Collective Effects of Individual, Behavioral, and ...

Collective Effects of Individual, Behavioral, and Contextual Factors on High School Students’ Future STEM Career Plans Alpaslan Sahin, Adem Ekmekci & Hersh C. Waxman

International Journal of Science and Mathematics Education ISSN 1571-0068 Int J of Sci and Math Educ DOI 10.1007/s10763-017-9847-x

1 23

Your article is protected by copyright and all rights are held exclusively by Ministry of Science and Technology, Taiwan. This eoffprint is for personal use only and shall not be self-archived in electronic repositories. If you wish to self-archive your article, please use the accepted manuscript version for posting on your own website. You may further deposit the accepted manuscript version in any repository, provided it is only made publicly available 12 months after official publication or later and provided acknowledgement is given to the original source of publication and a link is inserted to the published article on Springer's website. The link must be accompanied by the following text: "The final publication is available at link.springer.com”.

1 23

Author's personal copy Int J of Sci and Math Educ DOI 10.1007/s10763-017-9847-x

Collective Effects of Individual, Behavioral, and Contextual Factors on High School Students’ Future STEM Career Plans Alpaslan Sahin 1 & Adem Ekmekci 2 & Hersh C. Waxman 3

Received: 17 August 2016 / Accepted: 8 August 2017 # Ministry of Science and Technology, Taiwan 2017

Abstract The purpose of this study is to investigate how students’ high school experience, math and science efficacy, and student, parent, and teacher expectations affect their plans for college major choice after controlling for students’ gender, ethnicity, and parental variables. Over 1500 9th grade students participated in the study. Using logistic regressions, we found that males and students whose parents held degree from a U.S. college are more likely to consider STEM majors in college. Hispanic students were found less likely to consider STEM major in college compared their Asian counterparts. Students who completed more STEM PBL projects and attended STEM summer camps are more likely to consider STEM majors. Students with higher GPAs also indicated that they are more likely to study STEM majors in college. In addition, students with higher parent and teacher encouragement are more likely to consider selecting a STEM major after graduating from high school. Moreover, students who had higher math and science efficacy are also more likely to consider choosing a STEM major in college. Last but not least, we found that students’ future career choice is also positively associated with their interests and goals they develop during high school years. Other findings and interaction effects with gender and ethnicity are also discussed in the paper. Overall, this study demonstrates that students’ contemplations about STEM major selection in college is * Alpaslan Sahin [email protected]; [email protected] Adem Ekmekci [email protected] Hersh C. Waxman [email protected]

1

Harmony Public Schools, Houston, TX, USA

2

Rice University School Mathematics Project, Rice University, Houston, TX, USA

3

Texas A&M University, College Station, TX, USA

Author's personal copy A. Sahin et al.

influenced by the complex interplay between the individual, environment, and behavior, three major components of social cognitive career theory. Keywords PBL . Pygmalion effect . Social cognitive career theory . STEM

Introduction A plethora of international calls and reports urges the importance of science, technology, engineering, and mathematics (STEM) due to its critical role in nations’ economy and defense systems (European Commission, 2004; National Research Council [NRC], 2013; National Science Board, 2007; Office of the Chief Scientist, 2012; Organisation for Economic Co-operation and Development [OECD], 2014; President’s Council of Advisors on Science and Technology, 2012). Due its critical role, stakeholders including policymakers, educators, scientists, and researchers have been promoting students’ interest and achievement in the STEM fields (NRC, 2013; OECD, 2014). Similarly, substantial effort has been devoted to make policies and develop programs that would yield STEM persistence among youngsters, especially students nearing the end of their secondary education and getting prepared for college (Watt, 2008). Therefore, understanding the factors that may play important role in student persistence in STEM to inform such policies and programs seems essential. Although not a new field of study, STEM persistence studies are not a one-time/era issue as many dynamics (i.e., new approaches, interventions, policies, and programs) continue to evolve and may have different outcomes and implications for the STEM future of every nation in the world. The vast majority of existing evidence in the extant literature on student persistence on STEM pipeline is based on college-level experiences (Nakakoji, Wilson, & Poladian, 2014; Sass, 2015). This may conceal important impacts of pre-college experiences, preparation, and resources on pursuing STEM areas during postsecondary years. There is research indicating that factors that come into play in students’ lives at younger ages could be more influential in their future education and career path (Crosnoe & Muller, 2014; Eccles, Midgley, Wigfield, Buchanan, Reuman, Flanagan, & Mac Iver, 1993; Graham, Frederick, Byars-Winston, Hunter, & Handelsman, 2013; Tai, Liu, Maltese, & Fan, 2006; Wang, 2013). Therefore, the goal of this study is to investigate the factors during the high school years that contribute to student motivation and interest to continue in STEM fields. More specifically, 9th grade students’ high school STEM experiences, teacher and parental expectations, and students’ motivational beliefs were the factors of interest in this study. The present study examines the high school experiences of 9th grade students from a large, public school district that has emphasized STEM in the school curriculum. This district was purposefully selected because of its emphasis on integrating STEM across the curriculum and variety of STEM activities offered on its campuses and because of the large diversity within the student population. Theoretical Framework As an extension and application of social cognitive theory (Bandura, 1986) to career choice, Lent, Brown, and Hackett’s (1994) social cognitive career theory (SSCT)

Author's personal copy Collective Effects of Individual, Behavioral, and Contextual...

provides the theoretical grounds for this study. SCCT asserts that one’s career choice is influenced by the interest and goals the individual develops and refines through complex interplay between the individual, environment, and behavior (Lent et al., 1994, Lent & Brown, 1996). Individual (personal) factors include racial and ethnic background, gender, and socioeconomic status (Howard, Carlstrom, Katz, Chew, Ray, Laine, & Caulum, 2011; Lee, Min, & Mamerow, 2015). Moreover, SCCT suggests that the decision to pursue a field can be explained by interests and goals (Lent & Brown, 2006; Mujtaba & Reiss, 2014; Uitto, 2014; Wang, 2013); contextual, social, and academic supports and barriers (Lee et al., 2015; Maltese & Tai, 2011; Mujtaba & Reiss, 2014); and early experience in the content area including course-takings (Lee et al., 2015; Ma & Johnson, 2008). Although SCCT provides a proper framework for investigations related to student STEM persistence, there is paucity of studies that utilize SCCT to explain STEM persistence (Wang, 2013). The key factors within the SCCT framework relate to student motivation (e.g., task value, self-efficacy, interest, outcome expectations). These psychological variables are considered as the mediators connecting other personal and contextual factors to future career choice and decisions (Yu, Corkin, & Martin, 2016). Individuals’ behavior and actions are kneaded primarily by their sense of personal capabilities (self-efficacy) and their beliefs about likely consequences of performing particular actions (outcome expectancy; Bandura, 1986). Empirical research at the global scale has shown that students with higher self-efficacy and expectation for mathematics and science are more likely to persist and be successful in these areas (Andersen & Ward, 2014; Lee et al., 2015, Mujtaba & Reiss, 2014; Uitto, 2014; Wang, 2013). In addition to individual motivational traits, SCCT framework recognizes other immutable personal traits (i.e., race/ethnicity, sex), and environmental factors (supports and barriers at school and home) in molding individuals’ career aspirations and choices (Lent & Brown, 1996). Research that relates to SCCT model has investigated whether factors associated with student STEM persistence can vary substantially by students’ sociocultural characteristics (e.g., gender differences and differences across ethnically underrepresented minority student populations; Ing, 2014; Riegle-Crumb, Farkas, & Muller, 2006; Seymour & Hewitt, 1997; Uitto, 2014; Yu et al., 2016). SCCT also underlines the role of contextual supports and barriers in determination of choice actions (Maltese & Tai, 2011; Mujtaba & Reiss, 2014; Yu et al., 2016). In a high school setting, for example, students’ intentions to persist in STEM can be influenced by the demands of postsecondary education such as academic readiness, securing financial resources or aids, and social expectations, some of which can be supportive whereas others can act as a barrier (Andersen & Ward, 2014; Wang, 2013). Several social parties are engaged during this process including parents, teachers, and peers. This study integrates SCCT and previous global research on factors closely connected with academic choices of students and addresses the interconnectedness among personal and environmental factors and STEM choice under the SCCT framework (see Fig. 1). These factors are detailed below under literature review. This study analyzes survey responses and academic records of 9th grade students from a public school system with a STEM-focus. Several studies utilized national large-databases—for example, High School Longitudinal Study of 2009 (Andersen & Ward, 2014; Lee et al., 2015; Maltese & Tai, 2011; Wang, 2013)—when studying STEM persistence. The present study avoids the issues with secondary analysis of national databases such


as the use of proper weights and restriction of variables of interest (Rutkowski, Gonzalez, Joncas, & von Davier, 2010). STEM Experiences of Students According to SCCT, learning experiences is one of driving factors for developing self-efficacy expectations, outcome expectations, and, in turn, interests and choice actions related to career-related choice behavior (Lent et al., 1994). Research on students’ persistence in STEM fields has documented several factors related to learning experiences in STEM: (a) success in mathematics and science classes (Andersen & Ward, 2014; Hansen, 2014), (b) access to AP STEM courses (Ehrenberg, 2010; Museus, Palmer, Davis, & Maramba, 2011; Nakakoji et al., 2014), (c) opportunities to participate in STEM-related projects during high school (Gottfried & Williams, 2013; Graham et al., 2013), (d) interaction with successful peers (Museus et al., 2011; Riegle-Crumb et al., 2006), and (e) high self-efficacy (Andersen & Ward, 2014; Eccles & Wigfield, 2002). The number of research studies at the high school level is limited when compared to those at the postsecondary level. However, the focus on college coursework, instructors, and grades in understanding STEM major and career choices only provides limited information about what has happened to a student prior to attending college and, in turn, may not provide any connection between postsecondary STEM choices and pre-college indicators (Sass, 2015). Moreover, research indicates that high school and early grades are critical times for developing expectancies for interest and success in STEM fields (Mujtaba & Reiss, 2014; Tai et al., 2006; Wang, 2013). What schools can offer to every student for the best education possible matters substantially for student outcomes (Barber & Mourshed, 2007). Research shows that school factors related to both formal and informal activities and experiences such as the courses offered on campus and extracurricular activities including STEM clubs and project fairs are positively associated with students’ future choices and performance in STEM (Dawes, Long, Whiteford, & Richardson, 2015; DeWitt, Archer, Osborne, Dillon, Willis, & Wong, 2011; Gottfried & Williams, 2013; Ibrahim, Aulls, & Shore, 2016). Moreover, factors that have been found in the literature that influenced students’ persistence and perseverance in STEM fields include early exposure to math and science (Anderson & Kim, 2006; Graham et al., 2013); math and science curriculum

Fig. 1 Conceptual model operationalizing the SCCT framework


(Elliott, Strenta, Adair, Matier, & Scott, 1996); the number of courses taken (Chen & Solder, 2013; Maltese & Tai, 2011; Simpkins, Davis-Kean, & Eccles, 2006); STEM summer camps or internships (Gottfried & Williams, 2013; Kong, Dabney, & Tai, 2013), opportunities and support students receive (Seymour & Hewitt, 1997), and teacher quality and diversity (Andersen & Ward, 2014; Price, 2010). Research also indicates that participation in pre-college mathematics and science enrichment activities is positively associated with motivational beliefs such as self-efficacy, value, and interest in math and science in postsecondary years (for a review, see Sass, 2015). Additional research has indicated that developing expectancies for success and interests in math and science in pre-college years strongly increases the likelihood of students persisting in STEM fields (Tai et al., 2006). Therefore, one of the major factors of interest in this study is the STEM experience of students’ during high school years that would contribute to their career choice plans in STEM fields. Teachers’ and Parents’ Expectations SCCT considers personal and motivational aspects as the most important factors in shaping one’s career-related choices (Lent et al., 1994). However, contextual factors are also important as they may serve as supports and barriers during the choice-making process (Lent & Brown, 1996, 2006; Mujtaba & Reiss, 2014). Others’ expectations can be considered as part of contextual factors. Others’ expectations concerning how an individual would do in a particular situation or how she or he would be influenced by a particular experience may actually influence that individual’s actions (DeWitt et al., 2011; Schunk, Meece, & Pintrich, 2014). Some researchers call this the Pygmalion effect after the Rosenthal and Jacobson (1968) study where they found teacher expectations positively influence their students’ intelligence and reading achievement. Several studies focused on parents’ and teachers’ expectations about their children’s educational attainment. Entwisle and Hayduk (1988) found that children’s early social environment could have long lasting effect on their school performance. Aside from cognitive status, they concluded that parent and teacher expectations for elementary children were linked to children’s reading and math performance 4 to 9 years later. Using data from a national, cross-sectional study of 12-year-olds, Davis-Kean (2005) looked into the extent to which socioeconomic factors indirectly related to children’s academic achievement through parents’ beliefs and behaviors and found that parents’ beliefs/expectations about how much their child would attain in their education had significant positive mediating effects between socioeconomic factors and achievement outcomes. Similarly, Neuenschwander, Vida, Garrett, and Eccles (2007) studied the role of parents’ expectations on student’s self-concepts and achievement using data from two US longitudinal samples and a representative sample of Swiss sixth graders. Neuenschwander et al. (2007) found that parents’ educational expectations about their children predicted students’ standardized achievements in both mathematics and native language, even after controlling for prior performance in each of these subject areas. Moreover, parents’ expectations also mediated the link between family income and students’ standardized achievement tests. In a meta-analysis of quantitative studies about the relationship between parental involvement and students’ academic achievement, Fan and Chen (2001) found that parental expectation for children’s educational achievement had the strongest relationship to children’s academic achievement among all parental factors including home


supervision. Yet another review of the literature by Gunderson, Ramirez, Levine, and Beilock (2012) revealed that parents’ and teachers’ gender-biased expectancies for children’s mathematics competence can influence children’s mathematics attitudes and performance. Among the three directions proposed by Gunderson et al. (2012) in which future researchers should study is investigating the specific mechanisms underlying the causal link between parents’ and teachers’ beliefs and children’s motivation for STEM including mathematics attitude. Secondary Education Years The goal of this study is to understand the impact of earlier educational experiences (starting at 9th grade) on high school performance and the impact of these experiences on the transitions that students make from high school to adult roles. The transition from secondary education to postsecondary life is of special interest to federal policy and programs (Ingels, Dalton, Holder, Lauff, & Burns, 2011). Adolescence is a time of psychological and physical changes. Attitudes, aspirations, and expectations that are sensitive to adolescents’ environments and what adolescents experience influence the process of choosing among opportunities especially during high school years (Nagy, Trautwein, Baumert, Köller, & Garrett, 2006). While traditional research has noted the link between general ability and examination in predicting university entry and major choices (see Parker, Schoon, Tsai, Nagy, Trautwein, & Eccles, 2012), evidence suggests that high school experiences including academic achievement and motivational beliefs may account for much of this relationship (Bowen, Chingos, & McPherson, 2009). Maltese and Tai (2011) found that most students who chose STEM made that choice during high school, and that choice is related to a growing interest in mathematics and science. Hoogstra (2002) analyzed the National Educational Longitudinal Study of 1988 and found that the number of advanced mathematics courses taken during high school was predictive of choosing and persisting in mathematics and science majors in college. Similarly, Nagy et al. (2006) found that students’ college major choice reflected their specialized courses taken in high school. Moreover, researchers point out the importance of high school years as early as the 9th grade. For example, there is research indicating that students who take Geometry in 9th grade (as opposed to 10th) are more likely to persist in STEM majors (Ma & Johnson, 2008; Maltese & Tai, 2011). It is clear that students’ experiences with the STEM content starts as early as 9th grade. Indeed, students’ 9th grade experience is key in their career selection. According to the U.S. Department of Education, nearly 28% of high school freshmen declare interest in a STEM-related subject (Hom, 2014). Interestingly, 57% of those loose interest and major in non-STEM areas by the time they graduate from high school. Therefore, to capture all the factors contributing to STEM persistence and detect reasons why students change their initial choices, it is pivotal to include students’ 9th grade experience as well. Research on high school students’ persistence in STEM areas has also documented that demographics of students have influence on their career choices (e.g., gender and ethnicity; Eccles, 2005; Howard et al., 2011; Parker et al., 2012; Riegle-Crumb et al., 2006; Seymour & Hewitt, 1997). Therefore, the demographic variables are also included in this study. Purpose of the Study A very recent study investigated the effects of student’ mathematics and science efficacy and students, parents, and teachers’ expectations in students’ STEM-M(edicine) career selection (Lee et al., 2015). The present study extends


the Lee et al.’s (2015) study by utilizing SCCT framework more comprehensively and by adding additional school and out-of-school variables including academic activities (e.g., #AP, #STEM PBL projects) and extracurricular activities (e.g., #STEM clubs, #science fair projects) that might students to choose a STEM major. Thus, we might have a more solid understanding of how students choose a STEM career. With this goal in mind, we designed a study in collaboration with Texas A&M university, Rice university, and Harmony Public Schools to examine how students’ high school experience, math and science efficacy, and parent and teacher expectations, and the interplay between them affect their major choice decision when students’ gender, ethnicity, and parent variables are controlled. Our research questions are given below: 1. What are the impacts of school and out-of-school-related activities on students’ intention to pursue a STEM degree? 2. What are the impacts of both teacher and parental educational expectations on students’ intentions to pursue a STEM degree? 3. What are the impacts of a students’ self-efficacy in math and science and college expectations on the likelihood of pursuing a STEM degree? 4. What are the impacts of interaction effects between individual, environment, and behavior on students’ likelihood of pursuing a STEM degree?

Method Setting: Harmony Public Schools (HPS) Harmony Public Schools are a non-profit open-enrollment K-12 college preparatory public charter school system across Texas.1 As of April 2017, it operates 48 schools serving a diverse student population of over 31,000 where 60% of students receive free or reduced price lunch and 70% are underrepresented minorities. Participants Participants are selected from the second largest charter school system in the Nation, HPS. Two thousand one hundred and fifty-seven class of 2019 9th grade students from 20 HPS schools were invited to participate in the study. We obtained a total of 1520 students’ parent consents and administered the survey (~ 70%). This study reports the first-year findings from of a 4-year longitudinal study which aims to track class of 2019 students to examine their secondary school and out-of-school formal and informal STEM-related experiences that may have impact on their STEM career choice selections. Demographics of participants are given in Table 1 below. 1

HPS is an open-enrollment college prep school system. Because HPSs are public charter schools, they must follow all federal laws that apply to any other public schools. Therefore, they have to accept students by lottery and cannot choose its students based on their interests or achievements.

Author's personal copy A. Sahin et al. Table 1 Descriptive data for participants N

Answer range

Overall answers’ mean

Overall answers’ std. deviation

College degree in the USA

1518

0–1

0.45

0.50

Parent education level

1318

0–4

2.01

1.38

Household income

1235

0–2

0.89

0.63

Count STEM SOS project

1500

1–6

2.20

1.61

Count STEM SOS presentation

1353

0–6

1.83

1.23

STEM SOS website contest

1513

0–1

0.20

0.40

DISTCO contest

1514

0–1

0.18

0.39

Count science fair participation

1477

3 – 16

6.82

2.33

STEM summer camp

1516

0–1

0.12

0.33

STEM internship

1513

0–1

0.07

0.25

STEM choice of major

1515

0–1

0.58

0.49

Count STEM AP

333

1–2

1.17

0.37

Total AP taken

605

0–3

1.16

0.45

Educational degree expectation

1512

1–7

5.10

1.59

Parent expectation

1516

1–4

3.49

0.75

STEM teacher expectation

1515

1 –4

2.93

0.91

Good at math

1517

1– 5

3.62

1.14

Good at science

1517

1–5

3.50

0.94

GPA

1489

0–4

3.04

0.80

White

1520

0–1

0.15

0.36

Black

1520

0–1

0.15

0.36

Hispanic

1520

0–1

0.52

0.50

Male

1489

0–1

0.51

0.50

Data Collection Data collection began on March 2016 by surveying all the HPS high schools who have 9th graders. Twenty-three HPS schools were invited to take part in the study. Twenty of them agreed to be in the study. We were told it would be more convenient and easier to track students through Naviance because they have subscriptions to it and their students already have accounts to log on to the system. We entered the survey to Naviance. Naviance is a comprehensive K-12 college and career readiness platform (Naviance, 2016). Parent consent forms were shared with the parents of more than 2100 students through school websites and hardcopies were given during parent-teacher conferences. One thousand five hundred and twenty out of 2157 students’ parent consent forms were obtained. We had to remind college guidance counselors of each school 5 times to administer the completion of the survey for those 1520 students. Each counselor at HPS had an hour per week to meet with students of each grade to do their course selections for the upcoming year during every spring semester. They administered the survey during those hours between early March and late April. The response rate was almost


70%. One thousand five hundred and twenty participants include 725 females (48%), 242 white (15.9%), 238 Black (15.7%), 805 Hispanic (53%), and 228 Asian (15%). We used an online survey consisting of 23 questions to request information about four categories of variables: (a) student demographics, (b) family context, (c) school and out-of-school-related activities, and (d) Pygmalion effect variables (see Appendix). We used items from previously developed reliable and valid instruments. We adapted Pygmalion factor items from the Lee et al. (2015) study. In addition, math and science self-efficacy items were adapted from previously developed valid and reliable scales used in Longitudinal Study of American Youth (Miller, 2014). Individual factor variables (e.g., gender, ethnicity, SES) and high school STEM experience variables (based on what was offered on HPS campuses) were also included. Lastly, we also included students’ grade point average (GPA) to see how it relates to students’ STEM career choice. Variables Our dependent variable was STEM Choice. Students declared either 1 as considering a major in a STEM-related area or 0 indicating not considering STEM majors. We used National Science Foundation’ (2010) STEM profession classification. We had two group of independent variables. The first group included school and out-of-schoolrelated activities. STEM SOS is an interdisciplinary, standards-focused, and engaging STEM teaching approach that is teacher-facilitated, student-centered, and directed through sets of projects- and inquiry-based (P&IBL) projects (Sahin & Top, 2015). Samples of STEM SOS projects can be found at http://www.stemsos.org/#home. The second group of variables included students’ expectation about their educational attainment, parents and STEM teachers’ expectations, and students’ math and science efficacy. We used gender, ethnicity, parents’ education level, parents’ college degree, and household income as our control variables to see how other variables affect students’ STEM major selection when we control students and parent demographics. The conceptual model in Fig. 1 also provides a summary of variables. Analyses Descriptive analyses were conducted to see the percentages of those who contemplate majoring in STEM fields in college. Before we ran multiple logistic regressions to investigate which group of variables predict students’ probability of choosing a STEM major, we verified the assumptions of absence of multicollinearity, independence of errors, and linear relationship between the independent variables and the log odds (Meyers, Gamst, & Guarino, 2006). Table 1 provides the description for the variables used in the data analyses. Numbers for each question may vary because, for example, students in 9th grades may have not taken many AP courses.

Results The descriptive findings highlight the fact that HPS class of 2019 9th graders who responded to the survey were more than twice as likely to declare interest in a STEM-


related field major than the average of students (a) across the state of Texas, and (b) across the USA (see Table 2 below for more details). Correlational analyses among all the independent variables revealed low to moderate correlations among the variables of interest that eliminated the multicollinearity issue for logistic regressions (Kline, 2015). For the first research question, we ran a logistic regression analysis that revealed several significant results. Pertaining to model 1— individual factors and school and out-of-school factors as predictors of STEM choice outcome—we found that 9th graders contemplation of choosing a STEM major was significantly different by gender (see Table 3) with males being 1.9 times more likely to consider STEM majors in college. Moreover, US first generation students—students whose parents did not earn a degree from a US college—seemed to be in a disadvantaged situation. More specifically, students whose parents held a degree from a US college were about 1.8 more likely to consider STEM track after high school. In terms of school and out-of-school STEM experiences, a number of STEM SOS projects students completed were found to positively impact STEM choice. More specifically, one unit increase in the number of STEM SOS project completion was associated with about 1.3 more likelihood for STEM choice. Similarly, whether or not students attended a STEM summer camp was also significant: summer camp attendance was associated with 2.4 more likelihood for STEM choice. Finally, students with one standard deviation higher GPA were 1.8 times more likely to consider choosing STEM majors after high school graduation. To deepen our analysis and explore the moderating effects of gender and ethnicity on the relationships of school and out-of-school variables with STEM choice, we ran the interaction effects model (model 2). Interestingly, STEM summer camp participation was no longer a significant stand-alone predictor of STEM choice in the interaction effects model. Among all the interaction terms (gender and ethnicity [Hispanic, Black, and White] by all school and out-of-school variables) only GPA and gender interaction came out significant favoring females. More specifically, one standard deviation increase in GPA was associated with 0.4 less likelihood for STEM choice for male students when compared to their female peers.

Asian Students Were the Reference Category for the Dummy-Coded Ethnicity Variables For the second research question, logistic regression results for the model 3 (Pygmalion effects model without interactions) indicated that gender and Hispanic Table 2 STEM field majoring rates by gender and ethnicity compared to Texas and national average Overall

Male

Female

White

Blacks

Hispanic

Asians

HPS

57.9 (860)

63.4 (494)

51.8 (366)

62.6 (142)

63 (150)

51.3 (407)

72.1 (158)

State

25.7a

46a

14.4a

32.7a

23.7a

29.6a

35.9a

National

28b

44.9a

13.4a

30a

25.1a

29.3a

34.6a

a

My College Options (2012)

b

ASTRA (2015)

Author's personal copy Collective Effects of Individual, Behavioral, and Contextual... Table 3 Logistic regression results for school and out-of-school-related variables as predictors Model 1

Constant

Model 2

B

Sig.

Exp (B)

B

Sig.

Exp (B)

− 1.992**

0.009

0.136

− 3.316**

0.009

0.043

0.649**

2.932*

0.036

6.807

Individual factors Gender

0.004

1.915

White

− 0.369

0.323

0.691

− 0.481

0.278

.657

Black

− 0.348

0.389

0.706

− 0.461

0.294

.645

Hispanic

− 0.513

− 0.662

College_degree_in the USA Household_income

0.585* − 0.151

0.139

0.599

0.012

1.795

0.418

0.860

− 0.256

0.288

0.918

− 0.183

0.000

1.279

0.622**

0.083

.533

0.008

1.901

0.210

.782

School/out-of-school-related factors Count_project_present_place Count_STEM SOS_project

− 0.085 0.246***

0.101

.834

0.244*

0.010

1.279

SOS_website_contest

0.238

0.359

1.268

0.439

0.206

1.577

DISTCO_contest

0.284

0.301

1.329

0.652

0.113

1.825

Science fair_new

− 0.223

0.069

0.800

− 0.046

0.690

.932

0.013

2.353

0.960

0.296

1.722

− 0.120

0.776

0.887

− 0.502

0.160

.401

Total_AP_taken

0.058

0.822

1.060

−0.183

GPA

0.627***

0.000

1.872

STEM_summer camp STEM_internship

0.856*

0.244***

0.266

.622

0.000

3.299

.007

0.402

Interaction terms Gender X GPA

− 0.912**

To save space, although every interaction term between gender and ethnicity (Hispanic, Black, & White) and all the school and out-of-school variables were included in the analysis, only significant interaction effects were included in the table Note: *p < .05, **p < .01, ***p < .001

covariates as individual factors came out significant (see Table 4). That is, male students are 1.5 times more likely to consider STEM majors in college. Also, Hispanic students are 0.5 times less likely to consider STEM majors compared to their Asian peers (reference category). Beyond individual factors for this research question, both Pygmalion factors (parent and teacher encouragement) were found to be significant. Students who reported higher levels (in one unit increments) of both parent and STEM teacher encouragement about going to college were associated with greater likelihood of STEM choice—more specifically 1.2 and 1.5 times more likely for parent and teacher encouragement, respectively). Model 4 (Pygmalion effects including interaction terms with gender and ethnicity) revealed no significant interaction effects. However, STEM teacher encouragement was still significant in this model. The third research question relates to exploration of motivational factors’ impact beyond other individual factors on STEM major choice. Similar to earlier analyses, the difference between model 5 and model 6 is that the latter includes interaction terms to deepen the analysis made in model 5. As Table 4 displays,

Author's personal copy A. Sahin et al. Table 4 Logistic regression results for teacher and parental expectation variables as predictors Model 3

Constant

Model 4

B

Sig.

Exp (B)

B

Sig.

Exp (B)

− 1.482***

0.000

0.227

− 2.528**

0.004

0.080

0.371**

2.289


0.002

1.449

0.828

0.193

White

− 0.165

0.459

0.848

− 0.235

0.848

0.790

Black

− 0.322

0.140

0.724

2.167

0.051

8.732

Hispanic

− 0.593**

0.001

0.553

0.111

0.903

1.117

College_degree_in the USA

0.102

0.435

1.107

0.107

0.415

1.113

Household_income

0.092

0.364

1.097

0.096

0.353

1.100

Pygmalion factors Parent_encouragement

0.213*

0.011

1.238

0.330

0.186

1.391

STEM_teacher_encouragement

0.389***

0.000

1.475

0.600**

0.003

1.822

Interaction terms Gender/ethnicity X

None significant

Note: *p < .05, **p < .01, ***p < .001, None significant: p > .05

form model 5, male students are 1.3 times more likely to consider STEM majors in college. In addition, Hispanic students are .6 times less likely to consider STEM majors compared to their Asian peers (reference category). Pertaining to motivational factors (math and science self-efficacy and educational degree selfexpectation [from high school drop-out to earning a post-bac degree]), both selfefficacy in math and self-efficacy in science were found to be significant. More specifically, students who reported higher levels of math and science self-efficacy (in one unit increments) about going to college were 1.2 and 1.5 times more likely to consider STEM major in college, respectively. Model 6 (involving interaction terms with gender and ethnicity) revealed only one significant interaction effects: females who had higher levels of education degree self-expectation tended to contemplate about STEM major in college 0.9 less likely than male students who also had higher self-expectations (Table 5). Finally, the fourth research question addresses the interplay between individual, environmental, and behavioral (closely tied to motivation) factors in relation to students’ STEM major plans in college. In a sense, the fourth research question investigates the SCCT framework in a collective way in contrast to the first three research questions, which took a singular approach (only one of school context, Pygmalion, and motivational dimensions of SCCT). We believe both approaches add value to the research because in some educational contexts it may not be always possible to control all aspects of SSCT framework, in the case which singular approaches may be more realistic to address. Moreover, the fourth research question also builds on the earlier research questions: we only included the factors that were found significant in earlier research questions since the sample size, though it was large, was not large enough to produce powerful analysis with all the variables included (more than 20). We also

Author's personal copy Collective Effects of Individual, Behavioral, and Contextual... Table 5 Regression results for student expectation, math and science efficacy variables as predictors Model 5

Constant

Model 6

B

Sig.

Exp (B)

B

Sig.

Exp (B)

− 2.042***

0.000

0.130

− 2.149*

0.026

0.117

0.251*


0.041

1.285

0.794

0.260

2.212

White

− 0.229

0.313

0.795

− 1.530

0.261

0.217

Black

− 0.347

0.118

0.707

− 1.153

0.363

0.316

Hispanic

− 0.528**

0.003

0.590

− 0.339

0.733

0.713

College_degree_in the USA

0.121

0.360

1.129

0.122

0.362

1.130

Household_income

0.117

0.258

1.124

0.111

0.289

1.117

Motivational factors Self-education degree expectation

− 0.032

0.407

.968

0.075

0.472

1.078

Good_at_math

0.285***

0.000

1.330

0.144

0.369

1.154

Good_at_science

0.456***

0.000

1.578

0.483*

0.010

1.621

− 0.159*

0.046

0.853

Interaction terms Gender X Self-education degree expect Note: * p < .05, **p < .01, ***p < .001

created some additional interaction variables between school-related, Pygmalion, and motivational factors that were found significant from previous analyses. The results for the fourth research question indicated that gender and Hispanic covariates as individual factors came out significant (see Table 6; Model 7: without interactions model). More specifically, male students are 1.5 times more likely to consider STEM majors in college. Also, Hispanic students are 0.6 times less likely to consider STEM majors compared to their Asian peers (reference category). Beyond individual factors for this research question, count STEM SOS projects, STEM summer camp, GPA, STEM teacher encouragement, and math and science self-efficacy were found significant with greater levels of these variable were associated with greater likelihood of contemplating about choosing STEM majors in college. The Model 8 (with interactions model) revealed several significant interaction effects including both positive interactions (black X math efficacy, GPA X math efficacy, science efficacy X parent encouragement) and negative interactions (science efficacy X STEM summer camp, GPA X science efficacy, math efficacy X parent encouragement). More specifically, black students with higher math efficacy are more likely to major in STEM-related fields in college than Asian students with similar level of math efficacy. Students with higher GPA are more likely to choose STEM fields in college than reference group with similar level and math efficacy. Students with higher science efficacy are more likely to choose STEM majors compared to their peers with lower science efficacy but similar levels of parent expectation. Interestingly, students with higher GPA are less likely to major in STEM compared to students with similar level of science efficacy. Likewise, for summer camp by science-efficacy interaction, students who participate in summer camp are less likely to major in STEM areas than students who do not participate in STEM summer camp with similar level of science

Author's personal copy A. Sahin et al. Table 6 Regression results for individual, environmental, and motivational factors as predictors Model 7

Constant

Model 8

B

Sig.

Exp (B)

B

Sig.

Exp (B)

− 4.145***

0.000

0.016

− 1.409

0.587

0.244

0.401**

2.703


0.003

1.493

0.994

0.296

White

− 0.285

0.249

0.752

− 0.567

0.766

0.567

Black

− 0.308

0.207

0.735

0.101

0.956

1.106

Hispanics

− 0.443a

0.030

0.642

0.148

0.922

1.159

0.059

0.673

1.061

0.084

0.569

1.087

− 0.007

0.953

0.993

− 0.062

0.598

0.940

College_degree_in the USA Household_income School/out-of-school-related factors Count_STEM SOS_project

0.106a

0.014

1.112

− .501

0.185

0.606

STEM_summer camp

0.624**

0.004

1.866

2.641

0.200

14.023

GPA

0.415***

0.000

1.514

0.504

0.442

1.655

Pygmalion factors Parent_encouragement

0.091

0.311

1.095

0.018

0.977

1.018

STEM_teacher_encouragement

0.307***

0.000

1.359

− 0.041

0.945

0.960

Good_at_math

0.152a

0.012

1.164

− 0.456

0.349

0.634

Good_at_science

0.368***

0.000

1.445

− 0.397

0.477

0.673

0.528a

0.031

1.696

− 0.541a

0.049

0.582

Motivational factors

Interaction terms Black X math efficacy Science-efficacy X STEM summer camp GPA X math efficacy

0.250**

0.004

1.283

GPA X science efficacy

− 0.262*

0.026

0.769

Math efficacy X parent encouragement

− 0.214a

0.018

0.807

Science efficacy X parent encouragement

0.299**

0.004

1.348

Gender by count STEM SOS projects

1.298a

0.30

3.660

To save space, although every interaction term between gender and ethnicity (Hispanic, Black, & White) and all motivational variables were included in the analysis, only significant interaction effects were included in the table a

Only significant factors from Models 1–6

Note: *p < .05, **p < .01, *** p < .001

efficacy. Moreover, math efficacy had a negative interaction with parent expectation— higher parent expectation plays negative role in students’ STEM major selection for students who have similar levels of math efficacy.

Discussion In the present study, we examined 9th grade HPS students’ aspirations for choosing STEM as a major in college and what factors may be associated with


their current decision for choosing STEM as their college major. First, we found that 9th grade male students were more likely to contemplate choosing a STEM major than female students. Second, we found that students whose parents held a degree from a US college were more likely to consider STEM track after high school. Generally, students with higher GPAs in 9th grade were more likely to consider studying STEM majors after graduation than students with lower GPAs, but this finding was mediated by an interaction effect revealing that the higher GPAs were associated with 0.4 less likelihood for STEM choice of major for males than female students. Moreover, 9th grade students who completed more STEM SOS PBL and who participated in STEM summer camp were also found to indicate that they were significantly more likely to major in STEM fields. In exploring teacher and parental expectations after controlling for students’ gender, ethnicity, and parental variables (education and income), we found that male students are more likely to consider STEM majors in college while Hispanic students are less likely to consider STEM majors compared to their Asian peers. Students with higher parent encouragement and higher STEM teacher encouragement were significantly more likely to contemplate studying STEM majors in college. In addition, students with higher math and science self-efficacy were significantly more likely to choose a STEM major. We found that self-education degree expectation had more impact for males. This might be because even though females may think about going to college or completing even a higher degree, they tend to consider less about STEM majors in college or beyond as suggested by earlier results above. The findings for our last research question that tested the complete model we proposed in this study, indicated that gender, ethnicity, STEM summer camp participation, GPA, STEM teacher encouragement, and science self-efficacy impacted students’ choice of a STEM major. In a sense, the fourth research question investigates the SCCT framework in a collective way in contrast to the first three research questions, which took a singular approach (only one of school context, Pygmalion, and motivational dimensions of SCCT). We believe both approaches add value to the research because in some educational contexts it may not always possible to control all aspects of SSCT framework, in the case which singular approaches may be more realistic to address. In other words, we found that when testing the overall model (without interactions) that at least one individual, environmental, Pygmalion, and motivational factor significantly influenced students’ choice of STEM major. This finding lends support to the validity of the social cognitive career theory. The findings from this study raise several important issues regarding promoting students’ STEM interest and motivation in high school. First, there were several school-based activities that appeared to make a difference on students’ STEM aspirations such as completing STEM projects and participating in STEM summer camps. Other school districts may want to explore the specific types of STEM projects that HPS develops for all its students as well as how to develop and implement STEM summer camps. It may also be important for school districts to examine some of the STEM practices that were not found to increase the likelihood of students’ choosing a STEM major. Participating in science fairs, STEM internships, and STEM contests, for example, were not found to significantly


influence students’ STEM choice. Taking AP courses also was not found to relate to students’ choice of STEM. School districts may want to further explore why these activities do not appear to be related to students’ STEM interests. High STEM teacher encouragement as well as high parent encouragement also appears to make a difference on students’ future STEM aspirations. It is important to discuss this specific finding with teachers and parents so that can continue and increase their encouragement to children. Similarly, we found that students’ with higher math and science self-efficacy were more likely to indicate that they were going to choose a STEM major in college. These findings suggest that schools need to continually focus on the affective dimensions of STEM so that all students feel confident that they can be successful in math and science. Overall, the findings of our study are especially encouraging because they suggest that schools can implement programs and practices that may impact students’ future careers in STEM. It may be especially important to implement or programs that may benefit Hispanic high school students the most (Rodriguez, Rhodes, & Aguirre, 2015) since they were found less likely to consider STEM major in college. A second critical issue that emerges from our study is that there still is a wide gap on STEM aspirations between (a) male and female students, (b) Hispanic and non-Hispanic students, and (c) high- and low-achieving students. Prior research in the extant literature has found that science motivation and experience gaps often begin to occur in elementary schools and are generally stable across secondary school levels (Morgan, Farkas, Hillemeier, & Maczuga, 2016). The findings of the present study also suggest that some gaps pertaining to gender and ethnicity (e.g., less impact of teacher encouragement on Black students) exist at the high school level. It is important that the school districts should try to address these gaps that already exist in high school early on so that they are reduced or eliminated by the time students graduate from high school. A third important finding from this study is that it lends support to the social cognitive theory (Bandura, 1986) and SCCT (Lent et al., 1994). The SCCT is an elaborate theory and most prior research has only tested part of the theory and its assumptions (van Tuijl & van der Molen, 2015). The present study, however, tests all of the relationships in the theory and found that 9th grade students’ future career choice is influenced by the interest and goals the student develops and through the complex interplay between the individual, environment, and behavior (Lent et al., 1994; Lent & Brown, 1996). A final critical issue that this study addresses focuses on the exploration of possible interactions among individual, school, and Pygmalion factors. The examination of the significant interactions yielded several interesting results that need to be further explored in follow-up studies. We found, for example, that black students with higher math efficacy are more likely to major in STEMrelated fields in college than Asian students with similar level of math efficacy. This suggests that schools may need to specifically target programs to improve black students’ math efficacy. We also found a significant interaction for parent expectations and science efficacy. Students with higher science efficacy were found to be more likely to choose STEM majors compared to their peers with lower science efficacy but similar levels of parent expectations. These significant


interactions highlight the importance of examining math and science efficacy in combination with other individual, school, and Pygmalion factors. On the other hand, some of the significant interactions yielded surprising results. For example, the significant interaction of summer camp by science-efficacy students suggests that for students with similar levels of science efficacy, those who participated in summer camp are less likely to major in STEM areas than those students who did not participate in STEM summer camp. Follow-up studies are needed to explore why summer camp may have a deleterious effect. Similarly, we uncovered a puzzling finding in that students’ math efficacy had a negative interaction with parent expectations. This surprising finding needs to be explored in further studies to determine if there is an optimal range for parent expectations. Limitations and Future Research One of the limitations of this study is that it was conducted in one large school district that was implementing an integrated STEM curriculum for several years. For future studies, it would be interesting to include other large school districts so that we could possibly compare across districts on the how the three factors of (a) demographics, (b) school and out-of-school factors, and (c) Pygmalion effect variables such as student, teacher, parent expectations and students’ math and science efficacy differentially affect students’ choosing STEM majors in college. Additional studies could also include other research methods such as interviewing students, teachers, and school administrators to address research questions that focus on other personal, school, and out-of-school factors that may have motivated students to choose STEM majors in college. Conclusions This study makes several important contributions to the field. First, the design of the study included a large, representative sample of students from one of largest multiethnic charter school districts in the USA. The 70% response rate of participants is much higher that other similar studies. Second, the instrument used in the present study is quite comprehensive and includes relevant variables that adequately test the social cognitive career theory (SCCT). The actual testing of the SCCT may be the most important contribution of this study. Our findings indicate that at least one individual, environmental, Pygmalion, and motivational factor significantly influenced students’ choice of STEM major. In other words, it suggests that research on students’ interest in STEM should try to address all four of these factors in order to comprehensively examine the construct. It is important to acknowledge that this study focuses on a particular context where STEM education has been emphasized throughout the school district’s curriculum. It may be more difficult to generalize the findings from the present study to other school districts where STEM education has not been such a primary focus. The challenge of increasing students’ interest in STEM and pursuing STEM as a career is daunting. The findings of this study, however, are promising in that they suggest that they there are several school-based practices and motivational strategies that can successfully close the STEM opportunity gap.


Appendix


References Andersen, L., & Ward, T. J. (2014). Expectancy-value models for the STEM persistence plans of ninth-grade, high-ability Students: A comparison between black, Hispanic, and white students. Science Education, 98(2), 216–242. Anderson, E., & Kim, D. (2006). Increasing the success of minority students in science and technology. Washington, DC: American Council on Education. Bandura, A. (1986). Social foundations of thought and action: A social cognitive theory. Englewood Cliffs, NJ: Prentice Hall. Barber, M., & Mourshed, M. (2007). How the world’s best-performing schools systems come out on top. London, UK: McKinsey & Company. Bowen, W. G., Chingos, M. M., & McPherson, M. S. (2009). Crossing the finish line: Completing college at America’s public universities. Princeton, NJ: Princeton University Press. Chen, X., & Solder, M. (2013). STEM attrition: College students’ paths into and out of STEM fields (NCES 2014-001). Washington, DC: U.S. Department of Education and National Center for Education Statistics. Crosnoe, R., & Muller, C. (2014). Family socioeconomic status, peers, and the path to college. Social Problems, 61(4), 602–624. Davis-Kean, P. E. (2005). The influence of parent education and family income on child achievement: The indirect role of parental expectations and the home environment. Journal of Family Psychology, 19(2), 294–304. Dawes, L. A., Long, S., Whiteford, C., & Richardson, K. (2015). Why are students choosing STEM and when do they make their choice? In A. Oo & A. Patel (Eds.), Proceedings of 26th Annual Conference of the Australasian Association for Engineering. Geelong, Australia: Deakin University. DeWitt, J., Archer, L., Osborne, J., Dillon, J., Willis, B., & Wong, B. (2011). High aspirations but low progression: The science aspirations–careers paradox amongst minority ethnic students. International Journal of Science and Mathematics Education, 9(2), 243–271. Eccles, J. S. (2005). Studying gender and ethnic differences in participation in math, physical science, and information technology. New Directions for Child and Adolescent Development, 2005(110), 7–14. Eccles, J. S., Midgley, C., Wigfield, A., Buchanan, C. M., Reuman, D., Flanagan, C., & Mac Iver, D. (1993). Development during adolescence: The impact of stage-environment fit on young adolescents’ experiences in schools and in families. American Psychologist, 48(2), 90–101. Eccles, J. S., & Wigfield, A. (2002). Motivational beliefs, values, and goals. Annual Review of Psychology, 53(1), 109–132. Ehrenberg, R. G. (2010). Analyzing the factors that influence persistence rates in STEM field, majors: Introduction to the symposium. Economics of Education Review, 29(6), 888–891. Elliott, R., Strenta, A. C., Adair, R., Matier, M., & Scott, J. (1996). The role of ethnicity in choosing and leaving science in highly selective institutions. Research in Higher Education, 37(6), 681–709. Entwisle, D. R., & Hayduk, L. A. (1988). Lasting effects of elementary school. Sociology of Education, 6(3), 147–159. European Commission. (2004). Europe needs more scientists: Report by the high level group on increasing human resources for science and technology. Brussels, Belgium: Author. Fan, X., & Chen, M. (2001). Parental involvement and students’ academic achievement: A meta-analysis. Educational Psychology Review, 13(1), 1–22. Gottfried, M. A., & Williams, D. (2013). STEM club participation and STEM schooling outcomes. Education Policy Analysis Archives, 21(79), 1–27. Graham, M. J., Frederick, J., Byars-Winston, A., Hunter, A., & Handelsman, J. (2013). Increasing persistence of college students in STEM. Science, 341(6153), 1455–1456. Gunderson, E. A., Ramirez, G., Levine, S. C., & Beilock, S. L. (2012). The role of parents and teachers in the development of gender-related math attitudes. Sex Roles, 66(3–4), 153–166. Hansen, M. (2014). Characteristics of schools successful in STEM: Evidence from two states’ longitudinal data. The Journal of Educational Research, 107(5), 374–391. Hom, E. J. (2014). What is STEM education? Retrieved from http://www.livescience.com/43296-what-isstem-education.html Hoogstra, L. A. (2002). Divergent paths after high school: Transitions to college and work (Unpublished doctoral dissertation). Illinois: University of Chicago. Howard, K. A., Carlstrom, A. H., Katz, A. D., Chew, A. Y., Ray, G. C., Laine, L., & Caulum, D. (2011). Career aspirations of youth: Untangling race/ethnicity, SES, and gender. Journal of Vocational Behavior, 79(1), 98–109.

Author's personal copy A. Sahin et al. Ibrahim, A., Aulls, M. W., & Shore, B. M. (2016). Teachers’ roles, students’ personalities, inquiry learning outcomes, and practices of science and engineering: The development and validation of the McGill Attainment Value for Inquiry Engagement Survey in STEM disciplines. International Journal of Science and Mathematics Education, 1–21. https://doi.org/10.1007/s10763-016-9733-y. Ing, M. (2014). Gender differences in the influence of early perceived parental support on student mathematics and science achievement and STEM career attainment. International Journal of Science and Mathematics Education, 12(5), 1221–1239. Ingels, S. J., Dalton, B., Holder Jr, T. E., Lauff, E., & Burns, L. J. (2011). The High School Longitudinal Study of 2009 (HSLS: 09): A first look at Fall 2009 ninth-graders (NCES 2011–327). Washington, DC: National Center for Education Statistics. Retrieved from http://nces.ed.gov/pubs2011/2011327.pdf. Kline, R. B. (2015). Principles and practice of structural equation modeling (4th ed.). New York, NY: Guilford publications. Kong, X., Dabney, K. P., & Tai, R. H. (2013). The association between science summer camps and career interest in science and engineering. International Journal of Science Education, Part B, 4(1), 54–65. Lee, S. W., Min, S., & Mamerow, G. P. (2015). Pygmalion in the classroom and the home: Expectation’s role in the pipeline to STEMM. Teachers College Record, 117(9), 1–36. Lent, R. W., & Brown, S. D. (1996). Social cognitive approach to career development: An overview. The Career Development Quarterly, 44(4), 310–321. Lent, R. W., & Brown, S. D. (2006). On conceptualizing and assessing social cognitive constructs in career research: A measurement guide. Journal of Career Assessment, 14(1), 12–35. Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and academic interest, choice, and performance. Journal of Vocational Behavior, 45(1), 79–122. Ma, X., & Johnson, W. (2008). Mathematics as the critical filter: Curricular effects on gendered career choices. In H. M. G. Watt & J. S. Eccles (Eds.), Gender and occupational outcomes: Longitudinal assessments of individual, social, and cultural influences (pp. 55–83). Washington, DC: American Psychological Association. https://doi.org/10.1037/11706-002. Maltese, A. V., & Tai, R. H. (2011). Pipeline persistence: Examining the association of educational experiences with earned degrees in STEM among US students. Science Education, 95(5), 877–907. Meyers, L. S., Gamst, G., & Guarino, A. J. (2006). Applied multivariate research: Design and interpretation (3rd ed.). Thousand Oaks, CA: Sage. Miller, J. D. (2014). Longitudinal Study of American Youth: User’s manual—student, parent, and school data for cohorts 1 and 2 (1987-2011). Ann Arbor, MI: Inter-university Consortium for Political and Social Research, University of Michigan. Retrieved from http://www.icpsr.umich.edu/icpsrweb/ICPSR/studies/30263. Morgan, P. L., Farkas, G., Hillemeier, M. M., & Maczuga, S. (2016). Science achievement gaps begin very early, persist and are largely explained by modifiable factors. Educational Researcher, 45, 18–35. Mujtaba, T., & Reiss, M. J. (2014). A survey of psychological, motivational, family and perceptions of physics education factors that explain 15-year-old students’ aspirations to study physics in post-compulsory English schools. International Journal of Science and Mathematics Education, 12(2), 371–393. Museus, S. D., Palmer, R. T., Davis, R. J., & Maramba, D. C. (2011). Racial and ethnic minority students’ success in STEM education. ASHE Higher Education Report, 36(6), 1–140. My College Options (2012). Where are the STEM Students? What are their career Interests? Where are the STEM jobs? Retrieved from http://www.discoveryeducation.com/feeds/www/media/images/stemfaceacademy/Why_STEM_Students_STEM_Jobs_Full_Report.pdf. Nagy, G., Trautwein, U., Baumert, J., Köller, O., & Garrett, J. (2006). Gender and course selection in upper secondary education: Effects of academic self-concept and intrinsic value. Educational Research and Evaluation, 12(4), 323–345. Nakakoji, Y., Wilson, R., & Poladian, L. (2014). Mixed methods research on the nexus between mathematics and science. International Journal of Innovation in Science and Mathematics Education, 22(6), 61–76. National Research Council. (2013). Monitoring progress toward successful K-12 STEM education: A nation advancing? Washington, DC: The National Academies Press. National Science Board. (2007). A national action plan for addressing the critical needs of the US science, technology, engineering, and mathematics education system. Arlington, TX: National Science Foundation. National Science Foundation (US). (2010). Preparing the next generation of stem innovators: Identifying and developing our nation’s human capital. Arlington, TX: National Science Foundation. Naviance (2016). Connecting learning & life. Retrieved from http://www.naviance.com/ Neuenschwander, M. P., Vida, M., Garrett, J. L., & Eccles, J. S. (2007). Parents’ expectations and students’ achievement in two western nations. International Journal of Behavioral Development, 31(6), 594–602.

Author's personal copy Collective Effects of Individual, Behavioral, and Contextual... Office of the Chief Scientist. (2012). Mathematics, engineering & science in the national interest. Canberra, Australia: Commonwealth of Australia Retrieved from http://www.chiefscientist.gov.au/2012/05/mesreport/. Organisation for Economic Co-operation and Development. (2014). TALIS 2013 results: An international perspective on teaching and learning. Paris, France: OECD Publishing. https://doi.org/10.1787 /9789264196261-en. Parker, P. D., Schoon, I., Tsai, Y. M., Nagy, G., Trautwein, U., & Eccles, J. S. (2012). Achievement, agency, gender, and socioeconomic background as predictors of postschool choices: A multicontext study. Developmental Psychology, 48(6), 1629–1642. President’s Council of Advisors on Science and Technology. (2012). Engage to excel: Producing one million additional college graduates with degrees in science, technology, engineering, and mathematics. Washington, DC: Executive Office of the President and President’s Council of Advisors on Science and Technology. Price, J. (2010). The effect of instructor race and gender on student persistence in STEM fields. Economics of Education Review, 29(6), 901–910. Riegle-Crumb, C., Farkas, G., & Muller, C. (2006). The role of gender and friendship in advanced course taking. Sociology of Education, 79(3), 206–228. Rodriguez, E., Rhodes, K., & Aguirre, G. (2015). Intervention for high school Latino students in preparing for college: Steps for consideration. Journal of Hispanic Higher Education, 14(3), 207–222. Rosenthal, R., & Jacobson, L. (1968). Pygmalion in the classroom. The Urban Review, 3(1), 16–20. Rutkowski, L., Gonzalez, E., Joncas, M., & von Davier, M. (2010). International large-scale assessment data issues in secondary analysis and reporting. Educational Researcher, 39(2), 142–151. Sahin, A., & Top, N. (2015). STEM students on the stage (SOS): Promoting student voice and choice in STEM education through an interdisciplinary, standards-focused, project based learning approach. Journal of STEM Education: Innovations and Research, 16(3), 24–33. Sass, T. R. (2015). Understanding the STEM pipeline (CALDER #125). Washington, DC: American Institutes for Research. Schunk, D. H., Meece, J. R., & Pintrich, P. R. (2014). Motivation in education: Theory, research, and applications (4th ed.). New York, NY: Pearson. Seymour, E., & Hewitt, N. M. (1997). Talking about leaving: Why undergraduates leave the sciences. Boulder, CO: Westview Press. Simpkins, S. D., Davis-Kean, P. E., & Eccles, J. S. (2006). Math and science motivation: A longitudinal examination of the links between choices and beliefs. Developmental Psychology, 42(1), 70–83. Tai, R. H., Liu, C. Q., Maltese, A. V., & Fan, X. (2006). Planning early for careers in science. Science, 312, 1143–1144. The Alliance for Science & Technology Research in American (2015). Texas’ 2015 STEM report card. Retrieved from https://www.usinnovation.org/state/pdf_cvd/ASTRA-STEM-on-Hill-Texas2015.pdf. Uitto, A. (2014). Interest, attitudes and self-efficacy beliefs explaining upper-secondary school students' orientation towards biology-related careers. International Journal of Science and Mathematics Education, 12(6), 1425–1444. Van Tuijl, C., & van der Molen, J. H. W. (2015). Study choice and career development in STEM fields: An overview and integration of the research. International Journal of Technology and Design Education, 26(2), 159–183. https://doi.org/10.1007/s10798-015-9308-1. Wang, X. (2013). Why students choose STEM majors: Motivation, high school learning, and postsecondary context of support. American Educational Research Journal, 50(5), 1081–1121. https://doi.org/10.3102 /0002831213488622. Watt, H. M. G. (2008). Gender and occupational outcomes: An introduction. In H. M. G. Watt & J. S. Eccles (Eds.), Gender and occupational outcomes: Longitudinal assessments of individual, social, and cultural influences (pp. 3–24). Washington, DC: American Psychological Association. Yu, S. L., Corkin, D. M., & Martin, J. P. (2016). STEM motivation and persistence among underrepresented minority students: A social cognitive perspective. In J. T. DeCuir-Gunby & P. A. Schutz (Eds.), Race and ethnicity in the study of motivation in education (pp. 67–81). New York, NY: Taylor & Francis.