What Types of Feathers Flock Together? How Role Model Similarities Support STEM Motivation

• In Event: Motivational Messages to Broaden Participation in STEM: Important Features of Messages and Messengers

Abstract

Objectives and Theoretical Framework
With an aim to diversify STEM fields, educators often expose students to role models in order to develop their interest and confidence in STEM (e.g., Gonzalez-Perez et al., 2020; Shin et al. 2016). This approach, based in the social cognitive theory, argues that students look to social others when developing their motivational beliefs (Bandura, 1986). This theory posits that role models can promote students’ motivation, particularly when students perceive them as similar (Bandura, 1997; Morgenroth et al., 2015). However, students can be similar to role models in many ways (e.g., race, gender, academic background), and it is not clear what aspects of similarity are most powerful at affecting motivation. The purpose of this study is to understand how different perceptions of similarity to role models may predict college students' motivational beliefs in STEM.

Method
Using two unique samples, we examined perceptions of similarity to role models in STEM. Sample 1 (N = 264; 75.5% women; 77.4% white) comprised undergraduate students from an educational psychology subject pool who were asked about chemistry role models. Sample 2 (N =150; 74.3% women; 49.3% white) comprised undergraduate students in an introductory biology course who were asked about biology role models.
Students first indicated whether they had interacted in college with someone who was good at chemistry or biology. They then responded to an open-ended item, “Do you feel similar to them in any way? Why or why not?” A coding scheme was developed and two trained coders classified responses in terms of whether students mentioned the five types of commonly-referenced similarities (Cohen’s κs = 0.79 to 1.00; see Table 1 for example codes).
Students reported their motivational beliefs for chemistry or biology in terms of perceived competence (8 items; Eccles & Wigfield, 1995; α = .92) and task values (i.e., utility, attainment, and intrinsic value; 6 items; Eccles & Wigfield, 1995; α = .76).
Multiple regression analyses were used to examine whether each type of similarity predicted students’ motivational beliefs.

Results and Significance
More than two-thirds of students reported feeling similar to the role model they described (Sample 1 = 72.9%; Sample 2 = 68.8%). Across both samples, students most frequently described academically-related similarities. The most common categories were feeling similar in ability, effort, and interest. The different types of similarities uniquely predicted motivational beliefs (see Table 2). In sample 1, reporting similarities in ability and interest significantly predicted perceived competence in chemistry (β = .243 and β = .145, respectively). In sample 2, reporting similarities in ability and effort significantly predicted perceived competence in biology (β = .275 and β = .191, respectively). Moreover, these similarities also predicted greater perceived task values for biology, (β = .226 and β = .265, respectively).
Findings revealed that undergraduate students primarily write about academic similarities when considering STEM role models. These academic similarities (particularly in ability) predicted motivational beliefs (particularly competence-related beliefs) in powerful ways. Results shed light on key features of similarity that may optimally support students’ motivational beliefs in STEM.

Authors

• XIAOYIN CHEN, University of Georgia - Athens

• Emily Quinn Rosenzweig, University of Georgia