The United States is facing a teacher shortage, specifically for STEM-certified teachers (Sutcher et al., 2016). In addition to the need to address the shortage, there are calls to reform STEM (science, technology, engineering, and mathematics) education to increase the country’s global competitiveness (Bybee, 2010). In response to these issues, secondary teacher preparation programs, among others, are seeking ways to recruit and retain reform-minded STEM students (Logan et al., 2019). The past 2 decades have seen the emergence of identity as a theoretical lens for education research related to teaching and in relation to STEM disciplines.

This work has emphasized the importance of identity development for meaningful engagement and persistence in professions and roles such as teaching and with respect to disciplines such as science (Avraamidou, 2014; Bell et al., 2017; Carlone & Johnson, 2007). With the goal of recruiting STEM students into teaching and retaining them in the profession, it is important to understand the contribution of teaching identity and STEM identity to students’ persistence in teacher education programs. Furthermore, having an understanding of the intersection between teaching identity and STEM identity is essential for those who run the programs that are on the front lines of nurturing and preparing preservice STEM teachers in their transitions from student to professional and thus play a role in developing students’ science and mathematics teaching identity.

Background

A growing body of literature supports the notion that a professional teaching identity affects teaching practice, professional development, teacher agency, motivations to teach, and resilience (Beauchamp & Thomas, 2009; Beijaard et al., 2004; Day, 2002; Horvath et al., 2018; Wenger, 1998). For preservice teachers who are nearing the end of their programs, their strength of teaching identity predicts their intent to enter the profession and, later on, their engagement within the profession (Horvath et al., 2018). The influence of identity on recruitment into teaching and on future professional practice is significant enough to prompt calls for teacher education programs to specifically target teacher identity development (Beauchamp & Thomas, 2009; Luehmann, 2007).

In addition to teaching identity, there has also been considerable work done on disciplinary identity (e.g., science, mathematics; Bell et al., 2017; Carlone & Johnson, 2007). Although a great deal of this work focuses on students, there are studies that examine the science identity of teachers. In developing teachers’ identities, it is important to recognize the influence of context on professional identity development (Beauchamp & Thomas, 2009). The work on how STEM teachers construct their identity through the context of their discipline reveals differences based on time in-service. Furthermore, a qualitative study of preservice secondary science teachers completing student teaching found a hierarchical perspective on the intersection of their identities: Their teacher identity superseded both science identity and science teacher identity. As such, their science identities were left relatively unused in the context of teaching. However, their science identity was still part of their core identity (Chung-Parsons & Bailey, 2019). This contrasts with the literature on experienced science teachers who saw the role of subject matter as the most influential factor for their identity development (Helms, 1998).

Clearly, identity has been used as a lens for studying teacher preparation; however, this work has primarily been qualitative, with calls for more quantitative work in this area (Avraamidou, 2014). The focus of such work can include how individuals’ multiple identities align with a teaching identity, the ways in which they enact teaching identities, and the influence of different types of activities on teaching identity (Luehmann, 2007). In this article, we present a quantitative study that examines the relative contribution of a teaching identity and disciplinary STEM identity on an overall science and mathematics teaching identity for STEM students in recruitment courses for secondary science and mathematics teaching. We also examine the effect of a teaching identity and disciplinary STEM identity on persistence in the program. Given that a science and mathematics teaching identity is a complex intersection between having a teaching identity and a STEM identity, it is important to understand the role that each plays in developing an integrated disciplinary teaching identity, as well as these identities’ effects on student persistence. As such, we address the following research questions:

• What are the relative effects of having a teaching identity and a STEM identity on science and mathematics teaching identity?
• What are the relative effects of having a teaching identity and a STEM identity on persistence in a secondary science and mathematics teacher preparation program?

Methods

This study took place within the context of a preservice STEM teacher preparation program at a public university. The university is a replication site for the UTeach model of STEM teacher preparation, which relies on partnerships between education and STEM faculty to prepare teachers who have a substantive knowledge base in both disciplinary content and reform-based pedagogy (Luft et al., 2005).

As part of the program, a survey was developed and implemented to measure various aspects of identity development in relation to teaching, STEM, and science and mathematics teaching. The survey was administered electronically at the end of each fall and spring semester to students in the recruitment courses offered by the program between fall 2017 and fall 2019 (N = 389), representing five cohorts of students across 3 years of recruitment classes. Recruitment classes are offered free of charge for one or two credits, and students choose one of two possible recruitment courses as their first course in the program based on students’ stage in their degree program (e.g., the 1-credit course is typically taken by freshmen and has a subsequent 1-credit course; the 2-credit-hour course that combines the two 1-credit courses is typically taken by juniors). These courses serve as an introductory opportunity for students to immediately learn and implement inquiry-based teaching strategies in actual classrooms.

Although students may come to these recruitment classes with different prior experiences in STEM teaching (e.g., peer tutoring), all of the students are considered new to the teaching program, with the recruitment courses serving as their introduction to the program and the inquiry-based pedagogies. Because all of the students are new to the teaching program, we do not observe a wide variability in their STEM teaching knowledge, although they may have prior experiences such as tutoring or informally teaching classmates. The word persistence refers to the continuing enrollment of students who took at least one class in the program in a subsequent semester after taking the recruitment class.

The survey included specific items for students to report how they saw themselves (identity) in relation to teaching, science and mathematics, STEM disciplines, and science and mathematics teaching using a scale ranging from 0 to 4, where a higher rating indicated stronger agreement (see Table 1 for variables, means, and standard errors). We created an overall STEM identity measure by averaging the measures for science and mathematics identity and STEM identity. Note that a paired t-test revealed a significant difference between students’ STEM identities and teaching identities (t(382) = 9.996, p < 0.001), with the mean for STEM identity (3.14 ± 0.05) significantly higher than the mean for teaching identity (2.35 ± 0.06).

In terms of the relationship with science and mathematics teaching identity (addressing the first research question), we ran a regression model with teaching identity and STEM identity predicting the overall measure of science and mathematics teaching identity. Within this regression model, we tested the variable of a student’s year in college as a way to possibly account for variability in STEM identity based on students’ exposure to more STEM content. To examine the effect on persistence in the program (addressing the second research question), we ran a logistic regression with teaching identity and STEM identity predicting persistence. All analyses were run in R statistical software (R Core Team, n.d.).

Results

Table 2 summarizes the results of the regression, with teaching identity and STEM identity predicting science and mathematics teaching identity (Model 1). Note that we checked variance inflation factors (VIF) to assess whether there were multicollinearity issues and the VIFs were less than 1.2, where the acceptable threshold is typically less than 3 to 5 (Hair et al., 2014). In addition, after testing the model including year in college, we found that year in college was not significant and removed it from the model.

As expected, teaching identity and STEM identity are both significant predictors (p < 0.001) of the science and mathematics teaching identity variable, as represented in Figure 1. However, the effect size of teaching identity is large (0.77) compared to the effect size of STEM identity (0.12). This indicates that teaching identity lends a much larger contribution to science and mathematics teaching identity than does STEM identity. There was no significant interaction effect between teaching identity and STEM identity. The adjusted R2 (0.68) indicates that teaching identity and STEM identity together account for 68% of the variance in the science and mathematics teaching identity measure.

In terms of predicting persistence in the preservice program (Model 2), teaching identity was a significant predictor (p < 0.001), but STEM identity was not. Again, there was no significant interaction effect between teaching identity and STEM identity. The odds ratio indicates that a one-unit increase in teaching identity (on a scale of 0 to 4) translates into 2.23 higher odds of persisting to another course in the program. Figure 2 represents the predicted probability that students will persist in the program based on their level of teaching identity. The McFadden pseudo-R2 for Model 2 is 0.14, indicating that even though the two variables account for persistence to some degree, there are likely many other factors that account for students’ persistence in the program.

Discussion

The higher means of STEM identity as compared with teaching identity in the descriptive statistics (p < 0.001) clearly illustrate that STEM identity is much more salient than teaching identity, even for students who enroll in a science or mathematics teaching course. Furthermore, the results of Model 1 indicate that both teaching and STEM identity are significant predictors of the intersecting science and mathematics teaching identity. However, the relative contributions of teaching identity are far greater than STEM identity to the overall measure of science and mathematics teaching identity, as represented in Figure 3.

This resonates with Chung-Parsons and Bailey’s (2019) qualitative research on student teachers, which found a hierarchical structure for identity where teacher identity was dominant over science identity. Our study builds on their findings, indicating that students in secondary teacher preparation programs likely construct their identity through the lens of teaching more than through STEM. For teacher preparation programs attempting to recruit more science and mathematics majors into the profession, this construction of identity could complicate recruitment because science and mathematics majors likely have strong disciplinary identities, as reflected in Table 1.

This strong STEM identity is potentially easier to develop because there is a conception of increased financial rewards and status associated with STEM careers versus STEM teaching careers. Such misconceptions can perpetuate negative stereotypes of the teaching profession and stymie teaching identity development. That students with strong science or mathematics teaching identities identify less with their disciplines than they do with teaching could be a potential barrier to entry for students who identify strongly with STEM. This finding is particularly problematic because research has found that individuals who identify more strongly with science disciplines are more likely to feel competent and able to perform with respect to those disciplines (Carlone & Johnson, 2007), attributes that are also important for teaching. Nevertheless, many high-achieving students in their disciplines who aspire to teaching careers face continual barriers and scrutiny of their career choice (Parker, 2016). In sum, this demonstrates the need to value teaching identities and also focus on teaching identity development among students who are developing strong disciplinary identities, possibly at stages prior to college as well as within college disciplines (Marder et al., 2017). This finding also emphasizes the need for secondary science and mathematics teacher preparation programs to better leverage and connect with students’ STEM identities as part of the coursework and program.

The results of Model 2 indicate that teaching identity is a significant predictor (p < 0.001) of persistence in the teacher preparation program. These results build on the findings of Horvath et al. (2018), as they recognize that teaching identity is important for recruiting and retaining students in teaching preparation programs. Although teacher identity was a predictor of persistence, STEM identity had no significant effect on persistence. Again, these findings highlight a potential challenge for the recruitment and retention of science and mathematics majors into teaching.

These results offer insight for both STEM faculty and science and mathematics teacher educators, but there are limitations that future work can address. For example, this study uses single items rather than multiple items to measure certain constructs due to limitations in the data collected consistently over time. Furthermore, there are limitations to the use of self-reported information, although this approach has been relied on for identity research given the unique nature of identity work (Hazari et al., 2020). Although the results of the study align with qualitative work in this field, further research could add additional items to the survey to improve the reliability and validity of the measures.

Conclusion and implications

This study offers insight into an early stage of teacher identity development. Prior work has focused on committed preservice teachers (Chung-Parsons & Bailey, 2019) or experienced science teachers (Helms, 1998), but this study examined STEM majors who are deciding whether to pursue a teaching certificate as part of their programs. Our findings also provide a quantitative complement to a largely qualitative area and can serve as the foundation for a longitudinal study of teacher identity development within a preparation program (Avraamidou, 2014).

These findings also have implications for the recruitment of science and mathematics teacher candidates. Teaching identity is a strong predictor of persistence, so one implication is that programs should develop ways to identify incoming STEM majors who have this identity. For example, programs can partner with admissions departments to identify candidates at the point of application (Get the Facts Out, n.d.). However, these findings also indicate a challenge to recruiting future science and mathematics teachers from among those who have a strong disciplinary identity.

One implication is that programs should examine the role of STEM faculty in recruiting future science and mathematics teachers. In light of work that indicates that both STEM majors and faculty have negative misconceptions about the teaching profession (Logan et al., 2019; Marder et al., 2017), a concerted effort is needed to meaningfully engage with STEM faculty as partners in encouraging students who strongly identify with STEM disciplines to consider teaching and view science and mathematics teaching as a part of the STEM field. Initial recruitment efforts could capitalize on students’ STEM identity by emphasizing the connection between teaching science and improved science research practices (Feldon et al., 2011). Furthermore, secondary teacher preparation programs in science and mathematics can work to better integrate a focus on disciplinary identity as part of recruitment activities to help students realize that science and mathematics teaching requires a deep integration of both STEM disciplinary content and teaching practice (NGSS Lead States, 2013).

Ingelise Giles (gilesi@fiu.edu) and Nicole Cook (nicook@fiu.edu) are professors in the STEM Transformation Institute, Zahra Hazari (zhazari@fiu.edu) and Maria Fernandez (mfernan@fiu.edu) are professors in the Department of Teaching and Learning and the STEM Transformation Institute, and Laird Kramer (kramerl@fiu.edu) is a professor in the Department of Physics and in the STEM Transformation Institute, all at Florida International University in Miami, Florida.