Do Performance Goals and Fixed Mindset Explicate the Relations Between Stereotype Threat and Achievement? Examining Differences Between Racially Marginalized and White Students in STEM

Abstract

This study examined whether performance goal orientations and mindset beliefs explicate the negative relation of ethnic stereotype threat with achievement and whether these processes vary depending on students’ membership in a historically minoritized group. Multigroup analyses of undergraduate chemistry students (N = 1,376) indicated that perceived ethnic stereotype threat was associated with lower achievement regardless of whether students were from underrepresented minority groups (URM). For URM students, compared to White students, ethnic stereotype threat more strongly predicted performance-avoidance goals. Further, fixed mindset beliefs moderated the relation of ethnic stereotype threat with performance goals for White students only. The relations of stereotype threat with performance goals were stronger for White students with a greater fixed mindset. Findings imply that while ethnic stereotype threat has the potential to detrimentally impact both URM and White students, motivational beliefs may exert somewhat distinct influences within each group, shaping the outcomes of stereotype threat.

Educational Relevance

In our society, there are pervasive stereotypes about who can achieve success in competitive fields. Science learning contexts are no exception. In these contexts, and particularly among racially-marginalized students, anxieties about performing poorly and confirming stereotypes are prevalent, resulting in lower achievement for this student group (i.e., stereotype threat). In this study, we examined the consequences of ethnic stereotype threat for racially-marginalized and White students separately. Results revealed that ethnic stereotype threat led to lower science grades for both racially-marginalized and White students. For racially-marginalized students, ethnic stereotype threat also steered them toward goals focused less on growth and learning and more on avoiding looking incompetent to others (i.e., performance-avoidance goals). Further, for White students, a stronger belief that intelligence is fixed and cannot be improved (i.e., fixed mindset belief) exacerbated the negative effects of stereotype threat on their goal orientations, leading them to prioritize appearing intelligent and avoiding seeming incompetent to others. This research highlights the impact of ethnic stereotype threat on marginalized and non-marginalized students’ motivational beliefs and achievement.

Despite efforts to increase the persistence and achievement of students from underrepresented racial minority (URM) groups in STEM majors, enduring differences in performance and persistence suggest that universities are continuing to fail URM students (Musu-Gillette et al., 2016). According to the National Center for Science and Engineering Statistics (NCSES, 2019), one-third of URM students entered college with an interest in studying STEM, but only 16% completed a bachelor’s degree in a STEM field within six years. These statistics are evidence of continued disparities in the American higher education system and highlight the need to understand URM students’ experiences in education and identify levers of change to promote equity.

Decades of research on stereotype threat suggests that in situations when one’s performance is being evaluated (e.g., in school classrooms) students of marginalized groups experience anxiety about either being judged stereotypically or performing poorly and thus inadvertently confirming stereotypes about their group (Aronson et al., 2002; Aronson & Steele, 2005; Shapiro & Neuberg, 2007; Steele & Aronson, 1995). This anxiety interferes with marginalized students’ performance, leading to racial equity gaps (Nguyen & Ryan, 2008; Spencer et al., 1999). It is crucial for researchers and practitioners to go beyond documenting the existence of this obstacle to students’ success to understanding the processes underlying stereotype threat’s negative impacts. This serves as one important part of a multipronged approach to improve the educational environment and promote equity.

Research suggests that motivation-related processes can explain how stereotype threat leads to depressed performance (e.g., Aronson et al., 2002; Ryan & Ryan, 2005). Nevertheless, a precise understanding of the role of these processes remains unclear. This gap may be because the relation of stereotype threat with achievement is likely multifaceted with several moderating and mediating factors (Major & O’Brien, 2005). Therefore, in this study, we examined motivational moderators and mediators that potentially elucidate the ethnic stereotype threat-achievement relation. Specifically, we investigated whether the association of ethnic stereotype threat with achievement is explained by different types of performance goal orientations (i.e., whether students pursue goals related to demonstrating –or avoiding not demonstrating– their competence; Urdan & Kaplan, 2020), and whether the strength of these associations would depend on students’ mindset about intelligence (i.e., beliefs about whether intelligence is fixed; Dweck, 1999). In addition, we used a multigroup approach to examine how these relations differed depending on students’ membership in a URM group. Examining these dynamic relations will increase our understanding of the persisting racial equity gaps and shed more light on potential strategies to support URM students’ journeys in competitive STEM disciplines.

Stereotype Threat

Stereotype threat occurs when an achievement situation (e.g., a classroom) makes individuals aware of stereotypes of their group and fearful that they may confirm those stereotypes through low performance (Steele & Aronson, 1995; Aronson & Steele, 2005). There are also several situational factors that make provocation of stereotype threat more likely, including testing situations in which a stereotyped identity (i.e., race/ethnicity or gender) is made salient (Steele & Aronson, 1995) and settings where one is in the presence of positively stereotyped peers (Inzlicht & Ben-Zeev, 2000). Both individual and situational factors are clearly present for URM undergraduate science students at predominantly White institutions (PWIs), putting them at high risk of experiencing ethnic stereotype threat. However, it is also possible for White students to experience ethnic stereotype threat. For example, White students in STEM fields may experience concerns about unfavorable comparisons to other groups about whom there are positive stereotypes in STEM (e.g., Asian men; Aronson et al., 1999). Indeed, prior research has shown that White male students in STEM experience group-based performance anxiety, but at lower levels than their URM and female peers (Beasley & Fischer, 2012).

Accordingly, it is worthwhile to examine how ethnic stereotype threat may differentially influence the motivational processes of URM and White students.¹ In his early theorizing about stereotype threat, Steele (1997) posited that chronicity of stereotype threat would lead to different long-term psychological consequences, such as domain disengagement and disidentification. Students from minoritized racial and ethnic backgrounds are more likely than White students to experience chronic stereotype threat. Thus, URM students’ experiences with stereotype threat is qualitatively different than those of White students (Simon et al., 2013) and stereotype threat can have more severe consequences on motivation and achievement (i.e., larger effect sizes; Sherman et al., 2013) for URM students than for White students (Taylor & Walton, 2011).

Processes Underlying the Stereotype Threat-Achievement Relation

A variety of cognitive and emotional factors have been studied as processes in the negative relation of stereotype threat with achievement (see Pennington et al., 2016 and Spencer et al., 2016 for a review). For instance, a great body of research suggests that stereotype threat interferes with individuals’ achievement by undermining their working memory capacity (e.g., Schmader, 2010), increasing thought suppressions (e.g., Logel et al., 2009), and heightening their anxieties (e.g., Bosson et al., 2004; Laurin, 2013). More recent research also highlights the important role that a variety of social-cognitive factors such as motivational beliefs play in explaining the relation of stereotype threat with achievement (Cimpian et al., 2012; Deemer et al., 2013). This research suggests that, first, motivation may serve as a mediator (i.e., stereotype threat shapes motivation, which in turn, predicts performance); and second, motivation may serve as a moderator (i.e., the impact of stereotype threat on achievement depends on levels of motivation). Although a great number of motivational processes have been examined in stereotype threat literature (e.g., expectations for success, Hess et al., 2009; identification with and valuing of the domain, Smith et al., 2015), we place the focus of our examinations on performance goal orientation (e.g., Brodish & Devine, 2009) and fixed mindset beliefs (e.g., Good et al., 2012).

Researchers theorize that performance goal orientations (i.e., whether students are oriented towards demonstrating or avoiding not demonstrating competence) may mediate the relation of stereotype threat with achievement (e.g., Brodish & Devine, 2009; Smith, 2004). Moreover, fixed mindsets (i.e., implicit theories about whether intelligence is fixed; Dweck, 1999) may enhance or weaken the relation between stereotype threat and achievement (e.g., Cimpian et al., 2012; McKown et al., 2010). Most research, however, examined these constructs separately, which limits understanding of the interplay among these processes. Therefore, we examined these constructs concurrently by investigating performance goal orientations as mediators and mindset as a moderator of the relation between stereotype threat and achievement for both URM and White students.

Performance Goal Orientations as Potential Mediators

According to achievement goal theory, individuals have distinct goal orientations that provide them with a broad aim or purpose in achievement situations (Ames, 1992; Dweck & Leggett, 1988; Maehr & Midgley, 1996; Nicholls, 1984). This aim is used as an organizing schema that influences how one evaluates their performance in achievement situations, and in turn their beliefs about their level of competence and success. Achievement goal theory scholars commonly use a trichotomous achievement goal orientation model that distinguishes three types of goal orientations: mastery-approach, performance-approach, and performance-avoidance (Midgley et al., 2000; Scherrer et al., 2020; Urdan & Kaplan, 2020).

According to goal orientation theory (e.g., Ames, 1992; Elliot et al., 2011), a mastery-approach goal orientation is characterized by a focus on improving competence and learning for its own sake, and therefore is focused on intrapersonal comparison (i.e., how much progress one has made over time). In contrast, performance goals are characterized by either focusing on performing well to demonstrate to others that one has competence (performance-approach) or focusing on avoiding performing worse than others to avoid being labeled incompetent (performance-avoidance).²

Smith (2004) argues that performance goal orientations may be the missing piece in understanding the psychological mechanisms through which stereotype threat undermines achievement. The author explains that when one feels like their ability is being questioned, a common experience when one is threatened by ability-based stereotypes, one may pursue goals that assist them in looking competent (performance-approach goals) or prevent them from looking incompetent (performance-avoidance goals). On the other hand, since mastery-approach goals are focused on one’s desires to improve and learn for their own sake, experiencing stereotype threat, where interpersonal standards of competence are emphasized, may have limited influence on the development of mastery goal orientations.

Smith (2004) proposes an integrative model named STEP (stereotyped task engagement process) where she outlines the potential role that performance goals play in mediating the stereotype threat-achievement relation. In her model, stereotype threat is theorized to provoke performance goals (particularly performance-avoidance goals), because it motivates individuals to prove that competence-based stereotypes are untrue. Based on this model, stereotype threat may lead to the adoption of performance goals orientations, which would in turn, lead the individual to self-regulate their achievement behavior in a manner that results in negative outcomes.

Indeed, empirical research suggests that stereotype threat is positively related to performance-avoidance goals (see Thoman et al., 2013, for a review), supporting the notion that having a performance-avoidance goal orientation causes one to appraise tasks as threats rather than challenges (McGregor & Elliot, 2002). Additionally, prior research found that women under stereotype threat conditions in a quantitative field endorsed higher performance-avoidance goals compared to those in no-threat conditions (Chalabaev et al., 2008; Smith et al., 2007), which in turn led to a decline in performance (Brodish & Devine; 2009; Smith, 2006). The relation of stereotype threat with performance-approach goal orientation is not well documented, however, research suggests that this goal orientation tends to be associated with negative outcomes when the primary aim is framed as a goal to demonstrate or appear competent rather than outperform others (Hulleman et al., 2010).

It is also important to consider whether the relation of performance goals to other constructs, such as stereotype threat, varies based on individuals’ identification with a racially stereotyped group. Agreeing with the premise that stereotype threat is more consequential for racially stereotyped groups than non-stereotyped groups (Steele & Aronson, 1995; Spencer et al., 1999), we would then expect to see stronger negative relations of stereotype threat with motivation and achievement for URM students compared to White students. For URM students, being in competitive academic environments where their racial group is starkly underrepresented (e.g., NCSES, 2019) and their competence and intellectual ability is consistently questioned (e.g., Park et al., 2020), could increase pressure on avoiding appearing incompetent and hence, increase performance-avoidance goals. These relations may exist with a lesser strength for White students. For White students, although concerns about being compared to higher-performing student groups exists (e.g., Asian-Americans; Aronson et al., 1999), the well-representation of successful White role models in the field and the absence of historical race-based scrutiny may weaken any negative consequences that comparison-based stereotype threat has with achievement goals. Indeed, research suggests that achievement situations that emphasize performance goals (normative comparison) may be particularly harmful for Black students as they may have a greater need to engage in self-protective behaviors, such as self-handicapping, due to historical stigmatization experienced in school (Gutman, 2006). The current study seeks to extend this literature by investigating whether the relations of ethnic stereotype threat, performance goal orientations, and achievement vary between URM and White students. As such, this study aims to better understand how students’ ethnic identities affect their achievement goals, which has been highlighted as a need within achievement goal research (Urdan & Kaplan, 2020).

Mindset Beliefs as Potential Moderators

Performance goals were originally conceptualized as arising from students’ mindsets (Dweck & Leggett, 1988), so it is important to account for mindsets to fully understand how stereotype threat may shape students’ performance goal pursuit and in turn, their achievement. Furthermore, prior research suggests that students’ mindset beliefs may be shaped by their sociocultural backgrounds (Hwang et al., 2019), thus, these beliefs may be malleable to stereotype threat experiences.

Mindset theory proposes that when a person has a fixed mindset, they believe that performance is pre-determined by their stable, unchangeable level of intelligence (Dweck, 1999). Students with a fixed mindset are hypothesized to be more likely to endorse performance goals, exert less effort, and have lower persistence on difficult tasks (Dweck & Leggett, 1988). There is limited prior research linking mindsets, achievement goals, and stereotype threat, but there are some notable studies related to the potential relations among these constructs. For instance, Dweck (2008) surmised that negative stereotypes about the ability of one’s group are exacerbated by having a fixed mindset because the stereotype itself implies that ability cannot change. This assertion is supported by research that documented interventions that increase students’ growth mindset (i.e., beliefs that intelligence is malleable) are effective for addressing the negative impacts of stereotype threat on achievement (e.g., Aronson et al., 2002; Good et al., 2012). Furthermore, this research shows that college students’ mindsets change in response to interventions, so it is relevant to examine even though some of their beliefs may be more stable at this age (Aronson et al., 2002; Bråten et al., 2017; Shively & Ryan, 2013). Prior research has also demonstrated that interventions to promote growth mindset are most helpful for those facing challenges (e.g., ego threat; Burnette et al., 2013 meta-analysis) and who have a history of lower achievement (Yeager et al., 2019 national study), conditions that may be more common for URM students. Taken together, this literature suggests that mindsets may moderate the relation of stereotype threat with achievement, and that there is a need to conduct additional research to see if there are differences depending on students’ racial identities.

The Present Study

Despite the theorized importance of performance goals in explaining the relation of stereotype threat and achievement, and the theoretical expectation that this relation could be enhanced by having a fixed mindset (Dweck & Leggett, 1988; Ryan & Ryan, 2005), to our knowledge no research has explored the differential mechanistic role of these motivational beliefs for URM and White students. By using a multigroup approach paired with a comprehensive model that specifies hypothesized interrelations of students’ mindsets, ethnic stereotype threat, and performance goal orientations, the present study provides new understanding of the precise factors that could explain achievement and persistence disparities in STEM fields. Furthermore, by examining how students’ racial and ethnic identities are related to differences in their motivational beliefs, this study contributes to the scarce amount of race-focused motivation research (e.g., Kumar & DeCuir-Gunby, 2023; López, 2022; Urdan & Kaplan, 2020). Our approach has the potential to inform interventions to increase STEM students’ persistence and achievement and remove barriers to equity gaps between URM and White students. Accordingly, we addressed two main goals in our study.

Aim 1: Investigate the Role of Performance Goals in Mediating the Relations of Ethnic Stereotype Threat with Achievement for URM versus White Students

First, we expected that ethnic stereotype threat would negatively relate to achievement for both URM and White students but that this overall relation would be stronger for URM students. Further we expected that ethnic stereotype threat would be positively related to performance-avoidance goal orientation as this was found in prior research (Brodish & Devine; 2009; Smith, 2006). The relation between ethnic stereotype threat and performance-approach goal orientation is less clear, however, it is theorized that when performance-approach goal orientation is focused on demonstrating competence (i.e., appearance) rather than outperforming others (i.e., competitive), this orientation negatively relates to achievement (Hulleman et al., 2010; Maehr & Zusho, 2009). As situations that trigger stereotype threat are likely to make one concerned with appearing competent, we hypothesized that ethnic stereotype threat would be positively associated with performance-approach goal orientation. Importantly, given the theoretical expectation that ethnic stereotype threat is more consequential and experienced more chronically among URM students than White students (e.g., Steele & Aronson, 1995; Spencer et al., 1999), we expected that the relations of ethnic stereotype threat with achievement, performance-avoidance, and performance-approach would be stronger for URM versus White students,

Given the extant research that suggests achievement goals are key determinants of achievement outcomes (e.g., Maehr & Zusho, 2009; Urdan & Kaplan, 2020) and the theoretical expectation that performance goals could serve as mechanisms for the negative relation between stereotype threat and achievement (Smith, 2004), we expected that performance-avoidance and performance-approach goals would significantly mediate the relations between ethnic stereotype threat and achievement for White and URM students.

Aim 2: Investigate the Role of Fixed Mindset in Moderating the Relation of Ethnic Stereotype Threat with Performance Goals and Achievement for URM versus White Students

Theory and research suggest that a fixed mindset is positively related to students’ endorsement of performance-approach and performance-avoidance goal orientations (e.g., Blackwell et al., 2007; Cury et al., 2006; Dweck & Leggett, 1988) and is negatively related to performance (e.g., Blackwell et al., 2007; Grant & Dweck, 2003). Further, researchers theorize that a fixed mindset could exacerbate the consequences of the negative stereotypes that question the ability of individuals (Dweck, 2008). Therefore, we expected that a fixed mindset would moderate the relations of ethnic stereotype threat with performance-avoidance, performance-approach goals, and achievement. That is, we expected that the deleterious effect of perceived ethnic stereotype threat would be stronger when students viewed intelligence as fixed, and that these relations would be stronger for URM students due to the greater likelihood of experiencing chronic stereotype threat (Simon et al., 2013; Steele, 1997).

Method

Participants and Procedures

Participants were recruited as part of the larger study investigating persistence in STEM fields. We used a sample of 1,367 undergraduate students attending a large, public university located in the United States. The sample was majority female (57.0%), the mean age was 18.65 (SD = 1.72), and 16.0% were first-generation college students. Approximately, 87.0% of the participants were White-American, 6.7% were African-American, 4.0% were Hispanic or Latinx, 0.5% were American Indian or Alaska Native, and 1.8% of the participants selected multiple races (at least one of which were African-American, Hispanic, or Native-American). For the purpose of this study, a dichotomous URM variable was created to distinguish between URM and White students.³ URM students included African-American, Hispanic or Latinx, American Indian or Alaska Native students, as well as multi-racial students who selected at least one of these races/ethnicities (13.0%). We chose to exclude Asian students from this study because they are not underrepresented minorities in STEM fields (National Center for Science and Engineering Statistics, 2014), but are in the numeric minority at this university and within our sample (6.8% of students; 14% of sample). Additionally, since this racial group has historically experienced model minority stereotypes within STEM and other education disciplines (e.g., McGee et al., 2017; Trytten et al., 2013), we refrained from grouping them with White students. Table 1 includes sample descriptive statistics for URM and White students.

Table 1.

Descriptive Statistics and Correlations for the URM and White Student Samples Separately

	1	2	3	4	5	6
1. Stereotype threat	–	.33**	.12**	.08**	−.10**	−.10**
2. Fixed mindset beliefs	.17*	–	.17**	.14**	−.03	−.05
3. Performance-approach	.24**	.20**	–	.80**	.04	−.10**
4. Performance-avoidance	−.23**	.23**	.81**	–	−.02	.003
5. Final exam grades	−.29**	−.09	−.03	−.10	–	−.08**
6. Gender	−.002	.06	−.10	−.08	−.01	–
M (SD)
White Students	2.05 (.76)	2.66 (1.08)	3.00 (.85)	3.06 (.82)	131.37 (26.68)	–
URM Students	2.95 (1.09)	2.32 (1.06)	2.80 (.88)	2.94 (.96)	116.23 (30.66)
α
White Students	.95	.92	.90	.80	–	–
URM Students	.95	.89	.87	.83
N
White Students	1184	1185	1187	1189	1173
URM Students	178	178	177	178	174	178

Note. The bottom part of the table, with unhighlighted results represent the descriptive statistics for URM and White students separately. For correlations, the values below the diagonal represent correlations for the URM student sample and the values above the diagonal represent correlations for the White student sample.

Stereotype threat and performance goal orientations were rated on a 5-point Likert scale (1 = strongly disagree, 5 = strongly agree). Mindset beliefs were rated on a 6-point Likert scale (1 = strongly disagree, 3 = mostly disagree, 6 = strongly agree). Final exam scores were on a 0 – 180 scale.

^*p < .05

^**p < .01

All participants were enrolled in an introductory chemistry course, which was a requirement for many natural sciences and some engineering majors. In week 8 of the semester, students completed surveys related to their motivational beliefs and experiences in science. All students who completed the surveys received course credit as compensation. Data are only reported for those students who consented to participate in the research study. Data were collected in fall 2016. The treatment of participants in this study was conducted in accordance with the ethical guidelines set forth by the American Psychological Association. This study was approved as exempt by the university’s Institutional Review Board at the last author’s institution.

Measures

Measures included demographic questions and scales to assess ethnic stereotype threat, fixed mindset, and performance goal orientations. See Table 1 for the reliability information and the Appendix for the survey items.

Ethnic Stereotype threat.

An eight-item scale developed by Steele, James, and Barnett (2002; e.g., “If you do poorly on a test, people will assume that it is because of your ethnicity.”) assessed students’ stereotype threat. This measure has been used in prior literature to assess URM students’ experiences in science domains (Woodcock et al., 2012).

Fixed mindset.

Students’ fixed mindset beliefs were assessed using the theories of intelligence scale developed by Dweck (1999). This scale includes four items that assess fixed mindset beliefs (e.g., “You have a certain amount of intelligence, and you can’t really do much to change it.”). Although Dweck originally conceptualized entity (fixed) and incremental (growth) mindsets along a continuum, prior research found that when both fixed mindset beliefs and growth mindset beliefs were measured, they did not form a single factor but rather represented two distinct constructs (Cho et al., 2021; Dai & Cromley, 2014). As such, we refer to this scale as “fixed mindsets” and do not assume that low levels of fixed mindsets are equivalent to endorsing a growth mindset.

Performance goal orientations.

The Patterns of Adaptive Learning Scales (Midgley et al., 2000) was used to assess students’ performance goal orientations towards science. The performance-approach (PAP) subscale included 5 items measuring students’ orientation toward appearing smart and demonstrating competence to others (e.g., “It’s important to me that other students think I am good at science”). The performance-avoidance (PAV) subscale included 4 items measuring students’ orientation toward avoiding looking incompetent (e.g., “It’s important to me that I don’t look stupid in science”).

Achievement.

Students’ chemistry final exam grades were obtained from institutional records and used as indicators of academic achievement. Final exam grades were on a scale from 0 – 180 points.

Results

Preliminary Analyses

We first examined descriptive statistics and correlations using SPSS (v.24). Next, we conducted a multigroup confirmatory factor analysis to test the construct validity of measures for URM and White students. This is an important step in understanding whether the meaning of instruments is similar across racial groups. Model fit was examined via the Comparative Fit Index (CFI) and Root Mean Square Error of Approximation (RMSEA). Using Hu and Bentler’s (1999) guidelines, CFI > .90 and RMSEA < .08 indicates adequate fit, and CFI > .95 and RMSEA < .06 indicates excellent fit. We also tested for measurement invariance across racial groups to ensure that the variables retained their structural validity across URM and White students. Multigroup measurement invariance was established if the change in CFI was ≤ 0.01 (Cheung & Rensvold, 2002; Meade et al., 2008). For the purposes of this study, it was important to establish at least weak measurement invariance as this is the minimum threshold for comparing regression coefficients across different groups (Meredith, 1993; Widaman & Reise, 1997). Multigroup confirmatory factor analysis and measurement invariance tests were conducted using Mplus version 8.6 (Muthén & Muthén, 1998–2021).

Descriptive Statistics

Correlations and means for URM and White students separately are reported in Table 1 (see Table S1 for descriptive statistics for the overall sample).

Multigroup Confirmatory Factor Analyses

To test the construct validity of the measures for both the URM and White samples, we performed multigroup confirmatory factor analysis. One model with ethnic stereotype threat, fixed mindset beliefs, performance-approach goals, and performance-avoidance goals was tested. Results suggested excellent fit, χ² (400) = 1035.41, CFI = .96, TLI = .95, RMSEA = .05, SRMR = .04, therefore, we concluded that our constructs had acceptable construct validity.

Multigroup Measurement Invariance

We performed a multigroup measurement invariance test on ethnic stereotype threat, fixed mindset beliefs, and performance goals to examine whether the same constructs are being measured across URM and White students (Kline, 2015; Vandenberg & Lance, 2000). Because multigroup structural equation modeling (SEM) analyses involve comparing groups on path coefficients among variables, at least weak (or metric) measurement invariance was required for this study to ensure the constructs had the same meaning across groups (Meredith, 1993; Widaman & Reise, 1997). Therefore, we tested configural and metric invariance for our group of constructs. Model fit indices of the configural model, χ² (366) = 855.91, CFI = .966, TLI = .961, RMSEA = .045, SRMR = .033, and metric model χ² (383) = 880.41, CFI = .966, TLI = .963, RMSEA = .044, SRMR = .035 suggested excellent fit. Also, the change in CFI from configural to metric invariance models (< .001) was negligible, suggesting metric measurement invariance across groups.

Main Analyses

Due to the high correlation between performance-approach and performance-avoidance goals (rs = .82 and .80 for URM and White students, respectively) and concerns over high multicollinearity between these variables, we ran two separate models, with each model including either performance-approach or performance-avoidance goals. This decision was consistent with recommendations outlined by Linnenbrink-Garcia et al. (2012). Prior research (Kaplan et al., 2009; Ross et al., 2002) has found similarly high correlations between these variables and correlations above .50 are common (see Linnenbrink-Garcia et al., 2012 for a review).

We specified a measurement model within our SEMs. Given the results of our measurement invariance analyses, we assumed equal factor loadings across URM and White student groups in the measurement model. Then, we tested the relations among latent ethnic stereotype threat as the predictor, fixed mindset as the moderator, each of the performance goals as the mediators, and exam grades as the outcome. A latent interaction term was created using Mplus’s default method, Latent Moderated Structural Equations (LMS; Klein & Moosbrugger, 2000). This approach tests interaction effects using full-information maximum-likelihood estimation. Specifically, we multiplied the latent fixed mindset and stereotype threat variables using the XWITH command in Mplus (Muthén & Muthén, 1998–2021). Testing moderation effects with latent interactions offers several benefits, including accounting for measurement model error and increasing the power to detect moderating effects (Lodder et al., 2019). The effects of gender on stereotype threat, fixed mindset, performance goals, and grades were controlled for in both models (see Table S2 in the Supplemental Materials for the results of the relations of gender and other variables). Missing data (< 1% missing at both the item and scale level) were handled using full information maximum likelihood estimation (FIML).

First, we conducted single-group SEMs to explore the relations among ethnic stereotype threat, fixed mindset beliefs, performance goals, and grades for the overall sample (see the Supplementary Materials for the results of these analyses). Next, we conducted multigroup SEMs to examine whether there are any significant differences between URM and White students in the way these variables influence the ethnic stereotype threat-achievement relations. This approach is aligned with the cross-cultural approaches used in prior achievement goal research (Zusho & Clayton, 2011). The analyses were conducted in the following steps (see Jöreskog, 1971 and Sorbom, 1974 for guidelines). First, we allowed all parameters to be freely estimated across the URM and White student groups. This allowed us to obtain the coefficients for the relations among the variables for URM and White students’ separately. These results are described below under each research question and also presented in Figures 1 and 2. We evaluated the model fit for these freely-estimated, unconstrained models. Next, we constrained all parameters to be equal between the two groups. The results suggested the constrained model had a significantly worse model fit compared to the unconstrained model, as reflected by the significant increase in chi square (PAP model: Δχ² (11) = 123.65, p < .001; PAV model: Δχ² (11) = 127.24, p < .001). These results suggested significant overall differences between the URM and White student groups in the relations among the variables of interest. Finally, to locate the origin of difference between the groups, we constrained each path coefficient at a time and compared the fit of the model to that of the unconstrained model. A significantly worse model fit as indicated by significant changes in χ² would suggest a significant difference between the two groups on that path coefficient. Results of the multigroup differences are reported in Table 2 and described under each research question below.

Figure 1. Unstandardized Coefficients for Direct Effects of Stereotype Threat, Fixed Mindset, and Their Interaction on Performance-Approach Goals and Exam Grades.

Note. Bold estimates indicate significant direct paths.

Figure 2. Unstandardized Coefficients for Direct Effects of Stereotype Threat, Fixed Mindset, and Their Interaction on Performance-Avoidance Goals and Exam Grades.

Note. Bold estimates indicate significant direct paths.

Table 2.

Test of multigroup differences between URM and White students on each path.

Paths	χ2 (df)	Δχ2 (Δdf)	AIC	BIC
Performance-approach model
Unconstrained Model	61,928.56 (77)	–	62,082.57	62,484.36
Fully Constrained Model	62,052.22 (66)	123.65 (11) **	62,184.22	62,528.618
Constraining one path at a time
Stereotype → Grades	61,931.05 (76)	2.49 (1)	62,083.05	62,479.63
Fixed → Grades	61,928.71 (76)	0.148 (1)	62,080.71	62,477.29
Stereotype Threat → PAP	61,930.05 (76)	1.49 (1)	62,082.05	62,478.64
Fixed → PAP	61,929.672 (76)	1.11 (1)	62,081.67	63,478.25
ST × FB → PAP	61,931.24 (76)	2.68 (1)	62,083.24	62,479.82
PAP Grades	61,928.57 (76)	0.002 (1)	62,080.28	62,477.15
Performance-avoidance model
Unconstrained Model	60,647.18 (74)	–	60,795.18	61,181.32
Fully Constrained Model	60,774.41 (63)	127.24 (11) **	60,900.41	61,229.16
Constraining one path at a time
Stereotype → Grades	60,649.37 (73)	2.19 (1)	60,795.37	61,176.30
FB → Grades	60,647.29 (73)	0.12 (1)	60,793.29	61,174.22
Stereotype → PAV	60,653.40 (73)	6.23 (1) *	60,799.40	61,180.33
Fixed → PAV	60,648.79 (73)	1.61 (1)	60,794.79	61,175.72
ST × FB → PAV	60,649.54 (73)	2.37 (1)	60,795.64	61,176.47
PAV → Grades	60,647.18 (73)	<0.001 (1)	60,793.18	61,174.11

Note. Results of each model are compared to the results of the Unconstrained Model. Significant changes in χ² (marked in bold) suggest model fit significantly worsened after constraining a path to be equal across groups, thereby indicating a significant multigroup difference on that path.

^*p < .05.

^**p < .01.

Aim 1: Investigating the Relations Among Ethnic Stereotype Threat, Performance Goals, and Achievement for URM and White Students

To address Aim 1, we investigated whether ethnic stereotype threat is related to achievement and performance goals, whether performance goals mediate the relation between ethnic stereotype threat and achievement, and whether such relations are stronger among URM, compared to White, students.

Relation Between Ethnic Stereotype Threat and Achievement

We expected that ethnic stereotype threat would be negatively related to achievement for White students and, to a greater extent, for URM students. As hypothesized, ethnic stereotype threat was significantly and negatively related to achievement (i.e., final exam scores) for URM students (PAP model: β = −0.27, b = −8.44, SE = 2.11, p < .001; PAV model: β = −0.26, b = −8.13, SE = 2.12, p < .001) and White students (PAP model: β = −0.15, b = −4.67, SE = 1.14, p < .001; PAV model: β = −0.14, b = −4.56, SE = 1.14, p < .001), across both performance goal models (see Figures 1 and 2).

Difference Between URM and White Students.

Although the trends of the results suggested that the relation of ethnic stereotype threat with achievement may be stronger for URM students compared to White students, this group difference was not statistically significant (PAP model: Δχ² (1) = 2.49, p > .05; PAV model Δχ² (1) = 2.19, p > .05). See Table 2 for the results of intergroup differences.

Relations Between Ethnic Stereotype Threat and Performance Goals

We expected that ethnic stereotype threat would be positively related to performance-approach and performance-avoidance goals, and that these relations would be stronger for URM students than for White students. Results supported our expectations. We found that for URM students, ethnic stereotype threat was positively related to performance-approach (β = 0.14, b = 0.15, SE = 0.06, p = .019) and performance-avoidance goals (β = 0.19, b = 0.18, SE = 0.06, p = .003). On the other hand, we found that for White students, ethnic stereotype threat did not significantly predict performance-approach goals (β = 0.05, b = 0.06, SE = 0.04, p = .151) or performance-avoidance goals (β = 0.00, b = 0.00, SE = 0.04, p = .957; see Figures 1 and 2).

Difference Between URM and White Students.

Although the trends of results suggested that the relations of ethnic stereotype threat and performance goals may be stronger among URM students, compared to White students, the two groups only significantly differed with regards to the relations of ethnic stereotype threat with performance-avoidance goals (Δχ² (1) = 6.23, p < .05). There was no significant difference between the two groups in the relation of ethnic stereotype threat and performance-approach goals (Δχ² (1) = 1.11, p > .05; see Table 2).

The Mediating Role of Performance Goals in the Ethnic Stereotype Threat-Achievement Relation

We expected that there would be significant indirect effects of ethnic stereotype threat on achievement through performance-approach and performance-avoidance goals. We also hypothesized that these indirect effects would be stronger for URM students than for White students. Contrary to our hypothesis, results suggested no significant indirect effect of ethnic stereotype threat on final exam grades through either of the performance goals for URM students. See Table 3 for the results of the indirect effects. In fact, neither performance goals significantly predicted final exam grades, which could explain the non-significant mediation effects. Since the indirect effects were not significant in either student group, we did not conduct tests to probe the difference between the two groups on these indirect paths.

Table 3.

Unstandardized Indirect Effects of Stereotype Threat and Fixed Mindset Beliefs on Final Exam Grades for the URM and White Student Samples

	URM				White
Indirect Paths	b	SE	p	95% CI	b	SE	p	95% CI
Performance-approach Model
Stereotype –> PAP –> Grades	.23	.38	.543	[−.52, .98]	.09	.08	.286	[−.07, .24]
ST × FB –> PAP –> Grades	−.02	.12	.857	[−.25, .21]	.17	.12	.141	[−.06, .40]
Performance-Avoidance Model
Stereotype –> PAV –> Grades	−.01	.49	.979	[−.97, .94]	.00	.01	.961	[−.01, .01]
ST × FB –> PAV –> Grades	.00	.04	.979	[−.07, .07]	.01	.11	.914	[−.21, .23]

Note. 95% CI = 95% confidence intervals. Stereotype and ST = stereotype threat. Fixed and FB = fixed mindset beliefs. PAP = performance-approach goal orientation. PAV = performance-avoidance goal orientation. Grades = Exam Grades. Significant effects are represented in bold.

Summary of Aim 1 Results

In sum, we found that ethnic stereotype threat was associated with performance-approach and performance-avoidance goals among URM students but not among White students. Furthermore, we found that there were statistically significant differences between URM and White students in the relation of ethnic stereotype threat with performance-avoidance goals, but not in the relation of ethnic stereotype threat with performance-approach goals. Neither type of performance goal mediated the relation of ethnic stereotype threat to achievement for URM students or White students.

Aim 2: Investigating the Moderating Role of Fixed Mindset for URM and White Students

To address Aim 2, we investigated the role of fixed mindset in moderating the relations of ethnic stereotype threat with achievement and performance goals for URM and White students.

Moderating Role of Fixed Mindset the Relation of Ethnic Stereotype Threat with Achievement

We expected that the negative relation between ethnic stereotype threat and achievement would be stronger for students with a greater fixed mindset. We also hypothesized that this moderation effect would be stronger for URM students than for White students. Contrary to our hypothesis, the relation of ethnic stereotype threat and achievement was not significantly moderated by fixed mindset for either URM students (PAP model: β = −0.02, b = −0.52, SE = 2.15, p = .807; PAV model: β = −0.02, b = −0.49, SE = 2.15, p = .818) or White students (PAP model: β = 0.02, b = 0.58 SE = 0.89, p = .511; PAV model: β = 0.02, b = 0.73, SE = 0.89, p = .412) across the two performance goal models (see Figures 1 and 2). Since this moderation effect was not significant in either student group, we did not conduct tests to probe the difference between the two groups on this moderation effect.

Moderating Role of Fixed Mindset in the Relations of Ethnic Stereotype Threat with Performance Goals

We expected that the relation between ethnic stereotype threat and performance goals would be stronger for students with a greater fixed mindset and that these moderation effects would be stronger for URM students. Results partially supported our hypothesis. Consistent with our expectation, we found that for White students, fixed mindset significantly moderated the relations of ethnic stereotype threat with performance-approach goals (β = 0.10, b = 0.11, SE = 0.03, p < .001) and performance-avoidance goals (β = 0.10, b = 0.10, SE = 0.03, p = .001; see Figures 1 and 2).

To better understand these significant interactions, we conducted a series of simple slope tests for White students at high and low values of fixed mindset (+/− 1 standard deviation above and below the mean). As expected, among White students, the relations of ethnic stereotype threat with performance-approach and performance-avoidance goal orientations were stronger for students with high fixed mindset (PAP model: b = 0.17, SE =0 .04, p < .001; PAV model: b = 0.10, SE = 0.04, p = .012) in comparison with students with low fixed mindset (PAP model: b = −0.06, SE = 0.05, p = .312; PAV model: b = −0.10, SE = 0.05, p = .064). See Figure 3 for graphical representations of these interaction effects for performance-approach and performance-avoidance models.

Figure 3. Follow-Up Simple Slopes Tests for the Effects of the Interaction (Stereotype Threat x Fixed Mindset) on Achievement Goals for the White Student Group.

Note. High and low values of fixed mindset are calculated as +/− 1 standard deviation above and below the mean. The values on the Y axis represent the saved factor scores from the measurement invariance models for the achievement goals.

Contrary to our hypothesis, for URM students, fixed mindset did not significantly moderate the effects of ethnic stereotype threat on either of the performance goals (PAP model: b = −0.01, SE = 0.07, p =.851; PAV model: b = −0.01, SE = 0.07, p = .852). Although not specified as a hypothesis, we also tested fixed mindset as a moderator of the indirect effects of stereotype threat on exam grades through achievement goals. We did not find any significant effects. Results of these moderated mediation effects are included in Table 3.

Difference Between URM and White Students.

Even though the moderation effects were significant for White students but non-significant for URM students, the test of intergroup differences suggested no significant difference between the URM and White students on any of the moderated path coefficients (see Table 2).

Summary of Aim 2 Results

In sum, we found that for White students, a high fixed mindset was associated with stronger relations between ethnic stereotype threat and both performance goals. No significant moderation effects were found with the URM student sample. Additionally, there were no interactive effects of fixed mindset and ethnic stereotype threat on final exam grades after controlling for stereotype threat and performance goals for URM or White students.

Discussion

In this study, we examined whether and how motivational beliefs can intervene in the vicious cycle between heightened ethnic stereotype threat and lower achievement, particularly for students from historically marginalized groups. Our findings supported prior research that suggested ethnic stereotype threat may be problematic for students regardless of their racial background (Steele & Aronson, 1995). Nevertheless, we found that ethnic stereotype threat may have more negative consequences for URM students than for White students. These conclusions are substantiated by the stronger negative associations of ethnic stereotype threat with performance-avoidance goals among URM students.

It is notable that the negative relation between ethnic stereotype threat and achievement occurred for both students who are members of negatively stereotyped groups and those who are not. This suggests that White students may not be immune to the deleterious effects of ethnic stereotype threat, a notion supported in prior research (Aronson et al., 1999). We had expected to find stronger stereotype threat-achievement associations among URM students compared to White students. This expectation was based on the hypothesis that URM students may be more vulnerable to negative ethnic stereotypes because these stereotypes are deeply rooted in our society, which leads to chronic stereotype threat (Aronson, 2002). Although the group difference in this association was not statistically significant in our study, we found that the size of this association for URM students was considerably larger than that of White students. It would be plausible that research with a more balanced sample group distribution could detect significant group differences on the association between stereotype threat and achievement.

Our investigation of the relations between ethnic stereotype threat and performance goals showed that this relation depended on the students’ racial group membership. In the case of URM students, higher ethnic stereotype threat was associated with higher endorsement of performance-approach and performance-avoidance goals. The positive relation between ethnic stereotype threat and performance-avoidance goal is aligned with prior research (Brodish & Devine, 2009; Smith, 2006). This research suggests that stereotypically threatening environments can motivate URM students to avoid showing any signs of incompetence, as such signs could be judged by others as evidence supporting negative race-based stereotypes. We also found that performance-approach goals had similar relations with ethnic stereotype threat, which extends prior research. These findings suggest that under high levels of ethnic stereotype threat, URM students may be focused on not only avoiding appearing incompetent (Ryan & Ryan, 2005), but also on demonstrating competence to others as a way of disconfirming negative stereotypes about their racial group. However, it is important to note that the relation of ethnic stereotype threat with performance-approach goals did not differ significantly between URM and White students.

Inconsistent with theoretical expectations and prior empirical research (e.g., Brodish & Devine, 2009; Smith, 2004), we did not find evidence for the mechanistic role of performance goals in the ethnic stereotype threat-achievement relation for either student groups. Also, contrary to the results of past research (Darnon et al., 2009; Hangen et al., 2019; Maehr & Zusho, 2009; Urdan & Kaplan, 2020), we were not able to detect any significant associations between performance goals and achievement in our SEM models. This lack of relationship between performance goals and achievement was also reflected in the simple bivariate correlation results. One explanation for this lack of relation could be that we utilized an assessment of performance goals that was consistent with a broader, schema-based perspective on achievement goal orientations (see Zusho & Maehr, 2009; Urdan & Kaplan, 2020 for discussion). As this measure included items assessing both appearance-based on and normative-based elements of achievement goals, which prior research has found to be negatively and positively correlated with achievement (see Hulleman et al., 2010), it is perhaps not surprising that there was no significant relation between performance-approach goals and achievement. We were, however, somewhat surprised that performance-avoidance goals did not relate negatively to achievement, given prior research suggesting that these goal orientations are fairly consistently related to lower achievement (see Elliot & Hulleman, 2017; Hulleman et al., 2010).

We also found that for White students, fixed mindset strengthened the relations of ethnic stereotype threat with performance goals. When White individuals worry that their performance may be compared to that of a stereotypically high-achieving group (e.g., Asian students), they may be motivated to pursue goals that highlight their own performance. For these students, the belief that their ability is fixed, leaving no room for growth, leads them to interpret an occasional poor performance or mistake as indicative of their inherent lack of ability. For these students, such instances may serve as evidence to others of their perceived lower capabilities compared to students from other racial backgrounds. Therefore, for these students, demonstrating competence and avoiding appearing incompetent may become even more crucial. These speculations align with experimental research indicating that growth mindset interventions buffer students against the adverse effects of stereotype threat on their outcomes, underscoring the moderating role of mindsets (Good et al., 2012).

Contrary to our expectations, the fixed mindset did not significantly moderate the association between ethnic stereotype threat and performance goals for URM students. This unexpected finding contradicts prior research indicating that growth mindset interventions are particularly effective in mitigating the negative impacts of stereotype threat among students from underrepresented groups (e.g., Aronson et al., 2022; Good et al., 2012). We posit that the lack of a significant moderation effect in the URM student group may be attributed to the small sample size, resulting in insufficient statistical power to detect such an effect. Indeed, it is recommended that analyses involving latent interactions include a minimum of 1,000 participants to ensure adequate power (see Nagengast et al., 2011). According to these guidelines, our study had sufficient power to detect interaction effects in the White student sample but lacked the adequate power to detect them in the URM sample. We discuss this limitation in more detail in the subsequent limitations section. Additionally, the results of the multigroup tests revealed no significant differences between White and URM students regarding these effects. This suggests that our inability to detect significance in the URM groups may not necessarily indicate the absence of the effect but rather likely reflects the limitations of our study’s statistical power in the URM group.

The present study highlighted the way that motivational beliefs are related to ethnic stereotype threat. To our knowledge, this is the first study to document the association between ethnic stereotype threat and performance-avoidance and performance-approach goal orientations among URM students specifically, as prior research was focused on gender stereotype threat and used predominantly White samples. Furthermore, we found that ethnic stereotype threat and fixed mindset beliefs interacted to predict performance goals for White students. These findings will need to be replicated in future research, but they point to the potential that some motivational beliefs function differently for students from minoritized racial groups due to their experiences with systemic barriers and chronic stigmatization.

This research points to the need to continue to challenge systemic racism that leads to the stigmatization of URM students within STEM fields (Bell, 1995). Efforts should be invested in creating a welcoming environment that celebrates the uniqueness of students who identify with historically marginalized racial groups. This recommendation is aligned with research that suggests students report higher positive affect, sense of belonging, and academic self-efficacy when they have positive experiences related to their marginalized identities in college (Mendoza-Denton et al., 2002; Totonchi et al., 2022). Even so, such changes can be slow and difficult to make. Thus, in addition to pursuing structural changes, it will also be important to explore interventions that help students cope with the stereotype threats they currently face (McGee, 2020).

Limitations and Future Directions

There are several limitations in this study that should be considered. First, while we tested the relations of motivational variables to achievement at a later time point, we did not use an experimental design, so we were not able to draw causal conclusions. Future research using experimental design could provide important additional information about the potential causal links among these variables. Second, our sample represents students at a predominantly White institution (PWI). Thus, there is a limitation to generalizability of our findings. URM students at Minority Serving Institutions may not experience college in the same way, which may lead to differences in motivational patterns. Further, due to the racial composition of the student body of the university from which our sample was derived, the sample size of URM students was much smaller than White students. The limited sample size of URM students constrained the statistical power of the analyses for this demographic, thereby impacting the robustness of interpretations drawn from the results pertaining to this group. This imbalance in sample sizes, however, reflects the underrepresentation of students of color in real-world academic environments and could shed light on the experiences of these students when they are largely outnumbered by racial majority students. Finally, we examined the hypothesized detrimental role of a fixed mindset on science motivation and achievement, but we did not explore the potential protective effects of a growth mindset on students’ outcomes. Indeed, a popular perspective suggests that fixed and growth mindsets are not simply two sides of the same spectrum; rather, they are distinct dimensions that can coexist within an individual (Lüftenegger & Chen, 2017; Tempelaar et al., 2015). As such, fixed and growth mindsets may play different roles in moderating the effects observed in our study.

Despite these limitations, this study adds to the scant literature that focuses on how minoritized students’ motivation may be influenced by their distinct socio-historical experiences (Usher, 2018). In doing so, we identified several potential areas for future research. First, this study used a variable-centered multigroup approach to examine the relations between ethnic stereotype threat, mindset, and performance goals. A person-oriented approach may be another method of examining the questions posed as part of this study, as it may reveal more detailed information about students’ motivational profiles. Specifically, an interesting future direction may be to examine how students’ social-contextual perceptions, such as bias, stereotyping, and belonging, predict particular motivational profiles that include variables such as mindsets, self-efficacy, performance goals, and interest. Furthermore, future research that examines how URM students’ experiences in education settings may influence their achievement motivation could take a situative approach (Nolen et al., 2015). This approach would help us to understand how changes in individuals’ performance goals and mindset beliefs are related to more specific classroom practices by seeking to understand how these psychological phenomena arise in specific social contexts. Importantly, this type of approach would allow for examination of how students’ experiences pursuing a STEM degree reflect the broader social world – a world that has deeply entrenched negative stereotypes of URM students that reflect a history of racism (Steele & Aronson, 1995). We found evidence that URM students have distinct motivational patterns and future contextualized research could help us to better understand more precisely why this is the case. Finally, in future research, it would be informative to examine both fixed and growth mindsets as moderators of the effects of stereotype threat on goal orientations and achievement. This could help determine whether a growth mindset serves a protective role.

Conclusion

In this study, we used a multigroup approach to investigate whether the negative relation of ethnic stereotype threat with achievement is differentially explained by performance goal orientations depending on students’ mindset beliefs and racial group membership. We found that ethnic stereotype threat was associated with lower achievement for all students. Furthermore, results suggest that for URM students, ethnic stereotype threat explained more variance in goals than a fixed mindset did. This was not the case for White students, for whom fixed mindset was the main predictor of their goals and ethnic stereotype threat was not predictive of their goals. This study suggests that if we aim to narrow equity gaps in science domains, beyond implementing interventions to prevent a fixed mindset in URM students, there is a crucial need to implement structural changes that address systemic racism and reduce experiences of ethnic stereotype threat.

Highlights.

• Stereotype threat predicts science achievement for both racially-marginalized and White students.

• Stereotype threat predicts performance-avoidance goals more strongly among racially-marginalized students than among White students.

• A fixed mindset strengthens the relation between stereotype threat and performance goals among White students.

• The motivational variables explicating the effects of stereotype threat vary based on students’ marginalized status.

Appendix

Stereotype Threat (Steele et al., 2002)

How often do you feel that because of your ethnicity…

1. Some people believe that you have less ability.

2. If you are not better than average, people will assume that you are limited.

3. Professors expect you to do poorly.

4. Professors are less likely to encourage you.

5. You are not fully accepted or included by your program.

6. If you ask a simple question, people will think it is because of your ethnicity.

7. If you do poorly on a test, people will assume that it is because of your ethnicity.

8. People of your ethnicity face unfair evaluations because of their ethnicity.

Fixed Mindset (Entity) Beliefs (Dweck, 1999)

1. You have a certain amount of intelligence, and you can’t really do much to change it.

2. Your intelligence is something about you that you can’t change very much.

3. To be honest, you can’t really change how intelligent you are.

4. You can learn new things, but you can’t really change your basic intelligence.

Performance-Approach Goal Orientation (Midgley et al., 2000)

1. It’s important to me that other students think I am good at science.

2. One of my goals is to show others that I’m good at science.

3. One of my goals is to show others that science is easy for me.

4. One of my goals is to look smart in comparison to the other students in science.

5. It’s important to me that I look smart compared to others in science.

Performance-Avoidance Goal Orientation (Midgley et al., 2000)

1. It’s important to me that I don’t look stupid in science.

2. One of my goals is to keep others from thinking I’m not smart in science.

3. It’s important to me that my professors don’t think that I know less than others about science.

4. One of my goals in science is to avoid looking like I have trouble doing the work.