STEM teachers who believe that the ability is fixed have greater gaps in racial achievement and inspire less motivation from students in their classes



An important goal of the scientific community is to broaden the success and participation of racial minorities in the fields of STEM. Yet, faculty beliefs about stability of capacity may be an unwitting and neglected barrier for stigmatized students. Results from a university-wide longitudinal sample (150 STEM professors and more than 15,000 students) revealed that racial achievement gaps in courses taught by mentality teachers were twice as important as the achievement gaps in courses taught by growth-minded teachers. Course evaluations revealed that students were demotivated and had more negative experiences in courses taught by professors with a specific mentality (as opposed to growth). Faculty beliefs predicted student achievement and motivation beyond any other faculty characteristics, including gender, race / ethnicity, age, teaching experience, or tenure. These results suggest that teachers' beliefs have important implications for classroom experiences and the success of underrepresented minority students in STEM.


Despite decades of research and millions of dollars in federal funding to understand and improve the under-representation of diverse people in the STEM (Science, Technology, Engineering and Mathematics) pipeline, Black, Latino and Native American students [underrepresented racial/ethnic minorities (URM)] continue to underperform academically relative to their white peers (1). While these differences in race achievement are determined by multiple factors (economic and structural, for example), they can be exacerbated by subtle situational signals from STEM teachers that reinforce racial stereotypes that social groups are more or less likely to to have STEM skills (2).

The hypothesis of clues suggests that threatening situation indices in STEM environments, such as the diagnosis of a test (24), can lead URM students to fear being judged on the basis of stereotypes of ability, resulting in loss of motivation, intellectual underperformance and greater gaps in racial achievement in STEM classes (57). This study examines the role of a new stereotypical underperformance situational signal: STEM faculty teachers' beliefs about fixity or malleability of capacity (8) – and examines whether these beliefs are associated with the motivation of URM students and their academic success in these teachers' STEM courses.

People's mentalities (also called implicit theories or secular theories) are their beliefs about the fixity or malleability of human characteristics such as intelligence or personality (8). Faculty members who subscribe to strong beliefs subscribe to the idea that intelligence and abilities are fixed and innate qualities that it is impossible to change or develop much. In contrast, faculty members who subscribe to growth beliefs share the idea that capacity is malleable and that it can be developed through persistence, good strategies, and quality mentoring. Professors with a fixed state of mind are more likely to judge a student as having a low performance-based ability of only one test (9) and use unhelpful teaching practices, such as encouraging students to drop difficult courses (for example, "not everyone is expected to pursue a career in STEM") ()9).

Faculty members who subscribe to strong beliefs believe that some students have strong innate intellectual abilities, while others do not. What students could they be? Ubiquitous cultural stereotypes suggest that white and Asian students are more gifted in STEM than black, Latino and Native American students. Given that these American cultural stereotypes question the intellectual abilities of MRU students, we predicted that professors who adhere to concepts of a particular state of mind might be particularly demotivating to URM students. that would result in poorer performance among URM students in courses taught by teachers with a determined mindset (as opposed to growth). Classical conclusions about the influence of teacher beliefs on student performance show that when teachers' expectations are low, teachers become less motivated and perform poorly in their classes (ten). These Pygmalion effects are even stronger for URM students (11, 12).

It has been hypothesized that STEM faculty's steadfast beliefs about intelligence and ability would make URM students feel less motivated than their non-stereotypical peers, which is consistent with stereotypical threat theory. Classical studies that document stereotypical threat underperformance effects typically manipulate threatening (rather than non-threatening) situational cues in the learning environment, such as race / ethnicity / gender of an experimenter , and evaluate the intellectual performance of students as the main indicator of the stereotyped threat (2, 7, 13, 14). Drawing on this theoretical framework, this study examines the role of the college faculty mentality as a situational signal that triggers an underperformance of the MRU in STEM courses. We argue that if STEM professors who subscribe to well-established beliefs create a stereotypical threat to URM students, we should see lower student motivation and significantly greater differences in racial outcomes in these faculty courses. compared to courses taught by STEM professors who subscribe to beliefs about growth.

This study examines self-reported STEM faculty convictions and their implications for student motivation and performance. Previous research has examined students' perceptions of faculty beliefs (15However, to the best of our knowledge, no study has examined self-reported beliefs about STEM faculty as a predictor of student achievement. In addition, the effects of teachers' beliefs were only examined in young children (16) and have not been applied in undergraduate populations, where career choices and trajectories are more prominent. We test our hypothesis on a longitudinal sample of STEM faculty, the largest sample to date, beliefs of mentality faculties associated with student records.


To test our hypothesis, we examined the links between faculty beliefs and racial achievement gaps in these faculty courses over seven semesters (2 years) and more than 15,000 undergraduate student records. Use of two-point validated beliefs about intelligence measurement (8), we surveyed the STEM faculty (NOT = 150; 40.8% response rate) in a large, selective public university (for example, "To be honest, students have some intelligence and can not really do anything to change it"; α = 0.91, M = 3.87, SD = 1.46). The 13 STEM departments (for example, astronomy, biology, computer science, mathematics and physics) at the university were represented in the sample. More than half (55.3%) of the sample held and the average teaching experience in STEM was 18.4 years. The percentage of women and MSU teachers in the sample was similar to that of STEM teachers across the country (sample of professors: 26.7% women, 4.7% MRU; national level: 20.4% women, 5.2% MRM) (1).

University records provide course notes for all students[[[[NOT = 15,466; 7,172 women (46.4%); 1685 URM (10.9%)]enrolled in all courses (not = 634) taught by STEM faculty respondents in seven academic terms. Thus, the student data in this study represent a census (the entire population of individuals in an environment) rather than a sample used to estimate the population. A multilevel regression model captures the nested nature of the data (nested students in courses, nested within the faculty) and takes into account confounding factors such as past student achievement (SAT scores) and all characteristics of the student. courses and teachers available (17). All variables have been normalized so that the coefficients of the multilevel model can be interpreted as effect sizes (18). Finally, we added partially cross-checked random effects to the model, as students could enroll in multiple courses of the same faculty member or in courses of multiple faculty members of the sample over the course of a year. seven academic terms (19). Table S1 provides estimates of fixed effects from the model.

On average, all students score lower on STEM courses taught by teachers who have adopted more rigid beliefs (as opposed to growth) (B = 0.08, P = 0.011). However, in keeping with the stereotyped threat and the clues hypothesis, teachers' strong beliefs were more closely associated with lower course performance among Black, Latino and Native American (URM) students (B = 0.12, P = 0.001) than among white and Asian students (non-URM; B = 0.08, P = 0.010; group × faculty interaction: B = 0.04, P = 0.041; Fig. 1). On average, non-URM students scored an average of .14 point (on a scale of 4.0) higher than that of URM students, but in courses taught by professors who more closely adhered to a state of the art. determined spirit (-1), the racial achievement gap reached 0.19 GPA points (URM GPA = 2.71; non-URM GPA = 2.90). However, in courses taught by teachers who have more adopted a growth mindset (+1 ET), the racial achievement gap has been reduced to 0.10 GPA point (GPA URM = 2). 96, non-URM GPA = 3.06). Thus, the gap between racial outcomes was nearly twice as great in courses taught by university professors who subscribed to beliefs embedded in the mind (as opposed to growth) about student abilities.

Fig. 1 The faculty's beliefs predict the racial achievement gap in STEM courses.

The predicted values ​​are calculated from the interaction between the professors' convictions (fixed status = -1 SD, growth = +1 SD) and the URM status (black, Hispanic, Native American) of the students. The error bars represent ± 1 SE.

Which STEM teachers are more likely to endorse strong beliefs?

Do faculty members who subscribe to well-established beliefs tend to be men or women? White, Asian or URM? Teachers, both men and women, were just as likely to endorse strong beliefs (B = 0.14, P = 0.648; Table 1), and there was no difference in mentality by race / ethnicity (B = 0.03, P = 0.956). As social desirability and awareness of mind beliefs develop (20), it is possible that the explicit assent of firm beliefs is generational, so that older faculty members (compared to younger ones) are more likely to approve them. Similarly, it is possible that permanent (as opposed to non-permanent) professors with more experience (as opposed to) in college education may adhere to more rigid beliefs. However, we find no evidence that adherence to specific conceptions of mind differs according to the age of the teachers, their tenure or the number of years of experience in university education (all Ps> 0.35). It may also be that fixed beliefs have more weight in some STEM disciplines (21). However, we found that well-established beliefs transcended STEM disciplines and were equally adopted across the 13 STEM disciplines in our sample (all Ps> 0.14). Thus, it seems that well-established beliefs are neither gendered nor generational, but only to members of the majority group, they are simply a function of accumulated experience in teaching, or more concentrated in some STEM disciplines.

Table 1 Characteristics of the faculty predicting the beliefs of the faculty.

Higher scores on faculty beliefs reflect a growth mentality. The sex was coded as follows: female = 1, male = 0. The race / ethnicity was coded as follows: URM (Black, Hispanic, Native American) = 1, non-URM (White, Asian) = 0. The occupancy status was coded as follows: tenured = 1, nonenured = 0. Biology was used as the reference group for fictitious STEM discipline codes.

Explore other faculty characteristics as additional predictors of the underperformance of the URM

Are faculty characteristics alone exacerbating or mitigating the underperformance of the MRU and are beliefs based on a fixed state of mind more threatening when they come from a faculty presenting some demographic characteristics? For example, is it worse for MUR students when a white teacher subscribes to beliefs based on mentality (rather than growth)? Prototype studies of scientists and engineers show that students often evoke older white men as guardians of science (22Therefore, it is plausible that teachers with these characteristics are more likely to activate the stereotyped threat among MMR students, which would lead to greater differences in racial achievement in these faculty classes. We explored the role of all available faculty characteristics in our data set (ie, faculty gender, race / ethnicity, age, incumbent status, and teaching experience) as ( i) additional predictors of MRU underperformance and (ii) potential faculty moderators. effects of the state of mind.

Role models and exam supervisors of the same race have been shown to protect URM students from the underperformance of stereotypical threats in experimental laboratories (13, 23, 24However, we found that MRU professors (compared to non-MRUs) did not show smaller differences in racial achievement in their classrooms (B = 0.30, P = 0.215). Moreover, the racial identity of the teachers did not protect the URM students from the negative effects of fixed beliefs in the mindset of teachers (race / ethnicity × interaction of mentalities: B = -0.11, P = 0.502) – shared beliefs about mentality were just as bad for URM students when they were approved by professors from White or from URM. Similar results have emerged for faculty sex (all Ps> 0.24). Older professors who are more experienced in the field of education or experts in their field are more at risk from the point of view of the identity of URM students, especially when they subscribe to firm convictions. However, the age of teachers, their teaching experience, and their status did not predict racial achievement gaps in their classrooms (all Ps> 0,19), nor interact with their convictions to predict the students' URM scores (all Ps> 0.41). Demonstration of the strong impact of faculty beliefs on mentality, when student demographics, mentalities and URM status (and all interactions between these variables) were included in the model, teachers' beliefs remain the predictor constant of racial achievement gap their courses (Table S2). This suggests that faculty beliefs are strongly associated with the intellectual performance of MRU students, well beyond those of other faculty characteristics, such as gender, race / ethnicity, age, experience of teaching and teacher status.

What does it mean to be a student in classes taught by professors who subscribe more to a specific state of mind (as opposed to growth)?

If teachers communicate their convictions through verbal and non-verbal behavior (9), professors who subscribe to conventional wisdom should be less inclined to use pedagogical practices that emphasize learning and the potential for growth and development (9, 25, 26). What would be the point of focusing on learning, growth and development if you do not believe that students can develop their skills and abilities? Without faculty emphasis on learning, growth and development, we expected students to be less motivated to do their best in these teachers' classes. If students are less motivated, they should be less inclined to recommend the courses of these teachers to others. It is possible that professors who subscribe to firm principles create more demanding courses, forcing students to spend more time studying and preparing. If this is true, the differences in students 'performance and psychological experiences could be explained by the requirements of these courses (rather than the professors' beliefs).

The average student response to the four-semester course evaluation for all courses taught by all faculty teachers shed light on the experiences of students in these faculty courses. Since student-level responses were not available due to confidentiality issues, we were unable to examine the racial / ethnic differences between student experiences in the classroom. We tested multilevel models, controlling the characteristics of the course and faculty, to take into account the nested courses within the faculty.

In accordance with the theory that teachers' fixed beliefs are demotivating to students, they reported less "motivation to do their best work" in courses taught by professors who adhered to more rigid beliefs (B = 0.09, P = 0.028) (Fig. 2 and Table S3). Students also reported that teachers with a fixed state of mind were less likely to use teaching practices that "emphasize learning and development" (ibid.).B = 0.09, P = 0.005). Exploratory mediation analyzes of the responses to these two questions (see Additional material) revealed that these demotivating pedagogical practices statistically explained the effect of the faculty mentality on the URM and non-URM student grades, although this effect is more important for URM students. Thus, teachers who adopted more rigid beliefs used less motivating teaching practices (at least as reported by the students), and these practices were associated with lower course performance for all students on average and on average. especially for URM students.

Fig. 2 The convictions of the professors make it possible to predict the students' experiences in the STEM courses.

The predicted values ​​are calculated from the average of the state of mind of the faculties (fixed = -1 SD, growth = +1 DS). The error bars represent ± 1 SE. ns, not significant. *P <0.05 and **P <0.01.

Since professors who have adopted fixed-minded beliefs have used less motivational teaching practices than those who have embraced beliefs about the spirit of growth, it is not surprising that students less likely to recommend these courses to others (B = 0.08, P = 0.006). The professors' beliefs did not predict the length of time required for the course (B = -0.04, P = 0.350). This finding suggests that teachers with a specific mindset do not require more students – at least from a student's point of view – than faculty with a growth-oriented mindset; the time spent by students studying or preparing outside the classroom remained the same for all courses taught by professors with a determined and growing mentality.


Our results suggest that teachers 'beliefs can predict students' experiences in their STEM courses and the extent of racial achievement gaps in these courses. We found that racial achievement gaps in courses taught by more stable teachers were twice as high as those taught by more dynamic teachers. To our knowledge, this study examines the largest sample of courses in STEM (> 600) and students (> 15,000) to date, including more than 1,600 students in URM. In addition, he is the first to examine the association of teachers' stated beliefs with the marks of their own students, demonstrating the impact of faculty beliefs on the underperformance of IRMs in STEM courses. . Additional analyzes show that the teachers 'beliefs that are closest to the students' experiences (ie the beliefs of the specific teacher who teaches their class) matter more to the performance of the students in that class than the beliefs of the students. discipline teachers (ie average beliefs of teachers in a STEM discipline). Together, these findings suggest that the beliefs of STEM faculty members in the mindset shape the motivation and success of students in their classes, and that these beliefs are of particular importance to URM students in their classes.

Teachers' beliefs about the nature of intelligence are likely to shape how they structure their courses, how they communicate with students, and how they encourage (or discourage) student persistence (9). These malleable teaching practices have important implications for the motivation, learning, and success of all students in their classes. However, we argue that the beliefs of faculty on which students "have" STEM skills could be a greater barrier to SRM students, as strong beliefs can make stereotypes of group capacity salient, creating context threat of stereotypes. Recent research suggests that when stigmatized students expect to be stereotyped by fixed-minded institutions, they experience less belonging, less confidence in themselves, more anxiety, and less self-esteem. interested parties (27, 28), suggesting that teachers with a fixed state of mind could also cause these unfavorable results in students. In this study, we were unable to directly assess the stereotypical threat experiences of students, as this would have required a survey evaluation on a prohibitive scale (more than 15,000 students, for example). However, it is important to note that the majority of the literature on stereotypes, including the original demonstrations of the threat of stereotypes in the context of race and gender (2, 29), documented the presence of a stereotypical threat by assessing intellectual performance and demonstrating greater underperformance of stigmatized groups in the context of negative situational cues (eg, test diagnosis). Thus, our results are consistent with this tradition of measurement as well as with the stereotyped threat theory. Future research could measure students' threat experience in response to faculty beliefs.

We found that fixed beliefs are not concentrated in some STEM disciplines. Instead, they seem to be distributed relatively evenly among professors from one discipline to the next, suggesting that the negative effects of these beliefs are found in all departments, colleges, and probably in any school. 39, other universities. Concentrated beliefs within disciplines pose additional challenges to stigmatized students. Previous searches published in Science shows that teachers' beliefs about talent (ie whether maximum performance in a field requires talent) when aggregated at the discipline level correlate with the number of registered women and racial minorities to the American PhD program. programs (21), suggesting that brilliant beliefs – on the ground – can discourage the pursuit of higher education among stigmatized groups. This research complements this work by examining how more traditional beliefs – in this case teachers 'beliefs about the fixity (or malleability) of intelligence – shape students' classroom experiences, performance and racial inequalities in their courses. This work suggests that faculty beliefs could be an important predictor of future decisions regarding the pursuit of higher education in specific areas of STEM. Future research could test this possibility.

Fixed mind beliefs also did not match faculty identities (gender, race / ethnicity and age, for example) and experiences (eg, tenure and teaching experience), suggesting that beliefs fixed mental state are problematic for students, regardless of the faculty member's background. However, there are reasons to be optimistic here. Fixed mentality beliefs are variable. Studies have shown that cost-effective educational interventions can help people develop a growth-oriented mindset (30, 31). The teachers' mentality can therefore be a potential lever for the creation of safe collegiate classrooms (32) -Learning environments where all students, regardless of race or ethnicity, feel valued and encouraged to reach their full potential.

Millions of dollars in federal funding have been allocated to student-centered initiatives and interventions that address inequities in higher education and expand the STEM pipeline. Rather than imposing a burden on students and rigid structural factors, our work highlights teachers and how their beliefs relate to the under-performance of students being stigmatized in their STEM courses. By investing resources in interventions to strengthen the mindset of faculty, faculty members could understand the impact of their beliefs on student motivation and performance and help them create growth-minded cultures in their courses, at a low or no cost. Si davantage de professeurs créent des cultures d’esprit de croissance dans leurs classes, cela pourrait alors augmenter la motivation et l’engagement des étudiants dans les STEM, ce qui pourrait inciter davantage d’élèves URM à poursuivre une carrière dans les STEM. Même une légère augmentation du nombre de notes de cours en STEM pourrait faire toute la différence entre l&#39;obtention d&#39;un crédit pour le cours, le maintien de l&#39;aide financière et / ou l&#39;obtention d&#39;un diplôme en STEM. Dans cette étude, 150 professeurs ont enseigné à plus de 15 000 étudiants en seulement deux ans, soulignant ainsi l’influence omniprésente de chaque membre du corps professoral d’un collège. Les interventions centrées sur le corps professoral peuvent avoir le potentiel sans précédent de faire évoluer la culture STEM d’une culture du génie à mentalité fixe à une culture de développement axée sur la mentalité de croissance tout en réduisant les écarts de réussite raciale des STEM à l’échelle (33).



Tous les professeurs STEM actuellement employés (y compris les professeurs adjoints, chargés de cours, post-doctorants et étudiants diplômés) ayant enseigné au moins un cours à l&#39;université ont été recrutés par courrier électronique. Les courriels ont été obtenus à partir de dossiers universitaires. Au total, 483 professeurs de STEM ont été contactés et 197 ont fourni des données utilisables (40,8%). Nous avons exclu 45 professeurs qui n’avaient pas enseigné au moins un cours de premier cycle au cours des 2 années précédentes et 2 professeurs qui n’avaient pas répondu aux deux questions portant sur les convictions. L&#39;échantillon final comprenait 150 membres du corps professoral de 13 départements STEM: astronomie, biologie, biochimie, biotechnologie, chimie, sciences cognitives, informatique, économie, géologie, informatique, mathématiques, physique et statistiques. Voir le matériel supplémentaire pour une comparaison entre les professeurs de STEM ayant accepté de participer à l&#39;étude et ceux ayant choisi de ne pas participer.

Mesures d&#39;enquête auprès des professeurs

Les participants ont répondu au sondage en ligne et ont été invités à «prendre en compte les étudiants de premier cycle auxquels vous enseignez (ou avez enseigné) à [the university] en répondant à ces questions. "Les croyances des professeurs ont été mesurées à l&#39;aide de deux éléments (" Pour être honnête, les étudiants ont une certaine intelligence et ne peuvent vraiment pas grand chose à faire pour la changer "; vous ne pouvez pas beaucoup changer »(α = 0,91) sur une échelle de 1 (tout à fait d&#39;accord) à 6 (pas du tout d&#39;accord). Des scores plus élevés sur la mesure de croyance de la mentalité des facultés représentaient une mentalité plus axée sur la croissance. L’expérience d’enseignement a été mesurée à l’aide d’un élément («combien d’années avez-vous enseigné dans votre domaine?»). Les participants ont été invités à fournir leur sexe, leur race / appartenance ethnique et leur âge. Le statut d&#39;occupation a été recueilli à partir des archives de l&#39;université.

Variables étudiantes

Les archives universitaires indiquaient le sexe, la race / ethnie, le statut de première génération et le score SAT des étudiants pour tous les étudiants (NOT = 15 466; 46,4% de femmes) inscrites à tous les cours (n = 634) enseigné par les répondants du corps professoral STEM au cours de sept trimestres académiques. Les élèves noirs, hispaniques, amérindiens / natifs d’Alaska et d’origines hawaïenne / insulaire du Pacifique ont été classés dans la catégorie des minorités sous-représentées (URM; n = 1685; 10,9%). Les étudiants blancs et asiatiques ont été classés en tant que groupe majoritaire (n = 13 781, 89,1%). Les étudiants qui n’indiquent pas à leur université leur race ou leur ethnie ou qui ont été désignés comme ayant «deux races ou plus» ont été exclus de l’analyse (n = 3271). Les étudiants sont classés dans la première génération si aucun des parents / tuteurs n’a obtenu un diplôme universitaire de 4 ans (n = 2255; 14,6%). Si un élève a pris le ACT au lieu du SAT, son composite ACT a été converti en un score SAT. Les étudiants qui n’ont pas fourni à l’université un score SAT ou ACT ont été exclus de l’analyse (n = 440).

Notes de cours

Les notes de cours ont été obtenues à partir des archives universitaires de tous les étudiants (NOT = 15 466) dans tous les cours enseignés par les membres du corps professoral de notre échantillon pendant sept semestres (2 ans) précédant l’enquête auprès du corps professoral. Les notes ont été fournies sur une échelle de 4,0 (A / A + = 4,0, A− = 3,7, B ​​+ = 3,3, B = 3,0, B− = 2,7, C + = 2,3, C = 2,0, C− = 1,7, D + = 1,3, D = 1,0, D− = 0,7, F = 0,0).

Variables de niveau de cours

Les archives de l’Université ont fourni les caractéristiques du cours, telles que le nombre d’étudiants inscrits à chaque cours et le niveau du cours (niveau 100, 200, 300 ou 400). Un cours de niveau 100 est généralement un cours d&#39;introduction, alors qu&#39;un cours de niveau 400 est généralement un cours plus avancé. Sur les 634 cours inclus dans l&#39;échantillon, 24,0% étaient du niveau 100, 23,3% du niveau 200, 31,7% du niveau 300 et 21,0% du niveau 400.

Évaluations de cours

Les réponses des étudiants à l’évaluation moyenne des cours sur quatre semestres pour tous les cours enseignés par la faculté de notre échantillon ont été recueillies à partir de données universitaires. Les évaluations de cours dans cette université ont été normalisées pour tous les cours et destinées à être utilisées pour le développement du corps professoral (pour aider les professeurs à améliorer leur enseignement) et pour les décisions de titularisation et de promotion. À la fin du semestre, les étudiants ont répondu à deux questions concernant les pratiques pédagogiques du professeur (c.-à-d. "Dans quelle mesure l&#39;instructeur vous a-t-il motivé pour faire de votre mieux?" Et "Dans quelle mesure l&#39;instructeur a-t-il mis l&#39;accent sur l&#39;apprentissage et le développement des étudiants?") Et une question concernant leur recommandation générale par l&#39;instructeur (c.-à-d. «Quelle est la probabilité que vous recommandiez ce cours avec cet instructeur?») sur une échelle de 1 (pas du tout) à 4 (très / très probable). Les élèves ont répondu à une question concernant le temps requis pour le cours (c.-à-d. «Comparé aux autres cours que vous avez suivis, combien de temps a duré ce cours?») Sur une échelle de 1 (beaucoup moins de temps) à 5 (beaucoup plus de temps) . Des questions d&#39;évaluation supplémentaires ont été posées aux étudiants par l&#39;université. toutefois, seules les questions d&#39;évaluation mentionnées ci-dessus étaient accessibles en ligne au public; par conséquent, nos analyses ont été limitées à ces quatre questions. Les cours comptant moins de cinq étudiants inscrits n&#39;ont pas été inclus dans les analyses pour s&#39;assurer que les résultats n&#39;étaient pas biaisés par les faibles taux de réponse. Les réponses au niveau des étudiants n&#39;étaient pas disponibles en raison de problèmes de confidentialité; Pour cette raison, nous n’avons pas été en mesure d’examiner les différences raciales / ethniques entre les expériences des élèves en classe.

Modèles hiérarchiques

Nous avons utilisé la modélisation linéaire hiérarchique pour prendre en compte la structure imbriquée des données (17). Pour examiner les facteurs qui affectent les notes des cours, nous avons testé un modèle à trois niveaux dans lequel les étudiants (niveau 1) étaient imbriqués dans les cours (niveau 2) et les cours au sein du corps professoral (niveau 3). Le modèle incluait des effets aléatoires partiellement croisés car les étudiants pouvaient suivre des cours de plusieurs membres du corps professoral (19). Dans le modèle, nous avons contrôlé toutes les caractéristiques des étudiants disponibles (sexe, race / ethnie, statut de première génération et scores au SAT), toutes les caractéristiques des cours disponibles (inscription au cours et trois variables nominales prenant en compte le niveau du cours) et tous les professeurs disponibles. caractéristiques (sexe, race / ethnie, âge, années d’enseignement et statut d’occupation). Voir les tableaux S4 à S6 pour les corrélations entre les variables à chaque niveau. Les données manquantes ont été traitées par suppression liste par liste. La pente de la race / ethnie des étudiants a été autorisée à varier d’un cours à l’autre pour estimer l’interaction croisée entre l’esprit des professeurs et la race / ethnie des étudiants. Le coefficient de corrélation intraclasse (CCI) pour la section de cours (niveau 2) était de 0,06, ce qui indique que les sections de cours représentaient 6% de la variance des notes des étudiants. Le CCI pour le corps professoral (niveau 3) était de 0,09, ce qui indique que les professeurs représentaient 9% de la variance dans les notes des étudiants. Le modèle a été installé à l’aide du package lme4 (34) for R version 3.3.1 (35) using restricted maximum likelihood. We used the lmerTest package to obtain P values for fixed effects (36). T tests used the Satterthwaite approximations to degrees of freedom. All continuous variables were standardized. Categorical variables were coded as follows: female = 1, male = 0; URM (Black, Hispanic, Native American) = 1, non-URM (White, Asian) = 0; first-generation = 1, continuing-generation = 0; tenured = 1, nontenured = 0. We added three dummy codes to control for course level, with level 100 as the reference group (i.e., level 200 = 1 and level 100 = 0). Specifically, we estimated a model using the following R code, which was adapted from Bates et al. (34)

M1 <- lmer(Student_Course_Grade ~ Faculty_Mindset*Student_Race

+ Student_Firstgeneration + Student_Gender + Student_SAT

+ Faculty_Gender + Faculty_Teaching_Experience + Faculty_Tenure_Status

+ Faculty_Age + Faculty_Race

+ Course_Enrollment + Course_200Level + Course_300Level + Course_400Level

+ (1 | Student_ID) + (Student_Race |Faculty_ID/Course_ID)

To examine average course evaluations, we tested a two-level model in which courses (level 1) were nested within faculty (level 2). In this model, we controlled for the same course characteristics (course enrollment and three dummy variables that account for class level) and faculty characteristics (gender, race/ethnicity, age, years of teaching experience, and tenure status) as the previous model. The ICC for faculty (level 2) ranged from 0.51 to 0.60, depending on the question, indicating that faculty accounted for approximately 51 to 60% of the variance in students’ course evaluation responses. The following R code was used to estimate the models:

M2 <- lmer(Course_Evaluations ~ Faculty_Mindset

+ Faculty_Gender + Faculty_Teaching_Experience + Faculty_Tenure_Status

+ Faculty_Age + Faculty_Race

+ Course_Enrollment + Course_200Level + Course_300Level + Course_400Level

+ (1|Faculty_ID)


Supplementary material for this article is available at

Supplemental Analyses

Table S1. Fixed effects estimates predicting students’ grades in STEM courses.

Table S2. Testing the role of other faculty characteristics.

Table S3. Fixed effects estimates predicting course evaluations.

Table S4. Correlations among the variables at level 1 (student).

Table S5. Correlations among the variables at level 2 (course).

Table S6. Correlations among the variables at level 3 (faculty).

Table S7. Discipline-level mindset beliefs.

Fig. S1. Mediation models for URM and non-URM students.

References (3739)

This is an open-access article distributed under the terms of the Creative Commons Attribution-NonCommercial license, which permits use, distribution, and reproduction in any medium, so long as the resultant use is do not for commercial advantage and provided the original work is properly cited.


  1. National Science Board, Science and Engineering Indicators 2018 (National Science Foundation, 2018).

  2. M. C. Murphy, V. J. Taylor, The role of situational cues in signaling and maintaining stereotype threat, in Stereotype Threat: Theory, Process, and Application, M. Inzlicht, T. Schmader, Eds. (Oxford Univ. Press, 2011).

  3. C. S. Dweck, Self-theories: Their Role in Motivation, Personality, and Development (Psychology Press, 1999).

  4. R. Rosenthal, L. Jacobson, Pygmalion in the Classroom: Teacher Expectation and Pupils’ Intellectual Development (Holt, Rinehart and Winston, 1968).

  5. L. Jussim, J. Eccles, S. Madon, Social perception, social stereotypes, and teacher expectations: Accuracy and the quest for the powerful self-fulfilling prophecy, in Advances in Experimental Social Psychology (Academic Press, 1996), vol. 28, pp. 281–388.

  6. S. W. Raudenbush, A. S. Bryk, Hierarchical Linear Models: Applications and Data Analysis Methods (Sage Publications, 2002).

  7. D. M. Bates, Lme4: Mixed-Effects Modeling With R (Springer, 2010).

  8. D. M. Steele, B. Cohn-Vargas, Identity Safe Classrooms: Places to Belong and Learn (Corwin Press, ed. 1, 2013).

  9. R Core Team, R: A Language and Environment for Statistical Computing (R Foundation for Statistical Computing, 2016).

  10. National Science Board, Science and Engineering Indicators 2014 (National Science Foundation, 2014).

  11. L. K. Muthén, B. O. Muthén, Mplus User’s Guide (Muthén & Muthén, ed. 8, 1998–2017).

Acknowledgments: We thank N. Yel for statistical support throughout this project and the members of the Mind and Identity in Context Lab, Indiana University. Funding: This work was supported by NSF grants DRL-1450755 and HRD-1661004 awarded to M.C.M. and a Russell Sage Foundation grant (87-15-02) awarded to M.C.M. Author contributions: All authors designed the research and collected the data. E.A.C. analyzed the data. E.A.C. and M.C.M. wrote the manuscript with input from K.M. and D.J.G. Competing interests: The authors declare that they have no competing interests. Data and materials availability: All de-identified data, code, and materials are available upon request and by IRB approval. In compliance with IRB policies, group characteristics will only be shared when there are 10 or more individuals within the group to preserve participants’ anonymity. Additional data related to this paper may be requested from the authors.

Source link