Moderator | Relationship to role model’s effectiveness |
---|---|
Role model moderators | |
1) Perceived competence | The relationship between the perceived competence of the role model and ingroup students’ motivation had an inverted-U shape: Describing the role model as competent increased student motivation and performance, but only up to a point. When a role model’s competence was exceptional, the role model was often demotivating. This moderator either had no consistent effects on outgroup students or showed a reverse pattern whereby low-competence models had positive effects |
2) Perceived similarity to students | |
2a) Demographic similarity | When the role models belonged to groups that are underrepresented in STEM (e.g., women, Black people), they often had positive effects for all students, regardless of demographic similarity. In contrast, majority-group models (e.g., men, White people) did not motivate students from underrepresented groups and were sometimes demotivating |
2b) Psychological similarity | Characteristics that increased the role model’s psychological similarity to students (e.g., widely shared characteristics such as a preference for spending time with close others; characteristics that contradicted common stereotypes of scientists) generally had positive effects on student motivation. Prompting students to reflect on their similarity to the role models was also sometimes effective |
3) Success is perceived as attainable | Role model features and behaviors that increased the perceived attainability of the role model’s success (e.g., “demystifying” STEM careers for students by describing the responsibilities that come with such jobs and the concrete steps to pursue them) were positively related to student motivation. Notably, every instance in which exposure to a role model backfired, lowering STEM motivation (vs. increasing it or having no effect), was linked in some way to the perceived unattainability of the role model’s career |
Student moderators | |
1) Gender | The range of role models that were motivating for students from groups that are underrepresented in STEM (in particular, girls and racial/ethnic minorities) was narrower than that for majority-group students, and the probability of backfire effects was also greater among underrepresented-group students |
2) Race/ethnicity | |
3) Age | There was no difference in the extent to which role models were effective as a function of student age |
4) Identification with STEM | The prediction was that a general sense of similarity to people in STEM, regardless of whether this similarity is demographic or psychological, might be sufficient to attract students to STEM in the first place, whereas at subsequent stages (once students have identified with STEM) demographic similarity to role models might take on particular importance. There was not enough evidence to confidently evaluate this prediction |