







































Children and Teenagers 
ISSN 2576-3709 (Print) ISSN 2576-3717 (Online) 

Vol. 1, No. 1, 2018 
www.scholink.org/ojs/index.php/ct 

49 
 

Original Paper 

Secondary School Variables in Predicting Technology, 

Engineering, Mathematics (TEM) Major Choice 
Philip Ruse1* & Yonghong Xu1 

1 Department of Educational Psychology and Research, University of Memphis, Memphis, USA 
* Philip Ruse, Department of Educational Psychology and Research, University of Memphis, Memphis, 

USA 

 
Received: April 19, 2018          Accepted: May 9, 2018         Online Published: May 21, 2018 

doi:10.22158/ct.v1n1p49                          URL: http://dx.doi.org/10.22158/ct.v1n1p49 

 
Abstract 

This paper addresses the growing body of research into factors that can influence the decision for high 

school students to enter into a Technology, Engineering, Mathematics (TEM) major in college. A total n 

of 691, including 372 TEM majors (143 females and 229 males) were selected from the Education 

Longitudinal Study of 2002 (ELS, 2002) using propensity matching. A Structural Equation Modeling 

(SEM) methodology was utilized in the Social Cognitive Career Theory framework and showed good 

model fit in the whole group, female only and male only groups. Though intent to major was a strong 

predictor, observed gender differences were observed related to latent and endogenous variables. 

Keywords 

social cognitive career theory, major choice, secondary school, STEM 

 

1. Introduction 

The need for researchers to study the enrollment of students into Science, Technology, Engineering, and 

Mathematics (STEM) majors is predicated on two fundamental aspects: the individual benefits of such a 

degree and the collective need for the degree. This study aims to add to the growing body of research 

regarding the gender gap in STEM, specifically in computer science (technology), engineering, and 

mathematics (TEM). The ability of the country to remain competitive in specialized job markets, such as 

TEM, requires that policy makers understand the factors that lead students into a STEM field, to better 

meet the demands of a changing job market (Chen, 2013; Stater, 2011). In order to meet the growing 

demand of TEM related careers tapping into the under-represented groups of women and racial 

minorities becomes paramount. It is clear that secondary schools foster the development of student 

academic interest and subsequent career choices, therefore it is prudent to analyze the influence the 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
50 

Published by SCHOLINK INC. 

 

secondary school environment plays. The primary goal of the paper is to identify high school level 

factors that are viable indicators of student major choice into Technology, Engineering, and Mathematics 

(TEM) utilizing the Social Cognitive Career theoretical framework. An additional goal is to identify if 

any gender difference exists in the factors that lead to a TEM major. 

A multitude of studies have been dedicated to illuminating gender differences, with special focus on 

educational differences as they relate to the gender gap in STEM occupations, college majors, and high 

school interests (Blickenstaff, 2005; Ceci & Williams, 2010; Ceci & Williams, 2011; Riegle-Crumb, 

King, Grodsky, & Muller, 2012; Sax et al., 2015). Due to the growing need of STEM occupations 

anticipated for the coming years, policy makers must be clear on how these differences manifest 

themselves in STEM disciplines, what the reasons are for such a gender disparity in TEM, and why does 

it become critical to decrease the gender gap in TEM. One significant reason for reducing the gender gap 

is the potential economic benefit. The sustainable financial future of TEM careers could provide a viable 

path to economic mobility in under-represented groups. 

The statistics surrounding female involvement in STEM are somewhat disappointing. Though females 

are conferred degrees at a near equal proportion to males, there is a large discrepancy the matriculation of 

STEM degrees (Mann & DiPrete, 2013). Only 33% of females intend to major in STEM, 12% less than 

males and 3 times less likely to enter the majors (Moakler & Kim, 2014; National Science Board [NSB], 

2014). The explanation of the disproportionate enrollment in TEM fields by women was produced by 

Blickenstaff (2005). It was suggested that 9 possible reasons were influencing women interest in TEM, of 

the 9 one was attributed to biological differences and the other 8 were categorized as cultural exclusion of 

women in science. With the growing need of STEM professionals in the coming years it becomes critical 

to understand the reasons behind the gender disparity in TEM and make efforts to decrease it. 

The gender differences become apparent in academic indicators; females score lower on both math and 

science, 4% and 6% respectively (NEAP, 2009; NSB, 2014). More troubling is that the difference goes 

beyond achievement and enters self-concept. When females have similar math achievement their 

self-concept scores lower than male students, thus females are less likely to enroll in upper level math 

(Correll, 2001; Mann & DiPrete, 2013; Nagy et al., 2006; Sax et al., 2015; Wang, Degol, & Ye, 2015). 

Understanding these differences is even more vital because subsequent major choice is thought to be 

predicated on achievement, self-efficacy, and course selection (Lent, Brown, & Hackett, 2004; Wang, 

2012). 

There are multiple needs for research specifically into TEM. First, the greatest need for individuals is in 

TEM fields rather than STEM. TEM offers substantial income and professional opportunities compared 

to other STEM majors (NSB, 2014). This is made apparent by the lower unemployment rate and higher 

salaries of individuals when juxtaposed to lab-based sciences. Furthermore, the lab-based sciences have 

the lowest percentage of graduates employed in occupations related to their majors indicating that a 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
51 

Published by SCHOLINK INC. 

 

surplus may exist in lab-sciences (Xue & Larson, 2015; U.S. Census Bureau, 2014). 

Second, there is a more significant gender gap in TEM. Though women have continuously increased 

enrollment in STEM majors and occupations the growth is seemingly relegated to the biological 

science. Though women make up 24% of STEM professionals they are grossly under-represented in 

engineering (14%) and computer science/mathematics (27%) (Beede et al., 2011). Female students tend 

to place greater focus on social aspects rather than the material, this difference in value helps to explain 

why women preferentially select biology majors over the more object-oriented engineering and 

computer science majors (Ceci & Williams, 2010; Ceci & Williams, 2011; Wang, Degol, & Ye, 2015). 

In fact, males and females show no statistical difference between attitudes toward biology, however the 

same cannot be said about technology, math, and engineering. Men report a more positive attitude 

towards TEM fields and as a result, continue to dominate engineering and computer sciences. This has 

continued to the point that, though the participation of women in STEM has grown, there has been a 

decrease over the last several years in choosing to enter TEM majors (Christensen et al., 2014; Maltese 

& Tai, 2010; Riegle-Crumb, King, Grodsky, & Muller, 2012). This trend continues into the workforce 

where women make up 33% of the STEM professionals (NSF, 2012). Women workers are considerably 

under-represented in engineering (15%) and computer science/mathematics (25%). Where as in 

laboratory sciences women make up 48% and physical sciences 31%. The 2 categories that have seen 

the greatest change in participation since 1993 are a 5% reduction of females in math and computer 

science and a 14% increase in biological sciences (NSF, 2012). It is this that resonates with researchers 

and policymakers because the majority of job growth is expected to be in the technology, engineering, 

and computer science areas. 

Finally, the focus of this study on TEM is also determined by the nature of the data set used. As discussed 

later, the 2002 ELS surveyed math teachers, making the data more conducive to study TEM.  

Based on a conceptual understanding of the research questions and a review of the literature, Social 

Cognitive Career Theory (SCCT) provides a logical theoretical framework in which to create the model 

(Carrico & Tendhar, 2012; Lee, 2013; Lent et al., 2008; Maltese & Tai, 2010; Moakler & Kim, 2014; Sax 

et al., 2015; Sheu et al., 2010; Wang, 2012). Developed by Lent, Brown, and Hackett (1994) as an 

extension of Bandura’s Social Cognitive Career theory, SCCT was designed to explain the career and 

academic choices made by individuals. By incorporating self-efficacy, outcome expectation, interests, 

and choice goals SCCT is a comprehensive theory that provides a conceptually salient model that allows 

for the incorporation of the most critical constructs in the development of vocation and academic 

interests (Lent, Brown, & Hackett, 1994).  

In a review of the literature related to college major choice, SCCT is the dominant theoretical 

framework utilized by researchers (Carrico & Tendhar, 2012; Lee, 2013; Lent et al., 2008; Maltese & 

Tai, 2010; Moakler & Kim, 2014; Sax et al., 2015; Sheu et al., 2010; Wang, 2013). Further to the point 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
52 

Published by SCHOLINK INC. 

 

of this work SCCT has been shown to be effective in describing gender differences as well as in 

longitudinal data sets (Lent et al., 2008). Another advantage of SCCT is that it allows the researcher to 

investigate specific actions of individuals, i.e., the selection of a specific major (Carrico & Tendhar, 

2012; Lent et al., 2008). The ability of SCCT to elucidate the characteristics that influence college 

major selection and the flexibility of the inclusion of variables makes it a prudent choice for a 

theoretical framework. 

The foundation of SCCT is how self-efficacy, outcome expectation, interests, and choice goals 

influence both vocational and academic choice. It is customary for researchers to include supporting 

variables of supports and barriers, learning experiences, and intention of goal fulfillment in the model 

as well (Carrico & Tendhar, 2012; Garriott, Flores, & Martens, 2013; Lent et al., 2008; Sax et al., 2015; 

Sheu et al., 2010; Wang, 2013). Self-efficacy can be regarded as an individual’s personal belief about 

their ability to complete tasks related to achieving a particular level of success. This is related to 

maintaining actions to attain the goal through adversity. An outcome expectation is understood to be the 

perceptions that one has about the consequences of engaging in and completing an action. Social or 

cultural pressures placed on a woman in science could influence outcome expectation. Interests can be 

thought of as a measure of an individual’s enjoyment of a particular activity. Goals are positive future 

outcomes that individuals have undertaken legitimate steps to attaining and committed resources to the 

successful fulfillment of them (Carrico & Tendhar, 2012; Lent, Brown, & Hackett, 1994). 

1.1 Research Questions 

SCCT must be confirmed as a viable framework to study the factors associated with student entrance into 

TEM. Additionally, there is a need to improve the participation of under-represented groups in STEM 

fields; women attack problems in ways that are substantially different than men. Low female 

participation in TEM reduces the human capital the result of which can be detrimental to the 

technological and scientific growth of the nation (Sax et al., 2015).  

Therefore, this study will address: 

1) How are students impacted by academic experiences, their own self-efficacious beliefs of math and 

science, and the value of such a degree relate to their entrance into a TEM major? 

2) Do these relationships vary across gender?  

 

2. Method 

Data was acquired from the Education Longitudinal Study of 2002 (ELS, 2002). Base year data from 

2002 was collected from high school sophomores with subsequent follow-up surveys conducted in 

2004, 2005, 2006, 2012, and 2013. These surveys were administered to students, parents, teachers, and 

school support staff and consisted of an academic focus on math and English. ELS: 2002 was a suitable 

data set to use for this analysis because it connected high school level variables to subsequent college 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
53 

Published by SCHOLINK INC. 

 

major choice. In addition, the focus on math teachers and attributes was congruent with the point of this 

study. The original ELS data set consisted of 16197 individuals (47.2% male, 47.6% female), of which 

approximately 6500 reported attending post-secondary education in the 2006 follow-up. Of these 

respondents, only 10% reported enrollment in a TEM major. Further cleaning of the data was 

accomplished using case wise deletion on the utilized variables bringing the total N to 3254, with 10% 

TEM (44% male, 66% female).  

Before conducting the SEM analysis additional data management was required to bring the two groups 

into a more balanced data set. Matching the TEM majors (N = 327) to the non-TEM majors (N = 2927) 

using propensity scores can provide an effective way to balance the data and continue with the analysis. 

This method is further validated by Parsons (2004) who explains that, in order to detect differences 

between experimental and control groups there must be a reduction in the treatment selection bias. To 

maximize the number of TEM majors, propensity scores were generated via SAS (Parson, 2004; 

Relyea, 2016). All TEM cases were retained and 473 non-TEM majors with the highest propensity 

scores were retained as matches. The final total was N = 709, 79% male and 236 (33%) TEM majors. 

Outliers were removed using from the 709 individuals using Mahalonibis distance (χ2 p<.001) for a 

final N=690. 

2.1 Variables 

2.1.1 Math Self-Efficacy 

Self-efficacy, according to Lent, Brown, and Hackett (1994) is one of the pillars of SCCT and primary 

indicator of interest in academic and professional goals. Subject specific self-efficacy is an essential 

predictor of STEM interest and major choice; many more studies that could be named that implement 

SCCT include it as a predictor (Engburg & Wolniak, 2013; Moakler & Kim, 2013; Sax et al., 2015; 

Wang, 2012). Following the guidelines of SCCT self-efficacy has a path to outcome expectations, 

interest, intent, and major selection (Lent, Brown, & Hackett, 1994). 

For this latent variable 6 factors were included. One continuous measure of self-efficacy score from 

10th grade. In addition, five measures of 12th grade attitudes toward math were also included. These 

measures were rated on a 4-point Likert scale and were comprised of statements such as, can do 

excellent job on math assignments or can understand difficult math class. 

2.1.2 Outcome Expectations 

An outcome expectation as a latent variable has a direct path to interest, intention, and major choice 

(Lent, Brown, & Hackett, 1994). The value of intrinsic and extrinsic rewards along with expectation of 

graduate degree are factors that are consistent with other studies to load onto outcome expectations 

(Mann & Diprete, 2013). 

Three measures taken in the 2006 senior follow-up were used for the latent variable outcome 

expectations. The questions were based on a 3-point scale from not important (1) to very important (3). 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
54 

Published by SCHOLINK INC. 

 

These survey items inquired about the relative importance of success in jobs, the importance of making 

a lot of money, and the importance placed on family and children. 

2.1.3 Learning Experience 

Math and science achievement in high school is a strong predictor of interest in STEM and the later 

choice of major in college (Engburg & Wolniak, 2013; Maltese & Tai, 2010; Riegle-Crumb, King, 

Grodsky, & Muller, 2012; Wang, 2012). Achievement is measured at 10th grade and 12th grade using 

the ELS: 2002 data. The third measure, teacher recommendation for advanced courses in mathematics 

(1-3 not recommend to highly). 

2.1.4 Interest 

Previous studies suggest that course selection should be a predictor of interest in STEM. Students who 

enroll in upper level math and science courses as upper classmen in high school are more likely to have 

a greater interest in STEM and thus choose a STEM major (Correll, 2001; Engburg & Wolniak, 2013; 

Lee, 2013; Maltese & Tai, 2010; Mann & Diprete, 2013; Riegle-Crumb, King, Grodsky, & Muller, 

2012; Wang, 2012). 

Two variables were created to quantify the number of math and science electives were taken in high 

school. Upper level math course electives were number 0-4 to include courses trigonometry, 

pre-calculus, calculus, and business math. The science courses included were only physics and 

technology courses numbered 0-2. Only TEM science courses were included due to the scope of this 

study. 

2.1.5 Supports and Barriers 

The use of SES and the highest level of parent degree have both been used as supports and barriers in 

previous SEM studies (Garriot, Flores, & Martens, 2013; Sax et al., 2015; Wang, 2012). Loading the 

supports and barriers onto major choice is congruent with literature on STEM major choice within the 

SCCT framework. 

Three variables were loaded onto supports and barriers. A measure of SES, (BYSES2QU) which 

divided students into quartiles 1-4. The parents highest level of education (BYPARED) recorded as 1-8, 

did not graduate high school to doctorate respectively. Also included was the number of academic risk 

factors, this variable was recoded 1-6 in order to reflect 1 having 5 or more risk factors and 6 having 0 

risk factors.  

2.1.6 Intent 

Intent is simply a measure of an individual’s perspective major. This was coded 1 = TEM major and 

0=non-TEM major.  

2.1.7 Dependent Variable 

A dichotomous variable was recorded as 0 = Non-TEM and 1 = TEM major. Lab-based sciences such as 

chemistry or biology were excluded from the study variable (Table 1).  



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
55 

Published by SCHOLINK INC. 

 

Table 1. Variable Descriptions in Model 

*. Reverse coded. 

 

3. Result 

Four models were ran and model fit was assessed using Kline’s (2011) guidelines. The measurement 

model suggests a strong fit with a RMSEA = 0.027, with a 90% confidence interval upper bound below 

0.08 and CFI = 0.99 (see Table 2). All factor loadings were above 0.24 (Table 1). Three SEM analyses 

were ran, one assessing whole group and one each for males and females. Fit indices are provided in 

Table 2. 

 

Latent Variable 

Observed 

Variable 

Factor 

Loading Description Scale 

Learning 

Experience 

(L_E) 

BYTXMSTD 0.87 Math test standardized score (10 grade) Cont 

F1TXMSTD 0.98 Math test standardized score (12 grade) Cont 

BYTM19 0.44 Teacher recommendation for upper math 1-3 

Self-Efficacy 

(S_E) 

BYMATHSE 0.49 F1 mathematics self-efficacy Cont 

F1S18A 0.82 Can do excellent job on math tests 1-4 

F1S18B 0.83 Can understand difficult math texts 1-4 

F1S18C 0.82 Can understand difficult math class 1-4 

F1S18D 0.81 Can do excellent job on math assignments 1-4 

F1S18E 0.87 Can master math class skills 1-4 

Interest 

(Int) 

Math_ele 0.66 
Math electives (trig, precal, calc, Business math 

recoded as number of courses 0-4) 
0-4 

Sci_tech 0.53 
Science electives (Physics and technology)  

recoded as number of courses 0-2) 
0-2 

Outcome 

Expectation 

(O_E) 

F1S40A 0.44 Importance of being successful in line work 1-3 

F1S40C 0.47 Importance of having lots of money 1-3 

F1S40E 0.58 Importance of being able to find steady work 1-3 

Intent F2B15 1.00 
Field of study most likely to pursue upon 

entering 
0/1 

Supports and 

Barriers (S_B) 

BYPARED 0.87 Parents' highest level of education 1-8 

BYSES2QU 0.92 Quartile coding of SES2 variable 1-4 

BYRISKFC* 0.24 
Number of academic risk factors in 10th grade 

(recoded) 
1-6 

Major MJR_TEM 1.00 Major in 2006 2-digit code 0/1 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
56 

Published by SCHOLINK INC. 

 

Table 2. Goodness of Fit Indices for Measurement and Structural Models 

Model χ2 df χ2/df CFI RMSEA 

CI for 

RMSEA SRMR 

Measurement 185.16* 123 1.505 .99 .027 (.019, .035) .036 

Structural group 220.11* 126 1.580 .99 .033 (.026, .040) .047 

Structural male 199.08* 126 1.401 .99 .033 (.024, .041) .049 

Structural female  131.74 126 1.045 .99 .018 (.000, .046) .064 

*. P<.05. 

 

The whole group model demonstrated a strong fit to the data, RMSEA = 0.033 (90% CI = 0.026; 0.040) 

and the CFI = 0.99. The model fit indices fall within the suggested limits of RMSEA<0.06 and the 

CFI>.95 to show strong model fit (Kline, 2011). All path coefficients were significant (p<0.05) except 

learning experience to self-efficacy and to outcome expectation as well as self-efficacy to interest. The 

model explained 26% of the variance in Major. 

The model was then applied to only the males in the sample (TEM = 33% of 548). The model fit the 

sample well with RMSEA = 0.033 (90% CI = 0.024; 0.041) and a CFI = 0.99. The path coefficients had 

the same significance as the whole group. The variance explained in major saw a 11% increase to 37%. 

The same model was then ran using only females from the sample (TEM = 33% of 143). Again, the 

model fit was very strong, RMSEA = 0.018 (90% CI = 0.0; 0.046) and the CFI = 0.99. Learning 

experience showed a significant path to outcome expectation, which was not seen in the other models. 

The variance explained in major was reduced to 30%.  

 

Table 3. Covariance Matrix of Latent Variables (Whole Group) 

 1. 2. 3. 4. 5. 6. 7. 

1. S_E 0.19       

2. O_E 0.00 0.01      

3. Interest 0.10 0.00 0.53     

4. Major 0.00 0.00 0.02 0.22    

5. Intent 0.02 -0.01 0.04 0.12 0.24   

6. L_E 1.27 -0.18 0.66 0.05 0.21 56.33  

7. S_B 0.06 -0.01 0.03 -0.01 0.01 1.63 0.73 

 

 

 

 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
57 

Published by SCHOLINK INC. 

 

Table 4. Structural Equations of Unstandardized Path Coefficients (Whole Group) 

Latent Variable Structural Equation R2 

S_E 0.022*L_E + 0.029*S_B .15 

O_E -0.0034*S_E – 0.0031*L_E 0.045 

Int 0.50*S_E – 0.17*O_E 0.091 

Intent 0.080*S_E – 0.35*O_E + 0.066*Int  0.028 

Major -0.033*S_E – 0.063*O_E + 0.49*Intent – 0.013*S_B 0.26 

 

4. Discussion 

This paper continues to add to the growing body of literature dedicated to illuminating the influence of 

the high school experience on the decision to enter into a TEM major as well as identifying variables 

associated with the gender gap in TEM majors. Through the strong model fit of the SEM analysis in both 

the whole group and male group SCCT is clearly an adequate choice to identify such variables in 

predicting a college major choice. The female group had strong model fit but was not significant, which 

was likely due to the small sample size of the group. 

This paper further supports the use of SCCT to understand the factors that lead students to choosing a 

specific major. Oddly, the paths from learning experience were largely nonsignificant as was the paths 

from self-efficacy to interest, this could be due to sample composition and the relative homogeneity of 

students. It is not unprecedented to have small values from Outcome Expectations as small but 

significant paths from outcome expectation to interest have been reported in other studies (Garriott, 

Flores, & Martens, 2013). It does not come as a surprise to see academic achievement, outcome 

expectations, or self-efficacy as poor differentiators among college students. Many college students 

regardless of the major they choose will perform well in high school. The 2 largest and significant paths 

for the group were interest to intent (.103) and intent to major (.51) (Table 5). This suggests that course 

selection and intent to major are 2 of the more important and strong predictors of a student choosing to 

major in TEM. 

 

Table 5. Indirect Paths 

Path Whole Female Male 

S_EinterestIntentMajor .016 .025 .0106 

S_E IntentMajor .0357 .0324 .0488 

S_EMajor -.03 -.02 -.04 

S_E total .0217 .0374 .0194 

InterestIntentMajor .0525 .1134 .0366 

Standardized path coefficients.  



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
58 

Published by SCHOLINK INC. 

 

Several noticeable differences appeared between the genders. The only significant path from learning 

experience was to outcome expectation in females, and surprisingly it was negative. This means as math 

achievement increases outcome expectation decreases. The observed value could be a result of social and 

cultural constructs limiting female interest in TEM (Blickenstaff, 2005; Ceci & Williams, 2010; Ceci & 

Williams, 2011; Mann & DiPrete, 2013; Riegle-Crumb, King, Grodsky, & Muller, 2012; Sax et al., 2015). 

Female college students face added societal pressures to place greater value on intrinsic rewards and 

choosing a career path that favors home and family life, further widening the gender gap (Blickenstaff, 

2005; Mann & DiPrete, 2013; Sax et al., 2015). These factors, whether the cause or the effect, serve to 

place women in a “chilly climate” when pursuing STEM (Blickenstaff, 2005; Mann & DiPrete, 2013). 

All things considered, women are identified as an under-represented group in STEM (Engburg & 

Wolniak, 2013). 

Another curious difference was in supports and barriers. Females had a positive path coefficient 

compared to males’ negative value. As available resources increased females had a greater likelihood of a 

TEM major, but the same could not be said for males. The greatest difference was the path from interest 

to intent which was nearly 4 times larger for females. This resulted in greater influence from self-efficacy, 

which at .0374, was nearly double that of males. The results of this study suggest that females with 

greater exposure to TEM related courses in high school is the greatest predictor of their intent to major 

in TEM majors in college. Furthermore, interest is the greatest difference between the genders, coupled 

with the role of supports and barriers it is likely that providing women more exposure to TEM in high 

school could increase their potential decision to major in it in college. 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
59 

Published by SCHOLINK INC. 

 

              

       

       

       

    

   

   

   

   

    

  

  

  

  

      

      

      

         

 

Figure 1. Standardized Path Coefficients (Whole/Female/Male) 

Bold: p>.05 

 

Though the data set used, ELS 2002, is a large national data set there were several issues that presented 

themselves to the research questions. One such problem was that of the sample size. A small number of 

students choose to enroll in a TEM major and created an unbalanced data set. In addition, males were 

over-represented in the data set making comparison between genders difficult. This presented a 

secondary problem in that the number of females in the study is low for SEM. Adding more individuals 

to the study and increasing female representation could alleviate many issues. Additionally, a more up 

to date sample would benefit the study, the implemented data set is dated and technology has seen 

monumental changes in 15 years. SCCT is the dominant framework in determining student major 

choice, however, it is not alone. A comparison of SCCT, Pekrun’s Control Value Theory, could provide 

new and unique insights into student choice.   

 

 

 

L E 

S_E S_B 

Interest 

O_E 

Intent Major 

.37 

.38 

.37 

 

-.21 

-.25 

-.19 

 
 

-.01 

.02 

-.01 

 

-.08 

-.04 

-.08 

 

-.03 

.07 

-.02 

 

.30 

.22 

.29 

 

.07 

.06 

.08 

.103 

.21 

.06 

.51 

.54 

.61 

 

 

.06 

-.05 

.08 

 

-.02 

.09 

-.05 

 

-.03 

-.02 
-.04 

 

-.01 

-.01 

-.02 

 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
60 

Published by SCHOLINK INC. 

 

References 

Beede, D. N., Julian, T. A., Langdon, D., McKittrick, G., Khan, B., & Doms, M. E. (2011). Women in 

STEM: A gender gap to innovation. 

Blickenstaff, C. J. (2005). Women and science careers: Leaky pipeline or gender filter? Gender and 

Education, 17(4), 369-386. https://doi.org/10.1080/09540250500145072                                                                                                                                                                                                                                                                                                                                                                                                                   

Brown, S. D., Tramayne, S., Hoxha, D., Telander, K., Fan, X., & Lent, R. W. (2008). Social cognitive 

predictors of college students’ academic performance and persistence: A meta-analytic path analysis. 

Journal of Vocational Behavior, 72(3), 298-308. https://doi.org/10.1016/j.jvb.2007.09.003 

Carrico, C., & Tendhar, C. (2012). The Use of the Social Cognitive Career Theory to Predict Engineering 

Students’ Motivation in the Produced Program. In American Society for Engineering Education. 

American Society for Engineering Education. 

Ceci, S. J., & Williams, W. M. (2010). Sex differences in math-intensive fields. Current Directions in 

Psychological Science, 19(5), 275-279. https://doi.org/10.1177/0963721410383241 

Ceci, S. J., & Williams, W. M. (2011). Understanding current causes of women’s underrepresentation in 

science. Proceedings of the National Academy of Sciences, 108(8), 3157-3162. 

https://doi.org/10.1073/pnas.1014871108 

Chen, X. (2013). STEM Attrition: College students’ paths into and out of STEM fields. In Statistical 

Analysis Report. NCES 2014-001. National Center for Education Statistics. 

Christensen, R., Knezek, G., & Tyler-Wood, T. (2014). Student perceptions of Science, Technology, 

Engineering and Mathematics (STEM) content and careers. Computers in Human Behavior, 34, 

173-186. https://doi.org/10.1016/j.chb.2014.01.046 

Correll, S. J. (2001). Gender and the career choice process: The role of biased self-assessments. 

American journal of Sociology, 106(6), 1691-1730. https://doi.org/10.1086/321299 

Engberg, M., & Wolniak, G. C. (2013). College student pathways to the STEM disciplines (Doctoral 

dissertation, Teachers College). Retrieved from http://www.tcrecord.org/ 

Garriott, P. O., Flores, L. Y., & Martens, M. P. (2013). Predicting the math/science career goals of 

low-income prospective first-generation college students. Journal of Counseling Psychology, 60(2), 

200. https://doi.org/10.1037/a0032074 

Hall, C., Dickerson, J., Batts, D., Kauffmann, P., & Bosse, M. (2011). Are We Missing Opportunities to 

Encourage Interest in STEM Fields? Retrieved from http://scholar.lib.vt.edu/ 

Kline, R. B. (2011). Principles and Practice of Structural Equation Modeling. New York: Guilford Press. 

Lee, A. (2013). Determining the effects of pre-college STEM contexts on STEM major choices in 4-year 

postsecondary institutions using multilevel structural equation modeling. Journal of Pre-College 

Engineering Education Research (J-PEER), 3(2), 3. https://doi.org/10.7771/2157-9288.1059 

Lent, R. W., Brown, S. D., & Hackett, G. (1994). Toward a unifying social cognitive theory of career and 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
61 

Published by SCHOLINK INC. 

 

academic interest, choice, and performance. Journal of vocational behavior, 45(1), 79-122. 

https://doi.org/10.1006/jvbe.1994.1027 

Lent, R. W., Sheu, H. B., Singley, D., Schmidt, J. A., Schmidt, L. C., & Gloster, C. S. (2008). 

Longitudinal relations of self-efficacy to outcome expectations, interests, and major choice goals in 

engineering students. Journal of Vocational Behavior, 73(2), 328-335. 

https://doi.org/10.1016/j.jvb.2008.07.005 

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. 

https://doi.org/10.1002/sce.20441 

Mann, A., & Diprete, T. A. (2013). Trends in gender segregation in the choice of science and engineering 

majors. Social science research, 42(6), 1519-1541. 

https://doi.org/10.1016/j.ssresearch.2013.07.002 

Moakler, M. W., & Kim, M. M. (2014). College major choice in STEM: Revisiting confidence and 

Demographic factors. The Career Development Quarterly, 62(2), 128-142. 

https://doi.org/10.1002/j.2161-0045.2014.00075.x 

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. https://doi.org/10.1080/13803610600765687 

National Assessment of Educational Progress, The Nations Report Card. (2015). Mathematics and 

reading at grade 12. Retrieved from 

https://www.nationsreportcard.gov/reading_math_g12_2015/#mathematics/gaps 

National Assessment of Educational Progress, The Nations Report Card. (2015). Science Assessment. 

Retrieved from https://www.nationsreportcard.gov/science_2015/#gaps?grade=12 

National Science Board. (2014). Science and Engineering Indicators 2014. Arlington VA: National 

Science Foundation (NSB 14-01). 

Parsons, L. S. (2004). Performing a 1: N case-control match on propensity score. In Proceedings of the 

29th Annual SAS users group international conference (pp. 129-165). 

Relyea, G. (2016). Macro One to Many Match_Revised5.sas. Unpublished Protocol.  

Riegle-Crumb, C., King, B., Grodsky, E., & Muller, C. (2012). The more things change, the more they 

stay the same? Prior achievement fails to explain gender inequality in entry into STEM college 

majors over time. American Educational Research Journal, 49(6), 1048-1073. 

https://doi.org/10.3102/0002831211435229 

Sax, L. J., Kanny, M. A., Riggers-Piehl, T. A., Whang, H., & Paulson, L. N. (2015). But I’m Not Good at 

Math: The changing salience of mathematical self-concept in shaping women’s and men’s STEM 

aspirations. Research in Higher Education, 56(8), 813-842. 



www.scholink.org/ojs/index.php/ct                        Children and Teenagers                        Vol. 1, No. 1, 2018 

 
62 

Published by SCHOLINK INC. 

 

https://doi.org/10.1007/s11162-015-9375-x 

Sheu, H. B., Lent, R. W., Brown, S. D., Miller, M. J., Hennessy, K. D., & Duffy, R. D. (2010). Testing the 

choice model of social cognitive career theory across Holland themes: A meta-analytic path analysis. 

Journal of Vocational Behavior, 76(2), 252-264. https://doi.org/10.1016/j.jvb.2009.10.015 

Stater, M. (2011). Financial aid, student background, and the choice of first-year college major. Eastern 

Economic Journal, 37(3), 321-343. http://doi:10.1057/eej.2009.41                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                                      

U.S. Department of Commerce, U.S. Census Bureau. (2014). Census Bureau Reports Majority of 

STEM College Graduates Do Not Work in STEM Occupations (Release Number CB14-130). 

Retrieved from https://www.census.gov/newsroom/press-releases/2014/cb14-130.html 

Wang, M. T., Degol, J., & Ye, F. (2015). Math achievement is important, but task values are critical, too: 

Examining the intellectual and motivational factors leading to gender disparities in STEM careers. 

Frontiers in Psychology, 6. https://doi.org/10.3389/fpsyg.2015.00036 

Wang, X. (2013). Why students choose STEM majors motivation, high school learning, and 

postsecondary context of support. American Educational Research Journal. 

http://doi:10.3102/00028312134488622 

Xue, Y., & Larson, R. C. (2015). STEM crisis or STEM surplus: Yes and yes. Monthly Lab. Rev., 

138(1). https://doi.org/10.21916/mlr.2015.14 

 


