
Ryan (2008), for example, investigated stereotype threat effects for women and minorities in a
sample of 116 studies and showed an overall effect size of d = .29, which was slightly higher when
only the studies with an ethnicity-based stereotype were considered (d = .32). Appel et al. (2015)
investigated stereotype threat effects for various migrant groups in Europe and the United States in
a sample of 19 studies (also including Studies 1 and 4 from the current review) and yielded a
medium effect size of d = .63 in support of stereotype threat theory. Compared to these small to
medium effect sizes, the effect found in the current review (d = .098) is smaller (and non-signifi-
cant). What might account for this discrepancy in findings? On the one hand, the number of studies
in the present meta-analysis was considerably smaller than in the other meta-analyses, because the
current analysis is limited to the specific case of Turkish-origin migrants in Germany. Within such
a small set of studies, it is more difficult to successfully detect an overall stereotype threat effect.
On the other hand, studies have to be designed in very specific ways in order to detect stereotype
threat effects. For example, the performance test needs to be reliable and difficult, and test perfor-
mance needs to depend on working-memory capacity (for a review, see Schmader and Beilock,
2012). In two of the six studies included in the current review (Studies 2 and 3) we were apparently
not successful in creating a performance measure that was reliable enough to produce stereotype
threat effects, reflected by low internal consistencies. In such a small set of studies as in the current
meta-analysis, those studies that did not successfully establish stereotype threat effects have a high
impact on the overall mean effect, which might also explain why the effect did not reach statistical
significance. A further explanation is that the studies in the current meta-analysis were conducted
in school classes (i.e. real-world environments of learning and performance), whereas many stud-
ies included in the previously conducted meta-analyses were conducted in laboratory settings.
Outside of the controlled laboratory environment, multiple variables could have added error vari-
ance to the effect of stereotype threat on performance.
However, even a small-sized stereotype threat effect can have meaningful consequences for
the German educational system. Of the over 2.85 million Turkish-origin migrants living in
Germany, 900,000 are below 20 years old and are therefore likely to be in secondary education
(Bundeszentrale für politische Bildung, 2016). Some researchers have argued that every
16 European Educational Research Journal 00(0)

performance situation can invoke negative competence-related stereotypes and stereotype threat
(e.g. Aronson and Dee, 2012). We therefore argue that stereotype threat can hinder Turkish-origin
migrants from fulfilling their full academic potential in countless everyday performance and test-
ing situations. Further, it is possible that stereotype threat affects individuals repeatedly over time,
thus exacerbating its effect over the course of their education. The experience of chronic stereo-
type threat can lead to severe consequences beyond performance. For example, Turkish-origin
migrants might steadily decrease their identification with the academic domain and as a conse-
quence pursue less ambitious educational goals (e.g. Woodcock et al., 2012), and feel a lower
sense of belonging to the academic domain (Walton and Cohen, 2007), both of which increase the
likelihood of withdrawal from education. Stereotype threat can also reduce social approach moti-
vation (Martiny and Nikitin, in press) and might thereby negatively influence friendships that
Turkish-origin migrants have with their classmates. The ongoing threat and pressure could even
decrease their well-being and health (e.g. Blascovich et al., 2001; Guendelman, Cheryan, and
Monin, 2011; Inzlicht and Kang, 2010). For these reasons, we argue that the overall stereotype
threat effect for Turkish-origin migrants found in our studies is small, but nonetheless likely to
have an impact in the German educational context.
Further, results showed that in addition to performance reduction, stereotype threat can lead to
further consequences when members of the negatively stereotyped group try to restore their posi-
tive social identity (Tajfel and Turner, 1986): Turkish-origin students showed different strategies
for coping with stereotype threat. They identified more strongly with their ethnic group and
believed less in the meta-stereotype about ethnic performance differences (Study 1), but also
showed preferences for individual mobility (i.e. being motivated to temporarily leave the nega-
tively stereotyped group in order to join the higher-status group of Germans in a performance situ-
ation; Study 6). Thus, depending on the context, Turkish-origin students might react differently to
stereotype threat: they might increase their identification with their ethnic group and try to distance
themselves (and their group) from the negative stereotype. However, in performance situations in
which individual mobility is possible, they might prefer to leave their ethnic group in order to join
the higher-status outgroup of Germans. Further research should investigate the conditions under
which both coping strategies are used by Turkish-origin migrants under stereotype threat.

Hypothesis II: Moderators of stereotype threat

Although the current meta-analysis did not show heterogeneity of effect sizes, previous research
has emphasized that individuals vary in their susceptibility to stereotype threat (e.g. Nguyen and
Ryan, 2008). Therefore, and based on theoretical considerations, we proposed two new moderator
variables to explain the extent of Turkish-origin migrants’ performance reduction after the activa-
tion of negative stereotypes. We found evidence in three studies (Studies 4–6) that the endorsement
of an entity theory of intelligence and vertical collectivism predict a more pronounced performance
decrease under stereotype threat for Turkish-origin migrants. These moderator variables explained
the extent of stereotype threat effects with a mean effect size of d = .49, which can be considered
a medium effect (Cohen, 1988). Consequently, we were able to shed further light on who is affected
by stereotype threat and to what extent. Stereotype threat does not affect all members of the nega-
tively stereotyped group equally, and some individuals have a demonstrably stronger reaction to
the threat to their social identity than do others (e.g. Aronson and Dee, 2012). Both the belief that
one’s intelligence is fixed and the cultural orientation of vertical collectivism intensify stereotype
threat effects for Turkish-origin migrants in Germany. Both variables predict the extent to which
Turkish-origin migrants see the activation of a negative stereotype in performance situations as a
threat to their social identity (Steele et al., 2002). Because negative stereotypes imply that
Froehlich et al. 17

Turkish-origin migrants’ ability is lower than that of Germans, those Turkish-origin migrants who
think that their ability is fixed and unchangeable might worry that they will have difficulties in the
performance situation. In their eyes this would indicate that their ability is not high enough to be
successful on the task – an outcome that entity theorists experience as failure and as an indication
that they are not smart enough, which they think they cannot change (e.g. Dweck, 2006).
Furthermore, as Turkish-origin migrants value relatedness and connection to their ingroup mem-
bers, they internalize the high performance expectations by their family members (Verkuyten et al.,
2001). When stereotypes imply that Turkish-origin migrants have low ability, they might perceive
it as particularly difficult to fulfill their ingroup members’ high performance expectations despite
their own high achievement motivation. Thus, Turkish-origin migrants highly endorsing vertical
collectivism are strongly affected by stereotype threat. In sum, both newly identified moderator
variables predict the degree to which Turkish-origin migrants experience negative stereotypes as a
threat to their social identity.

Limitations
The database reported here, based on six experimental studies with a total of 981 students, is exten-
sive for experimental studies, but is still small for a meta-analysis, and the database needs to be
expanded. Further, the current sample of studies is too small to investigate effect sizes based on
methodological or contextual moderator variables as Appel et al. (2015) did in their meta-analysis
on stereotype threat effects for migrants in general. It would be interesting and relevant to investi-
gate whether effect sizes differ depending on school type (i.e. lower, middle, and higher track),
performance domain (i.e. mathematical or verbal performance), or the type of stereotype activation
manipulation (i.e. blatant or subtle manipulation). These investigations were not possible with the
existing set of studies. Another limitation of the current review is that the studies were conducted
with 9th/10th grade high-school students only. Future studies should investigate whether stereo-
type threat effects for Turkish-origin migrants also occur for students of different age groups to
determine the age group in which interventions against stereotype threat could be most effective.
Other research has shown that stereotype threat can have negative consequences over and above
test performance. For example, it can increase negative emotions (e.g. Adams et al., 2006), decrease
self-confidence (e.g. Koch, Müller, and Sieverding, 2008; Muzzatti and Agnoli, 2007) as well as
identification with the domain (e.g. Woodcock et al., 2012) or the group (e.g. Pronin, Steele, and
Ross, 2004). It can also decrease task motivation (e.g. Smith, Sansone, and White, 2007), social
motivation (e.g. Martiny and Nikitin, in press) and sense of belonging (e.g. Walton and Cohen,
2007) and have a negative impact on health and well-being (e.g. Blascovich et al., 2001;
Guendelman et al., 2011; Inzlicht and Kang, 2010). Therefore, future studies might investigate the
extent to which stereotype threat has consequences for Turkish-origin migrants beyond their imme-
diate performance.

Practical implications
Results showed that negative competence-related stereotypes about Turkish-origin migrants are
endorsed by preservice teachers and the activation of those stereotypes in test situations can hinder
Turkish-origin migrants from fulfilling their intellectual potential. These social-psychological vari-
ables provide a new explanation for ethnic performance differences in the German educational
system and can thus make a unique contribution to the unexplained performance differences
between Turkish-origin and German students. Further, the experimental approach shows that the
achievement gap does not depend on ability differences between Turkish-origin migrants and
18 European Educational Research Journal 00(0)

Germans, but rather that Turkish-origin migrants underperform only when negative stereotypes are
activated. Put differently, ethnic performance differences are dependent on the performance situa-
tion – in turn, the achievement gap would be reduced when ethnic group membership and the cor-
responding stereotypes were not salient in performance situations. Including moderator variables
that explain Turkish-origin migrants’ susceptibility to stereotype threat also shows that not all
individuals experience stereotype threat to the same extent. Instead, Turkish-origin students vary
in their perception of the performance situation as a threat to their social identity. These results
open up pathways to interventions designed to increase equal opportunities for all ethnic groups in
the German educational system.
One approach is to develop interventions that create learning and testing situations in which nega-
tive group-related stereotypes do not hinder performance (i.e. identity-safe testing environments, that
is, testing environments in which social identity is not threatened; Spencer et al., 2016). For example,
teachers as well as authors of testing materials could reduce the likelihood that students are reminded
of their membership in negatively stereotyped groups (e.g. instead of asking students to indicate their
gender, ethnicity, or native language before performance tests, they should always assess this infor-
mation after the test). Furthermore, preservice teachers as well as certified teachers should receive
training about stereotype threat. If teachers endorse negative competence-related stereotypes about
migrant students, these stereotypes can shape their attitudes and behavior. Stereotypes can even be
activated automatically and unconsciously and subsequently influence behavior (e.g. Bargh, Chen,
and Burrows, 1996), leading to inadvertent and unconscious discriminatory behavior by teachers
against students (e.g. having low performance expectations, not calling on Turkish-origin students
when difficult questions are asked, or providing dependency-oriented help instead of autonomy-ori-
ented help). When (preservice) teachers reflect on their own stereotypes during their training, they
might develop conscious and controlled strategies to recognize and counteract automatic stereotype
activation. In doing so, they may be able to develop strategies that control the influence of stereotypes
on performance expectations and behavior toward Turkish-origin students (for practical recommen-
dations, see Mok, Froehlich, and Scholz, 2015). One important aspect of teacher training is to inform
teachers that everyone has stereotypes, because stereotypes help to structure the perceived social
environment (e.g. Klauer, 2008). Thus, the goal of teacher trainings is not that teachers “get rid” of
their stereotypes, but that they are aware of them and try to limit their influence on their perceptions
of and behavior toward Turkish-origin students in classrooms.
Another approach is to develop interventions directed at the students with the goal of decreasing
their susceptibility to stereotype threat. In this case, the aim is to reduce the stereotype-induced
pressure that students experience in performance situations rather than to eliminate the stereotypes
themselves. These interventions would attempt to change students’ views that their social identities
are threatened by the expressed stereotypes (i.e. coping interventions, Spencer et al., 2016). For
example, interventions that teach students to endorse an incremental theory of intelligence (i.e. the
notion that one can expand one’s intelligence and ability by learning and effort; e.g. Dweck, 2006)
could increase students’ resilience to the activation of negative stereotypes. Typically, these inter-
ventions instruct students that “the brain is like a muscle – it changes as you train it,” which con-
veys that they can increase their ability by learning and effort. This message motivates students and
alleviates their concern about not being smart enough in performance situations (e.g. Blackwell,
Trzesniewski, and Dweck, 2007; Martinez and Mendoza-Denton, 2011; Yeager et al., 2016, but see
recent meta-analysis by Sisk et al., 2018).
Another group of interventions aims at increasing negatively stereotyped students’ feelings of
belongingness with the academic environment (e.g. Walton and Cohen, 2007, 2011). Although a
recent social belonging intervention with migrant students in Austria did not yield significant inter-
vention effects, the initial levels of belongingness of migrant students (who were actually in the
majority in their classrooms) were already relatively high and perhaps could not be further increased
Froehlich et al. 19

by the intervention (Weber, Kronberger, and Appel, 2018). Further research to adapt interventions
developed in the US-American educational context to the European and German context is needed.
A further group of interventions directed at students relies on affirmation of the self or an
important value of the self. With this approach, before taking a test, students are asked to indicate
a characteristic of themselves that they value the most (e.g. sense of humor, creativity, social
skills) and reflect on why this value is important to them. This activity affirms their self-integrity,
which would otherwise be threatened by the negative stereotypes. Consequently, students per-
ceive the subsequent performance situation as less threatening (e.g. Martens et al., 2006; Sherman
et al., 2013). However, a recent study conducting a values affirmation intervention with migrant
students in the Netherlands was not able to show beneficial effects of the intervention on migrants’
performance (de Jong et al., 2016). The authors suggest that their intervention might have failed
to produce the expected effects because the values affirmation exercise was originally developed
for a different population, that is, African-American students in the United States, and did not
generalize to predominantly Muslim migrant students in the Netherlands. More specifically, the
Muslim students frequently chose religious and communal values to reflect on, which might have
increased instead of attenuated stereotype threat, because of the common link between religious
and ethnic identity in this group (Verkuyten and Yildiz, 2007). As long as it is not yet feasible to
create identity-safe testing sessions in which stereotypes are never salient, the more promising
strategy for reducing the ethnic achievement gap is to reduce students’ susceptibility to stereotype
threat by interventions that teach them ways to cope with the social identity threats (Spencer
et al., 2016).

Conclusion
The academic performance of Turkish-origin migrants is partly shaped by their ethnic group mem-
bership – a category that is negatively evaluated in German society. The current review and meta-
analysis highlights the complex interplay of group membership, negative stereotypes, social
identity, and academic performance. The social-psychological approach based on stereotype threat
theory opens up new pathways to interventions that can reduce ethnic inequalities in the German
educational system. Teachers are encouraged to create identity-safe testing environments and
reflect on their own stereotypes about (Turkish-origin) migrants.

Declaration of conflicting interest

The authors declare that there is no conflict of interest.

Funding
This research was funded by the Federal Ministry of Education and Research (BMBF); grant number:
01JC1104.

Note
1. Project title: “Stereotype Threat als Ursache niedriger Leistungen von Schülerinnen und Schülern mit
Migrationshintergrund im deutschen Bildungssystem”; data are available upon request from the IQB
(Institut für Qualitätssicherung im Bildungswesen; project webpage: https://www.iqb.hu-berlin.de/fdz/
studies/Stereotype_Threat).

References
Adams G, Garcia DM, Purdie-Vaughns V, et al. (2006) The detrimental effects of a suggestion of sexism in
an instruction situation. Journal of Experimental Social Psychology 42(5): 602–615.
20 European Educational Research Journal 00(0)

Appel M, Weber S and Kronberger N (2015) The influence of stereotype threat on immigrants: Review and
meta-analysis. Frontiers in Psychology, 6: 900.
Aronson J and Dee T (2012) Stereotype threat in the real world. In: Inzlicht M and Schmader T (eds)
Stereotype Threat: Theory, Process, and Application. New York: Oxford University Press, pp.264–279.
Asbrock F (2010) Stereotypes of social groups in Germany in terms of warmth and competence. Social
Psychology 41(2): 76–81.
Autorengruppe Bildungsberichterstattung (2016) Bildung in Deutschland 2016: Ein indikatorengestütz-
ter Bericht Mit einer Analyse zu Bildung und Migration [Education in Germany 2016: An Indicator-
Supported Report Including an Analysis of Education and Migration]. Bielefeld: W. Bertelsmann Verlag.
Baier D, Pfeiffer C, Rabold S, et al. (2010) Kinder und Jugendliche in Deutschland: Gewalterfahrungen,
Integration, Medienkonsum: Zweiter Bericht zum Gemeinsamen Forschungsprojekt des
Bundesministeriums des Inneren und des KFN [Children and Adolescents in Germany: Experiences of
Violence, Integration, and Media Use]. Forschungsbericht N. 109. [Research Report No 109]. Hannover:
Kriminologisches Forschungsinstitut Niedersachsen (KFN).
Bargh JA, Chen M and Burrows L (1996) Automaticity of social behavior: Direct effects of trait construct and
stereotype activation on action. Journal of Personality and Social Psychology 71(2): 230–244.
Baysu G, Celeste L, Brown R, et al. (2016) Minority adolescents in ethnically diverse schools: Perceptions of
equal treatment buffer threat effects. Child Development 87(5): 1352–1366.
Baysu G, Phalet K and Brown R (2011) Dual identity as a two-edged sword: Identity threat and minority
school performance. Social Psychology Quarterly 74(2): 121–143.
Blackwell LS, Trzesniewski KH and Dweck CS (2007) Implicit theories of intelligence predict achievement
across an adolescent transition: A longitudinal study and an intervention. Child Development 78(1):
246–263.
Blascovich J, Spencer SJ, Quinn D, et al. (2001) African Americans and high blood pressure: The role of
stereotype threat. Psychological Science 12(3): 225–229.
Borenstein M, Hedges LV, Higgins JPT, et al. (eds) (2009) Introduction to Meta-Analysis. Chichester: Wiley.
Bundeszentrale für politische Bildung (2016) Die Soziale Situation in Deutschland: Migration:
Migrationshintergrund I [The social situation in Germany: Migration. Migation background I]. Available
at: http://www.bpb.de/nachschlagen/zahlen-und-fakten/soziale-situation-in-deutschland/61646/migra-
tionshintergrund-i (accessed 13 September 2017).
Cadinu M, Maass A, Lombardo M, et al. (2006) Stereotype threat: The moderating role of locus of control
beliefs. European Journal of Social Psychology 36(2): 183–197.
Cohen J (1988) Statistical Power Analysis for the Behavioral Sciences. 2nd ed. Hillsdale, NJ: L. Erlbaum
Associates.
Cuddy AJC, Fiske ST and Glick P (2007) The BIAS map: Behaviors from intergroup affect and stereotypes.
Journal of Personality and Social Psychology 92(4): 631–648.
De Jong EM, Jellesma FC, Koomen HMY, et al. (2016) A values-affirmation intervention does not benefit
negatively stereotyped immigrant students in the Netherlands. Frontiers in Psychology 7: 691.
De Paola M and Brunello G (2016) Education as a tool for the economic integration of migrants. Available at:
http://www.eenee.de/dms/EENEE/Analytical_Reports/EENEE_AR27.pdf (accessed 13 September 2017).
Die Beauftragte der Bundesregierung für Migration, Flüchtlinge und Integration (2016) 11. Bericht
der Beauftragten der Bundesregierung für Migration, Flüchtlinge und Integration – Teilhabe,
Chancengleichheit und Rechtsentwicklung in der Einwanderungsgesellschaft Deutschland [11th report
of the Federal Government Commissioner for migration, refugees and integration – participation, equal
opportunities and legal development in the immigration society of Germany]. Available at: https://
www.bundesregierung.de/Content/Infomaterial/BPA/IB/11-Lagebericht_09-12-2016.html (accessed 13
September 2017).
Dweck CS (1999) Self-Theories: Their Role in Motivation, Personality, and Development. Philadelphia, PA:
Psychology Press.
Dweck CS (2006) Mindset: The New Psychology of Success. New York: Random House.
Dweck CS, Chiu C and Hong Y (1995) Implicit theories and their role in judgments and reactions: A word
from two perspectives. Psychological Inquiry 6(4): 267–285.
Froehlich et al. 21

Elizaga RA and Markman KD (2008) Peers and performance: How in-group and out-group comparisons
moderate stereotype threat effects. Current Psychology 27(4): 290–300.
Ellemers N, Knippenberg A and Wilke H (1990) The influence of permeability of group boundaries and
stability of group status on strategies of individual mobility and social change. British Journal of Social
Psychology 29(3): 233–246.
Fiske ST, Cuddy AJC, Glick P, et al. (2002) A model of (often mixed) stereotype content: Competence and
warmth respectively follow from perceived status and competition. Journal of Personality and Social
Psychology 82(6): 878–902.
Froehlich L and Schulte I (2018) Warmth and competence stereotypes about immigrant groups in Germany.
Manuscript sumbitted for publication.
Froehlich L, Martiny SE, Deaux K, et al. (2016a) Being smart or getting smarter: Implicit theory of intelli-
gence moderates stereotype threat and stereotype lift effects. British Journal of Social Psychology 55(3):
564–587.
Froehlich L, Martiny SE, Deaux K, et al. (2016b) “It’s their responsibility, not ours”: Stereotypes about com-
petence and causal attributions for immigrants’ academic underperformance. Social Psychology 47(2):
74–86.
Grant H and Dweck CS (2001) Cross-cultural response to failure: Considering outcome attributions with dif-
ferent goals. In: Salili F, Chiu C-Y and Hong Y-Y (eds) Student Motivation: The Culture and Context of
Learning. New York: Kluwer Academic, pp.203–219.
Grant H and Dweck CS (2003) Clarifying achievement goals and their impact. Journal of Personality and
Social Psychology 85(3): 541–553.
Guendelman MD, Cheryan S and Monin B (2011) Fitting in but getting fat: Identity threat and dietary choices
among U.S. immigrant groups. Psychological Science 22(7): 959–967.
Güngör D, Karasawa M, Boiger M, et al. (2014) Fitting in or sticking together: The prevalence and adaptivity
of conformity, relatedness, and autonomy in Japan and Turkey. Journal of Cross-Cultural Psychology
45(9): 1374–1389.
Higgins JPT, Thompson SG, Deeks JJ, et al. (2003) Measuring inconsistency in meta-analyses. BMJ: British
Medical Journal 327(7414): 557–560.
Hofstede GH, Hofstede GJ and Minkov M (2010) Cultures and Organizations: Software of the Mind:
Intercultural Cooperation and its Importance for Survival. 3rd ed. New York: McGraw-Hill.
Inzlicht M and Kang SK (2010) Stereotype threat spillover: How coping with threats to social identity affects
aggression, eating, decision making, and attention. Journal of Personality and Social Psychology 99(3):
467–481.
Inzlicht M and Schmader T (2012) Introduction. In: Inzlicht M and Schmader T (eds) Stereotype Threat:
Theory, Process, and Application. New York: Oxford University Press, pp.3–14.
Jetten J, Branscombe NR, Schmitt MT, et al. (2016) Rebels with a cause: Group identification as a response to per-
ceived discrimination from the mainstream. Personality and Social Psychology Bulleti, 27(9): 1204–1213.
Jones BD, Ruff C and Paretti MC (2013) The impact of engineering identification and stereotypes on under-
graduate women’s achievement and persistence in engineering. Social Psychology of Education 16(3):
471–493.
Kassin SM, Fein S and Markus HR (2011) Social psychology. 7th ed. Belmont, CA: Wadsworth.
Keller J (2007) Stereotype threat in classroom settings: The interactive effect of domain identification, task
difficulty and stereotype threat on female students’ maths performance. British Journal of Educational
Psychology 77(2): 323–338.
Kervyn N, Fiske ST and Yzerbyt V (2015) Forecasting the primary dimension of social perception: Symbolic
and realistic threats together predict warmth in the stereotype content model. Social Psychology 46(1):
36–45.
Klauer KC (2008) Soziale Kategorisierung und Stereotypisierung. In: Petersen L-E (ed) Stereotype, Vorurteile
und Soziale Diskriminierung: Theorien, Befunde und Interventionen. 1st ed. Weinheim Basel: Beltz,
PVU, pp.23–32.
Klieme E, Artelt C, Hartig J, et al. (eds) (2010) PISA 2009: Bilanz nach einem Jahrzehnt [PISA 2009: Taking
Stock After a Decade]. Münster: Waxmann Verlag.
22 European Educational Research Journal 00(0)

Koch SC, Müller SM and Sieverding M (2008) Women and computers: Effects of stereotype threat on attribu-
tion of failure. Computers and Education 51(4): 1795–1803.
Komarraju M and Cokley KO (2008) Horizontal and vertical dimensions of individualism-collectivism: A
comparison of African Americans and European Americans. Cultural Diversity and Ethnic Minority
Psychology 14(4): 336–343.
Länderoffene Arbeitsgruppe “Indikatorenentwicklung und Monitoring” (2017) Vierter Bericht zum
Integrationsmonitoring der Länder 2013–2015. [Fourth report of the intergration monitoring of the
federal states 2013–2015]. Konferenz der für Integration zuständigen Ministerinnnen und Minister/
Senatorinnen und Senatoren der Länder (IntMK). Available at: http://www.integrationsmonitoring-
laender.de/sites/default/files/integrationsbericht_2015_endfassung_online.pdf (accessed 11 October
2018).
Lee TL and Fiske ST (2006) Not an outgroup, not yet an ingroup: Immigrants in the stereotype content model.
International Journal of Intercultural Relations 30(6): 751–768.
Levy SR, Stroessner SJ and Dweck CS (1998) Stereotype formation and endorsement: The role of implicit
theories. Journal of Personality and Social Psychology 74(6): 1421–1436.
Martens A, Johns M, Greenberg J, et al. (2006) Combating stereotype threat: The effect of self-affirmation on
women’s intellectual performance. Journal of Experimental Social Psychology 42(2): 236–243.
Martinez AG and Mendoza-Denton R (2011) The prospect of plasticity: Malleability views of group differ-
ences and their implications for intellectual achievement, mental/behavioral health, and public policy.
Social Issues and Policy Review 5(1): 137–159.
Martiny SE and Götz T (2011) Stereotype Threat in Lern- und Leistungssituationen: Theoretische Ansätze,
empirische Befunde und praktische Implikationen. In: Dresel M and Lämmle L (eds) Talentförderung –
Expertiseentwicklung – Leistungsexzellenz: Vol. 9. Motivation, Selbstregulation und Leistungsexzellenz.
Münster: LIT, pp.153–177.
Martiny SE and Nikitin J (in press) Social Identity Threat in Interpersonal Relationships: Activating Negative
Stereotypes Decreases Social Approach Motivation. Journal of Experimental Psychology: Applied.
Martiny SE, Mok SY, Deaux K, et al. (2014a) Effects of activating negative stereotypes about Turkish-origin
students on performance and identity management in German high schools. International Review of
Social Psychology 3(Tome 27): 205–225.
Martiny SE, Mok SY, Froehlich L, et al. (2014b) Does a Test Fairness Manipulation Trigger a Stereotype
Threat Effect on Verbal Performance of Turkish-Origin Students in Germany? Unpublished manuscript.
University of Konstanz, Germany.
Mok SY, Froehlich L and Scholz C (2015) Leistungsminderung durch negative Stereotype im Schulkontext
[Performance reduction due to negative stereotypes in school]. Lehren and Lernen 41(1): 28–31.
Mok SY, Martiny SE, Froehlich L, et al. (2014) Do Turkish-Origin Students in Germany Suffer From
Stereotype Threat Effects in Math When a Test is Labeled as Diagnostic for Students’ Intelligence?
Unpublished manuscript.
Mok SY, Martiny SE, Gleibs IH, et al. (2016) The relationship between ethnic classroom composition
and Turkish-origin and German students’ reading performance and sense of belonging. Frontiers in
Psychology 7(1071): 366.
Mok SY, Martiny SE, Gleibs IH, et al. (2017) The interaction of vertical collectivism and stereotype activa-
tion on the performance of Turkish-origin high school students. Learning and Individual Differences
56: 76–84.
Muzzatti B and Agnoli F (2007) Gender and mathematics: Attitudes and stereotype threat susceptibility in
Italian children. Developmental Psychology 43(3): 747–759.
Nadler JT and Clark MH (2011) Stereotype threat: A meta-analysis comparing African Americans to Hispanic
Americans. Journal of Applied Social Psychology 41(4): 872–890.
Nguyen H-HD and Ryan AM (2008) Does stereotype threat affect test performance of minorities and women?
A meta-analysis of experimental evidence. Journal of Applied Psychology 93(6): 1314–1334.
Pennington CR, Heim D, Levy AR, et al. (2016) Twenty years of stereotype threat research: A review of
psychological mediators. PloS One 11(1): e0146487.
Froehlich et al. 23

Phalet K and Claeys W (1993) A comparative study of Turkish and Belgian youth. Journal of Cross-Cultural
Psychology 24(3): 319–343.
Prenzel M (2013) PISA 2012: Fortschritte und Herausforderungen in Deutschland. [PISA 2012: Advancements
and challenges in Germany]. Münster: Waxmann.
Pronin E, Steel CM and Ross L (2004) Identity bifurcation in response to stereotype threat: Women and math-
ematics. Journal of Experimental Social Psychology 40(2): 152–168.
Rauch D, Mang J, Härtig H, et al. (2016) Naturwissenschaftliche Kompetenz von Schülerinnen und Schülern
mit Zuwanderungshintergrund [Science competencies of students with migration background]. In: Reiss
K, Sälzer C, Schiepe-Tiska A, et al. (eds) PISA 2015: Eine Studie Zwischen Kontinuität und Innovation.
Münster: Waxmann, pp.317–348.
Rydell RJ and Boucher KL (2010) Capitalizing on multiple social identities to prevent stereotype threat: The
moderating role of self-esteem. Personality and Social Psychology Bulletin 36(2): 239–250.
Sander A, Ohle A, McElvany N, et al. (2018) Stereotypenbedrohung als Ursache für geringeren
Wortschatzzuwachs bei Grundschulkindern mit Migrationshintergrund [Stereotype threat as a cause of
lower vocabulary increase for elementary school students with migration background]. Zeitschrift Für
Erziehungswissenschaft 21(1): 177–197.
Schmader T (2002) Gender identification moderates stereotype threat effects on women’s math performance.
Journal of Experimental Social Psychology 38(2): 194–201.
Schmader T and Beilock SL (2012) An integration of processes that underlie stereotype threat. In: Inzlicht
M and Schmader T (eds) Stereotype Threat: Theory, Process, and Application. New York: Oxford
University Press, pp.34–50.
Schmader T, Johns M and Barquissau M (2004) The costs of accepting gender differences: The role of stereo-
type endorsement in women’s experience in the math domain. Sex Roles 50(11–12): 835–850.
Schmader T, Johns M and Forbes C (2008) An integrated process model of stereotype threat effects on per-
formance. Psychological Review 115(2): 336–356.
Schmid M (ed) (2014) Italienische Migration nach Deutschland [Italian Migration to Germany]. Wiesbaden:
Springer Fachmedien Wiesbaden.
Schofield JW and Alexander K (2012) Stereotype threat, Erwartungseffekte und organisatorische
Differenzierung: Schulische Leistungsbarrieren und Ansätze zu ihrer Überwindung. In: Fürstenau S
and Gomolla M (eds) Migration und Schulischer Wandel: Leistungsbeurteilung. Dordrecht: Springer,
pp.65–88.
Schwippert K, Wendt H and Tarelli I (2012) Lesekompetenzen von Schülerinnen und Schülern mit
Migrationshintergrund [Reading competencies of students with migration background]. In: Bos W (ed)
IGLU 2011: Lesekompetenzen von Grundschulkindern in Deutschland im Internationalen Vergleich.
Münster: Waxmann pp.191–208.
Sherman DK, Hartson KA, Binning KR, et al. (2013) Deflecting the trajectory and changing the narrative:
How self-affirmation affects academic performance and motivation under identity threat. Journal of
Personality and Social Psychology 104(4): 591–618.
Shih M, Pittinsky TL and Trahan A (2006) Domain-specific effects of stereotypes on performance. Self and
Identity 5(1): 1–14.
Singelis TM, Triandis HC, Bhawuk DPS, et al. (1995) Horizontal and vertical dimensions of individual-
ism and collectivism: A theoretical and measurement refinement. Cross-Cultural Research 29(3):
240–275.
Sisk VF, Burgoyne AP, Sun J, et al. (2018) To what extent and under which circumstances are growth
mind-sets important to academic achievement? Two meta-analyses. Psychological Science 29(4):
549–571.
Smith JL, Sansone C and White PH (2007) The stereotyped task engagement process: The role of interest and
achievement motivation. Journal of Educational Psychology 99(1): 99–114.
Spencer SJ, Logel C and Davies PG (2016) Stereotype threat. Annual Review of Psychology 67: 415–437.
Spencer SJ, Steele CM and Quinn DM (1999) Stereotype threat and women’s math performance. Journal of
Experimental Social Psychology 35(1): 4–28.
24 European Educational Research Journal 00(0)

Stanat P, Rauch D and Segeritz M (2010) Schülerinnen und Schüler mit Migrationshintergrund [Students
with migration background]. In: Klieme E, Artelt C, Hartig J, et al. (eds) PISA 2009: Bilanz Nach einem
Jahrzehnt [PISA 2009: Taking Stock After a Decade]. Münster: Waxmann Verlag, pp.200–230.
Statistisches Bundesamt (2016) Datenreport 2016: Ein Sozialbericht für die Bundesrepublik Deutschland
[Data Report 2016: A Social Report for the Federal Republic of Germany]. Bonn: Reihe Zeitbilder.
Statistisches Bundesamt (2017) Bevölkerung und Erwerbstätigkeit: Bevölkerung mit Migrationshintergrund:
Ergebnisse des Mikrozensus 2016. [Population and Labor: Population With Migration Background.
Results of the Micro Census 2016]. Fachserie 1 Reihe 2.2. Wiesbaden: Statistisches Bundesamt.
Steele CM (1997) A threat in the air: How stereotypes shape intellectual identity and performance. American
Psychologist 52(6): 613–629.
Steele CM and Aronson J (1995) Stereotype threat and the intellectual test performance of African Americans.
Journal of Personality and Social Psychology 69(5): 797–811.
Steele CM, Spencer SJ and Aronson J (2002) Contending with group image: The psychology of stereotype
and social identity threat. In: Zanna MP (ed) Advances in Experimental Social Psychology, Vol. 34. San
Diego, CA: Academic Press, pp.379–440.
Strasser J (2012) Kulturelle Stereotype und ihre Bedeutung für das Verstehen in Schule und Unterricht
[Cultural stereotypes and their meaning for understanding in education and instruction]. In: Wiater W and
Manschke D (eds) Verstehen und Kultur. Wiesbaden: VS Verlag für Sozialwissenschaften, pp.191–215.
Tajfel H and Turner JC (1979) An integrative theory of intergroup conflict. In: Austin WG and Worchel S
(eds) The Social Psychology of Intergroup Relations. Monterey, CA: Brooks/Cole, pp.33–47.
Tajfel H and Turner JC (1986) The social identity theory of intergroup behavior. In:Worchel S and Austin
WG (eds) Psychology of Intergroup Relations. Chicago: Nelson-Hall, pp.7–24.
Varnum MEW, Grossmann I, Kitayama S, et al. (2010) The origin of cultural differences in cognition:
Evidence for the social orientation hypothesis. Current Directions in Psychological Science 19(1): 9–13.
Verkuyten M and Yildiz AA (2007) National (dis)identification and ethnic and religious identity: A study
among Turkish-Dutch muslims. Personality and Social Psychology Bulletin 33(10): 1448–1462. 04276
Verkuyten M, Thijs J and Canatan K (2001) Achievement motivation and academic performance among
Turkish early and young adolescents in the Netherlands. Genetic, Social, and General Psychology
Monographs 127(4): 378–408.
Walton GM and Cohen GL (2003) Stereotype lift. Journal of Experimental Social Psychology 39(5): 456–467.
Walton GM and Cohen GL (2007) A question of belonging: Race, social fit, and achievement. Journal of
Personality and Social Psychology 92(1): 82–96.
Walton GM and Cohen GL (2011) A brief social-belonging intervention improves academic and health out-
comes of minority students. Science 331(6023): 1447–1451.
Walton GM and Spencer SJ (2009) Latent ability: Grades and test scores systematically underestimate the
intellectual ability of negatively stereotyped students. Psychological Science 20(9): 1132–1139.
Weber S, Appel M and Kronberger N (2015) Stereotype threat and the cognitive performance of adolescent
immigrants: The role of cultural identity strength. Contemporary Educational Psychology 42: 71–81.
Weber S, Kronberger N and Appel M (2018) Immigrant students’ educational trajectories: The influence of
cultural identity and stereotype threat. Self and Identity 17(2): 211–235.
Wendt H, Schwippert K and Stubbe TC (2016) Mathematische und naturwissenschaftliche Kompetenzen
von Schülerinnen und Schülern mit Migrationshintergrund [Math and science competencies of students
with migration background]. In: Wendt H, Bos W, Selter C, et al. (eds) TIMSS 2015: Mathematische
und Naturwissenschaftliche Kompetenzen von Grundschulkindern in Deutschland im Internationalen
Vergleich. Münster and New York: Waxmann, pp.317–331.
Whitley BE and Kite ME (2016) Psychology of Prejudice and Discrimination. 3rd ed. New York: Routledge.
Woodcock A, Hernandez PR, Estrada M, et al. (2012) The consequences of chronic stereotype threat: Domain
disidentification and abandonment. Journal of Personality and Social Psychology 103(4): 635–646.
Yeager DS, Walton GM, Brady ST, et al. (2016) Teaching a lay theory before college narrows achievement
gaps at scale. Proceedings of the National Academy of Sciences of the United States of America 113(24):
E3341–E3348.
Froehlich et al. 25

Author biographies
Laura Froehlich is a Post-Doctoral Researcher in the Department of Social Psychology at the FernUniversität
in Hagen, Germany. She received her PhD in Psychology from the University of Konstanz in 2015. Her
research interests are stereotypes about gender and ethnicity, self & identity processes for immigrants as well
as cross-cultural approaches to social psychology.
Sog Yee Mok is a Post-Doctoral Researcher at the Institute for Educational Science at the University of
Zurich, Switzerland. She received her PhD in Empirical Educational Research from the University of
Konstanz in 2015. Her research interests are stereotype threat and identity processes for students with migra-
tion background, as well as evidence-based teacher training and teacher-student-interactions.
Sarah E. Martiny is a Full Professor of Social and Community Psychology at UiT The Arctic University of
Norway in Tromsø, Norway. She received her PhD from the University of Jena in 2008 and worked as a Post-
Doctoral Researcher with Kay Deaux at New York University in 2009.She was an Assistant Professor in
Empirical Educational Research at the University of Konstanz from 2010 to 2014. Her research interests are
in the area of social identity and intergroup relations including topics such as migration, gender, and social
inequality.
Kay Deaux is a Distinguished Professor Emerita at the Graduate Center of the City University of New York
and a Visiting Scholar at New York University, USA. She received the Distinguished Scholar Award from
the Society of Personality and Social Psychology in 2016. She has a long history in research on stereotypes,
discrimination, and social identity in the areas of gender and immigration.