Home Articles Reader Opinion Editorial Book Reviews Discussion Writers Guide About TCRecord
transparent 13
Topics
Discussion
Announcements
 

The Road to Becoming a Scientist: A Mixed Methods Investigation of Supports and Barriers Experienced by First-Year Community College Students


by Xueli Wang, Kelly Wickersham, Seo Young Lee, Na Lor, Ashley N. Gaskew & Amy Prevost - 2020

Background: Although a long line of research has been devoted to transfer pathways in general, there remains limited work on the capacity for community colleges to cultivate STEM baccalaureate transfer. In particular, both quantitative and qualitative evidence is extremely sparse on how STEM-aspiring students beginning at community colleges experience supports and barriers on their journey to pursue a STEM baccalaureate.

Purpose: This mixed-methods study addresses the question: What salient factors are associated with beginning community college STEM students’ decisions to transfer into baccalaureate STEM programs, and how do students describe the supports and barriers they experienced specifically pertaining to these factors?

Research Design: Guided by the STEM Transfer model, we carried out this research using an explanatory sequential mixed-methods design. We incorporated survey, administrative, and interview data from three large two-year institutions in a Midwestern state. We applied Artificial Neural Network (ANN) techniques to identify factors associated with beginning community college STEM students’ decisions to transfer into baccalaureate STEM programs. Based on the factors that emerged from ANN, we analyzed the interview data to give meaning to the identified factors using students’ rich descriptions of their experiences.

Findings: Results from the ANN revealed that students’ initial attitudes toward science was the most salient factor related to transfer in STEM. Following that, GPA, students’ initial attitudes toward math, transfer capital, being employed full time, major declaration, science preparation in high school, income levels above middle level, and transfer efficacy also turned out to be important variables shaping students’ transfer in STEM. Qualitative results further illustrated how the factors from the ANN exerted their impact.

Conclusions: This mixed-methods research illuminated significant factors shaping the road to becoming a scientist, as well as how those factors manifested their influences within the contexts of students’ educational journeys. Through this approach, we were able to establish the significance of influential factors without presuming directionality and leverage the interview data to disentangle how these factors functioned independently and together in sophisticated and nuanced ways. Our study brings forth a deeper understanding of community college students’ STEM pathways, including the many plot twists and processes involved to overcome challenges and maintain progress.



INTRODUCTION


The demand for preparing the next generation of the nation’s innovators in science, technology, engineering, and mathematics (STEM) fields is a long-standing topic of federal, state, and higher education interest. Numerous policy initiatives address the need to maximize national defense and economic productivity through the expansion of STEM education (National Economic Council and Office of Science and Technology Policy, 2015; National Science Board, 2010, 2015). In this effort, postsecondary institutions, particularly community colleges, are increasingly recognized as a major player in cultivating pathways to STEM careers requiring a bachelor’s degree and beyond (Hagedorn & Purnamasari, 2012; Hoffman, Starobin, Laanan, & Rivera, 2010; Wang, 2015).


Nearly half of all undergraduate students attend community colleges (American Association of Community Colleges, 2016). Also of note, almost half of all science and engineering graduates enroll at a community college at some point in their educational career (National Science Board, 2016). A vast share of community college students are from historically underserved populations based on race/ethnicity, socioeconomic status, age, as well as marital and employment status (Cohen, Brawer, & Kisker, 2014; Mooney & Foley, 2011; Tsapogas, 2004), reinforcing the critical role of these institutions in advancing diversity and equity in STEM education and the broader higher education system. Moreover, community colleges are fundamental in supplying a diverse talent pool to four-year institutions via upward transfer, potentially culminating in a broader group of students receiving baccalaureate degrees in STEM fields. The community college path to a STEM baccalaureate is especially important in light of a high demand for jobs that require science and engineering training at the bachelor’s degree level or beyond (National Science Board, 2015). As indicated by findings from the National Science Foundation’s National Survey of Recent College Graduates 2008, 75% of community college students in STEM-related fields reported that their primary reason for attending community college was to earn credits toward a bachelor’s degree (Mooney & Foley, 2011).


While community colleges’ potential in the cultivation of STEM pathways is undeniable, questions remain as to how to realize this promise. Historically, higher education scholars and policymakers have wrestled with the “democratizing” role of community colleges in serving historically underprivileged groups who would not otherwise have attended college (Doyle, 2009; Hagedorn & Purnamasari, 2012; Leigh & Gill, 2003, 2004; Rouse, 1995) versus these institutions’ potential “diversion” effect (Clark, 1960, 1980; Karabel, 1972) in “cooling out” the baccalaureate aspirations of transfer-intending students. It is especially intriguing to explore how the democratization/diversion discourse plays out in the STEM context, given these fields’ high demands and perceived academic rigor. Although a long line of research has been devoted to transfer pathways in general, there remains limited work on the capacity for community colleges to cultivate STEM baccalaureate transfer students (e.g., Hagedorn & Purnamasari, 2012; Jackson, Starobin, & Laanan, 2013; Wang, 2015). In particular, both quantitative and qualitative evidence is extremely sparse on how STEM-aspiring students beginning at community colleges experience supports and barriers on their journey to pursue a STEM baccalaureate. As Bahr (2013) argued, unpacking the pathway that leads to attainment yields great value for developing intentional interventions that promote community college student success, which is undoubtedly the case for cultivating the STEM pathway through community colleges. Accordingly, this mixed methods study investigates factors associated with beginning community college STEM students’ decisions to transfer in STEM fields, and how these factors may pose either supports or barriers that undergird students’ decisions to stay or leave the STEM transfer pathway.


BACKGROUND LITERATURE AND CONCEPTUAL FRAMEWORK


Community colleges have clearly evolved into a potential mechanism in broadening participation in baccalaureate STEM education through their upward transfer function (Dowd, 2012; Wang, 2013a). However, this policy priority has not been matched with a robust line of empirical work to foster systematic change. Nevertheless, a small body of research in this vein provides valuable insight into some factors that may contribute to STEM transfer. In this section, we review this literature, identify gaps that our study intends to fill, and describe a conceptual framework that informs our study.


BACKGROUND LITERATURE INFORMING THE STUDY OF TRANSFER IN STEM


Employing transcript data, several studies reveal the importance of succeeding in introductory math and science courses (Hagedorn, Cypers, & Lester, 2008; Hagedorn & DuBray, 2010; Wang, 2013b). In further exploration of STEM pathways starting at community colleges, Wang (2015) found that, although the overall STEM baccalaureate attainment rates among students beginning at community colleges were lower than those starting at public four-year institutions, community colleges are better positioned to maximize STEM momentum through robust course-taking that leads to future STEM baccalaureate success.


Prior research has also highlighted STEM transfer experiences and outcomes of women and students of color (e.g., Jackson, 2010; Jackson & Laanan, 2011; Jackson, Starobin, & Laanan, 2013; Myers, 2013; Packard, Gagnon, LaBelle, Jeffers, & Lynn, 2011; Packard, Gagnon, & Senas, 2012; Reyes, 2011; Starobin & Laanan, 2008). These studies were often qualitative in nature and illuminated several major dynamics underlying STEM transfer, such as access to transfer advising, articulation between two-year and four-year STEM programs, and a nurturing and supportive institutional culture.


As a whole, the literature on factors underlying community college students’ paths to transfer in STEM is small but growing. To summarize, across different studies reviewed, a number of factors have been identified as salient backgrounds, experiences, and attributes that may shape the STEM transfer path, including high school experiences (Adelman, 2006); students’ socio-demographic backgrounds such as race, gender, and socioeconomic status (Adelman, 2005; National Science Board, 2015; Wang, 2013b); self-efficacy beliefs (Jackson et al., 2013; National Science Board, 2010; Wang, 2013b); and institutional supports (Espinosa, 2011; Reyes, 2011). However, these elements are often examined in isolation from one another. To enhance this body of research, a more holistic approach needs to be adopted in order to produce a comprehensive understanding of these factors collectively. The fact that a significant portion of community college students start off with aspirations in STEM, but are unable to sustain those aspirations (Wang, 2013b) is of great concern and speaks to the complexities of this issue that need to be addressed from multiple lenses and methods.  


Specific to methodological choices, prior work in this area has drawn upon varying qualitative and quantitative approaches. Notwithstanding their unique strengths and contributions, most of the studies are cross-sectional in nature, only capturing a snapshot of student experiences at a given time. In light of the nuanced and evolving nature of how community college students form their aspirations, decisions, and eventual transfer paths in STEM, a longitudinal approach is much warranted to uncover the complexities during the process of transfer that are difficult for short-term studies to reveal. Furthermore, to date, this line of research has relied solely on either quantitative or qualitative methods. Corresponding to the topic’s richness, a mixed methods approach presents an opportunity to analyze the complex nature of students’ experiences and decision-making along the path to STEM transfer.


CONCEPTUAL FRAMEWORK


The STEM Transfer model (Wang, 2016) provides the conceptual grounding for this study. This framework draws upon the social cognitive career theory (SCCT) in the context of STEM transfer, and is informed by the small body of extant literature on the topic reviewed earlier. SCCT (Lent, Brown, & Hackett, 1994, 2000) delineates individuals’ career development and choices through the influence of cognitive-personal factors, situated within personal backgrounds and environmental contexts. In addition to SCCT, prior scholarship on upward transfer has identified individual, motivational, and environmental factors that offer further context to the STEM Transfer model. According to the model (Figure 1), eventual transfer in STEM is affected by an intersecting set of contextual factors, learning experiences, and motivational beliefs related to STEM learning and transfer, as well as person inputs and postsecondary factors.


Figure 1. Conceptual framework of the study (STEM Transfer Model).

[39_23000.htm_g/00002.jpg]

Note. For a more detailed conceptual description of the STEM Transfer model, see Wang (2016).

Guided by the conceptual framework, we adopt a mixed methods approach to address the following research questions: What salient factors are associated with beginning community college STEM students’ decisions to transfer into baccalaureate STEM programs, and how do students describe the supports and barriers they experienced specifically pertaining to these factors? Through the quantitative strand, we bring together a comprehensive set of factors to reflect the holistic nature of the topic and identify the most prominent factors associated with students’ decisions to transfer in STEM. In particular, by adopting Artificial Neural Network (ANN) analysis, we account for the iterative and nonlinear ways in which these factors may intersect with one another. With the qualitative strand, we give meaning to the identified factors using students’ rich descriptions of their experiences from in-person interviews, allowing us to contextualize and interpret our quantitative findings in more intimate and impactful ways. Overall, our study sets out to fill in the major gaps in the literature on the community college STEM transfer pathway, thus broadening the knowledge base on how to expand STEM baccalaureate participation through community colleges.


RESEARCH DESIGN


We carried out this research using an explanatory sequential mixed methods design that consists of two distinct strands: quantitative followed by qualitative (Creswell, 2013; Plano Clark & Ivankova, 2016; Teddlie & Tashakkori, 2009). This mixed-method design is well positioned to first identify prevalent patterns and factors across participants’ experiences through quantitative analysis, followed by qualitative data collection and analysis that help explore and explain initial quantitative results (Plano Clark & Ivankova, 2016). In our study, the mixing of the two strands occurred at two points, first in qualitative sampling when we used the survey data to identify participants for qualitative interviews, and second and more notably, during interpretation when we made sense of quantitative and qualitative results together.


STUDY SAMPLE AND DATA


We utilized data from a longitudinal, mixed methods project that investigates factors influencing students’ upward transfer in STEM. The target population included students beginning at three large two-year institutions in a Midwestern state in Fall 2014 with either a major declaration in STEM or course enrollment within STEM subjects. Through the project’s baseline survey, approximately 3,000 students were invited to participate in the longitudinal study, and 1,668 students took the Fall 2014 survey (with a response rate of 56.6%). The survey sample was futher restricted to respondents who indicated whether they decided to transfer in STEM, and after removing missing cases, the final analytic survey sample is 1,197. During Summer 2015, based on analysis of the survey data, individual interviews were conducted with 43 survey participants to deepen the findings from the survey data.


DATA COLLECTION

 

Quantitative Strand


Data for the quantitative strand were collected in Fall 2014 using the Expanding STEM Talent Survey, an instrument developed to capture beginning community college students’ intent and decisions regarding upward transfer in STEM fields, and to measure factors that may influence students’ transfer in STEM described earlier. Aligned with the STEM Transfer model, the key domains included in the survey are: person inputs (such as demographic background, prior academic preparation, and initial attitudes toward math and science), learning experiences in STEM classes and programs, motivational attributes (such as self-efficacy in math and science, outcome expectations, interest, and intent with regard to transfer and STEM education), contextual factors including background influences and supports as well as transfer-oriented interactions with faculty, peers, and advisors.


The survey instrument was validated through the following means. First, a pilot study was conducted in Summer 2014 to establish the instrument’s initial validity and reliability that involved feedback from an expert review panel, cognitive interviews with students, and exploratory factor analysis. Following the baseline survey data collection, confirmatory factor analysis and item response theory were adopted, with findings largely confirming the survey’s technical quality. A full description of the survey instrument and detailed procedures to assess its validity and reliability is provided in Wang (2016). Following the tailored design method (Dillman, Smyth, & Christian, 2009), we collected the survey data using a multi-mode approach by contacting respondents by both email and mail. Survey fielding procedures are detailed in Wang and Lee (2019). Table 1 describes the variables used in the study along with the survey sample’s descriptive statistics.


Table 1. List of Variables and Survey Sample Descriptives

Variable

Description

Sample Characteristics

N

(%)

Dependent Variable

   

Decision to Transfer in STEM

Binary indicator of whether students decide to transfer into STEM programs at baccalaureate institutions.

1=Yes



368



(30.74)

0=Otherwise

829

(69.26)

Independent Variables

  

Sex

1=Female

523

(43.69)

0=Male

674

(56.31)

Race/Ethnicity

A series of dummy variables

  

Asian and Other race/ethnicity

197

(16.46)

Black

79

(6.60)

Hispanic

113

(9.44)

White

808

(67.50)

Age

1=24 or older

376

(31.41)

0=Under the age of 24

821

(68.59)

English as a second language

1=English as a second language

201

(16.79)

0=Otherwise

996

(83.21)

First generation

1=Respondent is the first to go to college in their family

368

(30.74)

0=Otherwise

829

(69.26)

Financial dependency

1=Being financially dependent on parents or legal guardians

654

(54.64)

0=Otherwise

543

(45.36)

Family income

A series of dummy variables

  

<$30,000

421

(35.17)

$30,000-$60,000

293

(24.48)

>$60,000

483

(40.35)

Employment status

A series of dummy variables

  

Full-time employment

260

(21.72)

Part-time employment

639

(53.38)

No employment

298

(24.90)

Marriage status

1=Married

148

(12.36)

0=Otherwise

1049

(87.64)

Single parent

1=Single parent

51

(4.26)

0=Not single parent

1146

(95.74)

Enrollment intensity

1=Attending college full time

357

(29.82)

0=Otherwise

840

(70.18)

Institution type

1=Transfer as primary mission

421

(35.17)

0=Comprehensive in mission

776

(64.83)

Major declaration

1=Respondent declared a program of study in STEM

760

(63.49)

0=Respondent enrolled in STEM courses without declaring any majors

437

(36.51)

  

M

(SD)

GPA

GPA in fall 2014

3.02

(1.01)

Financial Support

Financial support respondent received to attend current institution. Item based on 5-point scale: 5 indicating “A great deal” and 1 indicating “None”.

2.52

(1.05)

Emotional Support

Mean scale score of emotional support respondent received from their family and friends consisting of two survey items measured on a 5-point scale: 5 indicating “Extremely” and 1 indicating “Not at all”.

3.64

(0.97)

Math preparation

The number of years of high school coursework in math. 1=Did not take, 2=0.5 year, 3= 0.5-2 years, 4=2-3 years, 5=3-4 years, 6=4 years or more

5.34

(0.98)

Science preparation

The number of years of high school coursework in science. 1=Did not take, 2=0.5 year, 3= 0.5-2 years, 4=2-3 years, 5=3-4 years, 6=4 years or more

5.18

(1.04)

Attitudes toward math

Mean scale score of respondent's self-reported attitudes toward math consisting of four survey items measured on a 5-point scale: 5 indicating "Extremely" or "A great deal" and 1 indicating "Not at all"

3.58

(0.88)

Attitudes toward science

Mean scale score of respondent's self-reported attitudes toward science consisting of four survey items measured on a 5-point scale: 5 indicating "Extremely" or "A great deal" and 1 indicating "Not at all"

3.60

(0.98)

Self-efficacy in math

Mean scale score of respondent's self-reported math self-efficacy consisting of five survey items measured on a 5-point scale: 5 indicating "Extremely" and 1 indicating "Not at all"

3.85

(0.85)

Self-efficacy in science

Mean scale score of respondent's self-reported science self-efficacy consisting of five survey items measured on a 5-point scale: 5 indicating "Extremely" and 1 indicating "Not at all"

3.74

(0.86)

Learning experience

Mean scale score of respondent's self-reported active learning experience consisting of fifteen survey items measured on a 5-point scale: 5 indicating "Very often" and 1 indicating "Never"

3.36

(0.67)

Transfer capital

Mean scale score of respondent's self-reported transfer capital consisting of seven survey items measured on a 5-point scale: 5 indicating "Very often" and 1 indicating "Never"

2.29

(0.91)

Transfer efficacy

Respondent’s self-reported confidence about their ability to handle the process and requirements for transferring to a 4-year college or university. Item based on a 5-point scale: 5 indicating “Extremely” and 1 indicating “Not at all”

3.66

(1.06)


Qualitative Strand


In this explanatory sequential mixed methods design, it was important that participants in the qualitative strand had also responded to the survey. Guided by a purposeful sampling approach, we primarily relied on maximum variation when sampling survey participants due to this technique’s capacity to identify essential elements and experiences across the variation among diverse participants (Patton, 2002). Capitalizing on the large number of survey respondents, we integrated random sampling techniques through stratification of the survey respondents by research sites and STEM fields, followed by a random selection of 100 survey participants from the joint distributions. We then augmented this information by adding respondents’ self-reported demographic background characteristics such as sex, race/ethnicity, and age. Finally, from this list, we purposively identified students who helped us achieve maximum variation along these noted dimensions.


A total of 43 students consented to and participated in the interviews, which were semi-structured in nature and lasted between 40 and 90 minutes (see Table 2 for a description of the qualitative interview sample.) Semi-structured interviews allowed us to draw upon preconceived protocols to guide the interview process (Mills & Birks, 2014), yet provided us the opportunity to follow additional responses or topics that emerged throughout the conversation. In general, our interview questions explored students’ experiences in STEM programs or courses, their educational and career plans, and other areas closely aligned with key domains in the survey instrument described earlier. The overarching protocol used for the interviews can be found in Appendix A. Prior to conducting the interviews, we as a research team underwent multiple rounds of training in which we iteratively reviewed and practiced interview protocols and calibrated among ourselves until we were able to ensure that we consistently addressed all facets of the interview process.


Table 2. Background Characteristics of Interview Participants

Study Name*

Race/Ethnicity

Sex

Age

Program of Study

Institution Type

Alexander

Hispanic/Latino

Male

26

COMP/MAT

Transfer

Bethany

White

Female

23

BIO/AG/ENV

Comprehensive

Bill

White

Male

19

PHYS

Comprehensive

Bubbles

Hispanic/Latino

Male

19

ENG

Transfer

Callan

Black

Male

48

BIO/AG/ENV

Comprehensive

Clyde

Black

Male

23

ENG

Comprehensive

Darian

White

Male

25

BIO/AG/ENV

Comprehensive

Elizabeth

Asian

Female

19

PHYS

Transfer

Gertrude

Native American

Male

47

COMP/MAT

Transfer

Greer

White

Female

18

ENG

Transfer

Gwyneth

White

Female

29

PHYS

Comprehensive

Hunter

Unknown

Male

18

ENG

Comprehensive

James

Black

Male

54

COMP/MAT

Comprehensive

Jasmine

White

Female

30

BIO/AG/ENV

Comprehensive

Jennifer

White

Female

25

BIO/AG/ENV

Comprehensive

Jennipher

White

Female

18

COMP/MAT

Comprehensive

Jim

White

Male

36

PHYS

Comprehensive

J.J.

White

Male

26

ENG

Transfer

John

White

Male

19

BIO/AG/ENV

Transfer

Jordan

White

Male

19

PHYS

Transfer

Joshua

White

Male

35

PHYS

Comprehensive

Kanda

Native American

Female

19

COMP/MAT

Comprehensive

Katy

White

Female

33

PHYS

Comprehensive

Kelly

White

Female

31

BIO/AG/ENV

Comprehensive

Kevin

Native American

Male

19

ENG

Transfer

Kirsten

White

Female

20

ENG

Transfer

Kooks

Hispanic/Latino

Male

19

ENG

Transfer

Kwesi

Black

Male

25

COMP/MAT

Comprehensive

Lotty

White

Female

54

BIO/AG/ENV

Comprehensive

Mathais

Native American

Male

24

ENG

Comprehensive

Nelkowicz

White

Male

35

ENG

Comprehensive

Nico

White

Male

22

PHYS

Comprehensive

Norman

White

Male

59

COMP/MAT

Comprehensive

Paul

Unknown

Male

29

COMP/MAT

Comprehensive

Robert

Hispanic/Latino

Male

20

COMP/MAT

Comprehensive

Rain

Hispanic/Latino

Male

30

COMP/MAT

Comprehensive

Scott

White

Male

19

ENG

Transfer

Seamus

Multiracial/Black

Female

23

BIO/AG/ENV

Comprehensive

Shinichi

Asian

Male

19

BIO/AG/ENV

Comprehensive

Stella

White

Female

33

BIO/AG/ENV

Comprehensive

Temperance

White

Female

33

ENG

Comprehensive

Tom

Hispanic/Latino

Male

28

COMP/MAT

Comprehensive

Vern

White

Male

27

COMP/MAT

Comprehensive

Note. ENG=engineering/engineering technology, BIO/AG/ENV=biological/agricultural/

environmental life sciences, PHYS=physical sciences, COMP/MAT=computer/mathematical sciences.

*Students chose their study names.


DATA ANALYSIS


Quantitative Strand


We applied Artificial Neural Network (ANN) techniques to quantitatively analyze the survey data. ANN departs from conventional approaches, such as a logistic regression, that are commonly adopted in higher education research exploring categorical outcome variables (Luan, 2002; Paliwal & Kumar, 2009). These traditional methods typically require a presumed relationship between independent and dependent variables. For example, a regression model without any quadratic or cubic terms assumes that there is a linear relationship between independent and dependent variables (Cohen, Cohen, West, & Aiken, 2003). While a regression analysis has advantages, such as ease of interpretation of model parameters and the ability to validate presumed relationships through model testing, this type of analysis is best applied when relationships can be modeled in advance, which is an assumption that can be untenable for underexplored topics.


Given these limitations, we opted to use ANN, an increasingly appealing methodological approach well suited for discovering relationships between independent and dependent variables without establishing any form of relationships in advance (Garson, 1998; González & DesJardins, 2002; Kantardzic, 2011; Paliwal & Kumar, 2009). Exploratory in nature, ANN excels at uncovering underlying trends or patterns in an existing dataset so as to reflect the complex dynamics of the data (Luan, 2002). Similar to how a biological neuron in the human brain registers, processes, and adopts external inputs in order to return an output from the processed inputs, ANN derives an output by using a set of independent variables through an iterative learning process to extract underlying empirical relationships between independent and dependent variables.1 As such, ANN is an optimal fit for the quantitative component of our study given our mixed methods approach, because the underlying mechanism will be deeply analyzed through the qualitative strand of our study. For the independent variables, we used a rich set of baseline survey items and administrative records in Fall 2014 (refer to Table 1 for survey item description). The dependent variable in the output layer is a binary indicator of students’ decisions to transfer in STEM.


A multilayered feedforward neural network was constructed with three layers: input, hidden, and output layers. We applied backpropagation algorithm as a learning process. After several exploratory investigations, a final model was decided with 32 independent variables in the input layer, 1 node in the hidden layer, and 1 dependent variable in the output layer. Furthermore, we applied 10-fold cross-validation by separating the data into 10 mutually exclusive subsets with almost equal sample sizes and repeating ANN 10 times, with nine subsets for building the relationships, and the remaining one for validating the identified relationships. The model was finalized by aggregating the results from the 10 repeated analyses (Kuhn, 2008). All analyses were conducted using the caret package (Kuhn, 2008) in R (R Core Team, 2015).


Qualitative Strand


The interviews were recorded, transcribed verbatim, and initially coded using an open coding scheme in order to capture a multitude of concepts and patterns in the data. Based on the factors that emerged from ANN, we further analyzed the interview data for relevant quotes and codes. As an example, for the factor “attitudes toward science,” we isolated interview data that included codes containing a range of words to describe how the students felt either broadly about science or specific learning experiences within various science subjects. For instance, students might have mentioned that they “love biology,” that “science is interesting,” or they found a course to be “more interesting than expected.” We also searched beyond the codes and looked for quotes that referenced ways in which students talked about science or their perceptions of science and how scientists behave, such as how “science applies to life,” that you “can’t argue with science,” “science people [are] more introverted,” or “chemistry [is a] guy’s thing.”


The research team members engaged with this analysis first independently and then as a group to discuss the codes and quotes identified from the individual process. The multiple researchers and iterations of analyses served as an innate triangulation strategy (Creswell, 2013). After identifying all relevant candidate codes and quotes for each factor, we refined them together based on their depth and ability to illustrate the various ANN factors. Using the factor “GPA” as an example, if students only mentioned their grade or performance without actually describing how they achieved it or how it influenced their journey in STEM, we were inclined to eliminate those codes or quotes.


Through this process, we were able to resolve inconsistences arising from our individual analysis and arrive at a consensus regarding the codes and their associated quotes. We then went through our selected list of codes and quotes under each factor to develop preliminary themes. We again approached the theme development process first individually and then together to reach a consensus, triangulate, and ensure credibility and trustworthiness. Throughout the analytical process, we took note of meaningful codes, quotes, and themes in terms of their grouping and organization. A sample of the coding and theme development process can be found in Appendix B.


FINDINGS


RESULTS FROM ANN


Table 3 shows ANN’s performance of the prediction. Of the analytic sample, 78.61% of the participants were predicted correctly by the final ANN model, indicating reasonably good model performance. We then investigated the relative strength of the independent variables in predicting the dependent variable. Following Garson (1991), we quantified the contribution of independent variables using weights connecting each layer so as to determine the relative importance of each independent variable. In this method, absolute connection weights are used to compute the relative variable importance instead of drawing conclusions regarding the directionality of the relationship between the independent and dependent variables. This approach is appropriate as we assume that the nature of such relationships is rather complex and nonlinear, which will be deeply investigated by the qualitative strand of our study.


Table 3. Distribution of Categories of Dependent Variable, Observed and Predicted

  

Observed

 
  

Transferred in STEM

Did not transfer in STEM

Total

  

N

(%)

N

(%)

N

Predicted

Transfer in STEM

203

(16.96)

91

(7.60)

294

Did not transfer in STEM

165

(13.78)

738

(61.65)

903

 

Total

368

(30.74)

829

(69.26)

1,197

Note. Bold indicates correct prediction.

Table 4 shows the relative variable importance in descending order. Note that the sum of the relative importance is 100. The most important variable was students’ initial attitudes toward science, indicating that, out of the 32 independent variables we considered, this factor had the strongest influence on students’ decisions to transfer in STEM. Following that, GPA, students’ initial attitudes toward math, transfer capital, being employed full-time, major declaration, science preparation in high school, income levels above middle level, and transfer efficacy also turned out to be important variables shaping students’ transfer in STEM. See Table 4 for the full set of results.


Table 4. Relative Variable Importance


Independent variables

Relative Variable Importance

Attitudes toward science

12.249

Academic achievement (GPA)

6.346

Attitudes toward math

6.068

Transfer capital

6.052

Employment: Full-time

5.340

Major declaration

4.814

Science preparation

4.581

Family income: Upper level (>$60,000)

4.389

Family income: Middle level ($ 30,000~$60,000)

4.355

Transfer efficacy

3.740

Employment: Part-time

3.569

Income: Lower level (<$30,000)

3.551

Employment: Not employed

3.399

Self-efficacy in science

3.333

English as second language

2.873

Not traditional age

2.744

Enrollment: Part-time

2.597

Learning experience

2.223

Race: Black

2.175

Self-efficacy in math

2.162

Marital status: Married

2.108

Transfer focused institution

2.008

Race: Asian and others

1.912

Gender: Female

1.649

Race: Hispanic

1.455

Race: White

1.154

Single parents

1.013

Math preparation

0.881

First generation

0.627

Financial dependency

0.292

Financial support

0.185

Emotional support

0.155


RESULTS FROM QUALITATIVE INTERVIEWS TO DEEPEN FACTORS IDENTIFIED BY ANN


Qualitative analyses allowed us to further illustrate how the ANN results might be understood. To reiterate, ANN does not provide information about the specific ways in which the identified factors impact students’ STEM career trajectories. Therefore, following the ANN with qualitative data collected from students themselves was essential for understanding how each of these factors exerted its impact. In the qualitative analysis, we focused on the 10 most influential independent variables identified by the ANN results. Because our study is the first to integrate ANN with qualitative interviews and since the independent variables’ importance is evaluated relatively, there is no established cutoff value to determine the precise number of variables to emphasize in the subsequent qualitative strand. We decided to highlight variables with a relative importance value greater than 3.125, which represents the mean value of the relative importance accounting for all 32 independent variables included in our analysis. Next, we delved into the interview data to identify qualitative evidence that contextualized these 10 factors.


Initial Attitudes Toward Science


The most influential factor, initial attitudes toward science, underscored how students felt about their experiences with science content early on. Three themes emerged from the data related to this factor: curiosity to learn, connections to students’ lives, and surviving and thriving in STEM. The following sections describe these three themes in detail along with relevant findings to illustrate them.


Curiosity to learn. The first theme stemming from initial attitudes toward science reflected a curiosity to learn, whether that meant learning science more broadly or detail-oriented science subjects. One participant, Jim, discussed his curiosity toward discovering how objects function, “I’m like a compulsive tinkerer. I love taking things apart and trying to find out how things work.” Echoing this, another participant, Ben, expressed a similar passion for learning, “I like to learn stuff, and I like learning new things. I read a lot. I love to read, and so just—as long as I’m learning something or doing something interesting, I’m okay with it.” Having an innate curiosity and being actively engaged in the learning process appears to help students enter STEM courses or programs with the mindset they need to succeed in their coursework. A general curiosity to learn can give rise to students’ interest and abilities as scientific thinkers and subsequently act as a catalyst toward a progression in the pursuit of a STEM degree.


A general curiosity toward learning extended to specific science subjects. One participant, Stella, illustrated her curiosity for learning various aspects of biotechnology:


The biotechnology program specifically, I thought it would be like testing blood samples and bacteria and that sort of thing but it’s getting a lot more in depth in genetics than I thought it would and it’s really interesting and I think it’s really awesome. I wasn’t really expecting that.


It was the unexpected content that further intrigued Stella, who came into the program anticipating that she would work with particular samples, such as blood and bacteria, yet ended up learning details about genetics. Ben echoed this sentiment, “Biotechnology? It’s like an emerging field. Well, it already has emerged, but there’s a lot of new stuff to learn in it and it seems really exciting.” Ben added, “But I like all the classes I’ve been in so far, working in the lab and doing math. Making solutions, just everything. Learning about science stuff . . . ” These students’ paths toward becoming scientists were essentially fueled by their continued curiosity and discovery in general and with respect to the specific science subjects included in their programs. Being actively involved in the inquiry process and perceiving science as a “new frontier” ripe for discovery shaped their attitudes in a positive manner.


Connections to students’ lives. The second theme we identified spoke to how the provable, applied nature of science connected directly to students’ lives. Kelly mentioned that her passion for science took shape as she came to realize the nature of science:


I used to think my passion was history, but you can only hear it so many times, you know? And oh, someone’s got a different take on history, okay fine. I like also that it’s more or less provable, you know like history is a lot of conjecture and there’s no arguing with solid science.


The assumption that scientific principles can be verified gives STEM fields a perceived objectivity not present in many other fields. This “provability,” at least on the surface, helps learners use their scientific knowledge to model approaches that can be applied to solving problems in general. As Kelly said, “Knowing more about science, it makes you a better problem solver, you know for every day encounters even.” Bethany also captured this point as she discussed the development of insulin for individuals with diabetes:


we talked about the fact that it was the biotech industry that came up with insulin replacement because like people were using it from like pigs and like crazy stuff before and so being able to grow it in bacteria for, you know, diabetic patients like that was kind of like the very first thing I learned about that application and sparked my interest.


Thus, students’ attitudes toward science are shaped by the ability to extend scientific knowledge and/or problem-solving to everyday objects or challenges, allowing them to develop a deeper understanding of and appreciation for science.


Surviving and thriving in STEM. The third theme represents students’ attitudes toward science in terms of surviving and thriving despite the biases that persist in STEM disciplines. One of the more general perceptions is that students in STEM may prefer to work alone and may be more reserved. Gwyneth spoke of this, “I think sometimes people that just go into science don’t always get, they’re a little bit more introverted.” These types of assumptions, albeit potentially misleading, may serve as a support or barrier in students’ pathways toward becoming scientists depending on whether they feel that their personality aligns with such perceptions, and whether or not they feel that they fit with others in the field.


These assumptions also extend to gender biases that students may encounter. Attitudes about whether women can “make it” in STEM can result in either a deflating or conversely motivating impact. Gwyneth experienced gender bias in her coursework, which influenced her decision to choose a major field of study in STEM, “I was thinking about going to school for chemistry, people telling me that they didn’t think I could handle it, that it was too difficult, that it’s more of a guy’s thing.” Kirsten also noticed a gender disparity and potential bias that exists in STEM:


I think it’s encouraging to see other females and you know, when there’s not that many it’s just kind of a weird feeling. It’s a little bit isolating in a sense. . . . I think that the STEM is very underrated. STEM gets such a bad rep for being, I don’t know. For having obviously not a lot of women.


Regardless, these students persisted. For Gwyneth, she was determined to stay on her path despite encountering the attitude that chemistry was not a field that women were able to enter and stated, “I’m still in science and I’d like to continue doing more science after this program.” Although the lack of women in STEM courses could feel isolating, Kirsten felt “comfortable and relaxed.” As a result, perceptions regarding gender in science have the potential to militate against an individual’s pathway to becoming a scientist, but in some cases, students may use them as a motivating factor to persist in the face of such challenges. In these instances, it can be critical to leverage this motivation to help students stay on their path in STEM.


GPA


Grade point average (GPA) emerged as an important factor related to students’ pursuit of a STEM transfer path. Two themes emerged from the interview data with respect to GPA: the perception that students were defined by their grades and instances in which students felt validated to persist due to earned grades.


Defined by grades. Students discussed their GPA in terms of a perception that they were defined by their grades, which would in turn determine their progress toward transfer in STEM. One of the participants, Mathias, expressed his concern with grades possibly influencing his path to a four-year institution, “But if my grades don’t reflect what they’re looking for, chances are, I might not be able to get into that four-year college.” He expanded on this, “Just by looking at a person’s grades, they may or may not accept you, and then it’s almost like they can get stuck.” As such, depending on performance indicators in the form of grades and GPA, students may feel restricted from becoming scientists. In a sense, the perception of the grades needed to transfer in STEM may serve as a barrier as much as the grades themselves. Students may choose to drop out or switch programs rather than persist in a STEM transfer pathway if they believe that they do not have the grades to make it in these majors.


Validation to persist due to earned grades. On the other hand, GPA or grades can serve as a validation mechanism. Specifically, students mentioned that the grades they received served to affirm their pursuit of a STEM field. Tom illustrated this when reflecting on a grade received in one of his STEM courses:


But then when they start to give me my grade it was better than expected. It was a higher grade than I expected. I was like, “Oh, wow.” He [the instructor] said, “I noticed that you did improve in your time in the class. You were asking questions. You were participating in my class.” That’s why he gave me a better grade. That was nice.


In Tom’s case, the instructor explained that the higher grade was influenced by active participation in class and not solely test scores or assignments. As a result, Tom’s efforts, not only in terms of what he may have achieved through traditional measures, but also with respect to his actions and behaviors in class, reinforced his decision to stay on the STEM path, as he felt validated and more confident in continuing his journey. Expanding or redefining the types of assessment measures used in STEM courses can make a significant difference in students’ confidence and decisions to persist in STEM.


Other students expressed a similar sentiment with respect to their performance in STEM disciplines. For example, J.J. was unsure of what to expect when he first started college. It was through his first and second semester classes that he gained confidence in and prepared for the increasing challenge of his STEM courses:


my first semester, I didn’t know what to expect. I didn’t take obviously the hardest classes, I was just taking generals, but I was able to maintain a 4.0 throughout the semester and that was really reassuring that got me ready for second semester where I took some harder classes and I actually got a 3.8 that semester. So, I was really pleased with my, you know, with my experience.


As such, obtaining good grades may help students feel as though they are following the appropriate academic and career pathway in STEM, that they have academic aptitude in STEM subject areas. Overall, grades may represent a determining or validating factor as to whether becoming a scientist is an obtainable goal.


Initial Attitudes Toward Math


Similar to initial attitudes toward science, initial attitudes toward math came to the forefront of the ANN findings. Three themes based on the interview data emerged from this factor: the perception that math takes work, developing an appreciation for math, and the applied or definitive nature of math.


The perception that math takes work. Many students mentioned that they had to work hard to do well in math, but with a positive attitude, they were able to find ways to make the work enjoyable, and ultimately came to appreciate the usefulness of math. For instance, although Gwyneth did not feel that she particularly excelled at math, she was aware that she would need it if she were to pursue a transfer pathway in STEM. As such, she remained positive and willing to put in the effort to understand and succeed in math.


In Bill’s case, after taking an exam with little preparation, he realized he would need to work harder if he wanted to do well in his math class. By taking the time and intentionally studying, Bill was able to transform his performance in his math class:


I never spent any time out of school studying or anything. And I’d never taken Calculus before. The last taken was Pre-Calc and that was two years ago. . . . So I hadn’t been that ready for math. And after, I got a 50 on the first quiz. And after that, I decided I need to put more time into it . . . and it turned my grade around a lot now.


With hard work, time, and deliberate practice, students are able transform initial attitudes toward math and stay on the STEM path.


Developing an appreciation for math. Regardless of some students’ initial attitudes toward math, they found ways to appreciate or even enjoy math. John shared that, despite feeling average in terms of his math ability, he was not opposed to the subject:


I didn’t really mind math too much. I was always like, I mean it never really was like wow I’m actually good at this or no I’m actually bad at it. I’d say I’m about average at math, like it’s one of those things that I don’t mind doing.


John did not see himself as exceptional in math, but he knew that he needed to be able to have an understanding of it in order to stay in the STEM pipeline. In fact, for some students, initial attitudes in math were negative. For example, Kirsten explained how she used to feel about math, “I used to hate it, I couldn’t stand math. But the more I’ve matured I realize math is really incredible and it’s a great skill, and it’s really, it’s an amazing tool.” She elaborated on her initial attitude toward math:


I remember freshman year of high school thinking that that was just the coolest thing in the world that I didn’t have to take math my senior year. I was so excited, I could not wait. And obviously I took math my senior year and it will continue, and I remember thinking how cool it would be that I wouldn’t have to take math in college, like I knew there was no way. I was so excited to be done. And here I am. I just think that a better attitude, like a more expanded attitude has really helped me appreciate things that I wouldn’t think about.


Kirsten had no intention of taking math once she attended college, but a shift in her attitude toward math guided her toward a pathway in STEM. Although students may not have had the most positive initial attitudes toward math, they came to understand, accept, and eventually appreciate the utility of math, which helped transform those initial attitudes to help them stay on the path in STEM.


Applying the abstract. Other than general, initial attitudes toward math, students also noted that, when they could see how math was applied, this had a positive impact on their attitude toward it. For instance, Kooks described how he could see the application of math beyond the classroom: “the way that I see mathematics can be applied in real life and how mathematics helps me understand certain concepts with electricity and things that go on within electric components.” In a sense, he could see how math was relevant in his everyday life, but also importantly in electrical engineering. When math is contextualized within a field, it can be seen as a practical tool rather than merely numbers that are put into equations to solve for something abstract. In this case, the student was encouraged by his coursework in math to continue his career path.


The perceived definitive nature of math also shaped students’ attitudes toward math. For example, Scott described how math was verifiable, which made it appealing to him:


where math, you can check that stuff. You can look into it and see if you have the answer and it’ll be yes it works or no it doesn’t. So that’s kind of like there’s more a definite answer and that kind of stuff. So that’s what I kind of like about it. It’s easier to work with, I guess.


Scott found math to be easier to handle since he felt that there was no ambiguity in figuring out answers to math problems. He knew that his work could always be proven. Similar to attitudes toward science, the perceived applicability and certainty found in math within and beyond students’ programs of study helped them develop positive attitudes, in turn supporting students’ pathways in STEM.


Transfer Capital


Transfer capital refers to the accumulation of tools and knowledge that help students navigate the transfer process (Laanan, Starobin, & Eggleston, 2010). This capital can be acquired as networks formed through interactions with instructors, advisors, or peers, as well as resources and materials (e.g., transfer guides, websites, etc.) that students reference. In our study, students often spoke of various aspects of transfer capital contributing to their journey toward transfer and becoming a scientist. These mainly revolved around three themes: encouragement and support from advisors, the need to seek out resources, and relying on the knowledge and support of others (i.e., in general, faculty, friends, or family).


Encouragement and support from advisors. Jim discussed a conversation that he had with his advisor regarding next steps in his college program and the prospect of transfer:


“I would highly suggest you go to the four-year university,” and then he handed me the pamphlets and told me a little bit about each one. And then he says, “Well, you know, look into it and make sure you come back to me with enough time before you graduate so I can make sure that you get into something.” So, he was definitely on my side as far as corresponding with the right people.


His advisor was able to provide guidance and information about transfer, making Jim feel that someone was looking out for him. Jennifer had a similar experience with her advisor, who helped guide her toward courses to prepare her for transfer:


I knew that you have to have a 2.0 and like basically the standard stuff. You want to have like some of your general education done and that, those types of things. But yeah then I, it comes back to my advisor he talked to me. When he found out that I was planning on transferring he’s like, “Okay. These are some classes that you’re gonna want to take because it will, even if they don’t transfer it will prepare you for what you’re getting, you know what you’re going into.”


For these students, the support from advisors represented pieces of critical transfer capital they needed to understand and navigate the transfer process, an important step on the pathway to a bachelor’s degree toward a STEM career. Information gained in this way helps students prepare for academics at four-year institutions, complete the appropriate courses prior to transfer, and carry over the necessary credits toward a bachelor’s degree.


The need to seek out resources. Other students, with or without the help of advisors, sought out transfer resources on their own in order to understand and undergo the transfer process. For instance, Elizabeth used online resources, particularly a four-year institution’s website, to determine and keep track of the requirements she would need for the degree she wanted:


I’ll go to the Centerville website and stuff and see for pre-med—even when I’m there, these are the requirements or for general transfer, this is the stuff, and every so often check off, okay, I have this down, this down, this down.


Students also utilized other transfer resources. For example, Kanda attended events where she could inquire further about transfer: “I went to a college career expo. They had a bunch of colleges representing their programs they offered, and I asked a lot of them if they can transfer.” As such, students cultivated their own transfer capital to help inform their STEM pathways. These resources highlight the importance of making transfer information available through various venues to help students navigate the transfer process.


Relying on the knowledge and support of others. Besides advisors and transfer resources, students engaged in discussions with other people that they knew and trusted. For instance, John would talk to his stepbrother about transfer:


The one other person [who knew] about transferring in general though, [was] my stepbrother Mason. He went here for his first year, he’s a year older than me, and he did the same thing, just went for a year and he transferred to Porter University, so he kind of, I got to see his experience of going here and transferring out to a different school, so yeah.


For Kooks, he relied on his cousin to help him with the transfer process since his cousin had transferred to the same university that Kooks wished to attend:



I have a cousin that goes to Lake Spirit Community College, or Fieldcrest University right now and he transferred there, I’m not sure where from, but he, he helped me with making sure, you know, that I got all my credits accounted for besides the help I got from the counselors.


As illustrated, many of the students relied on individuals who already experienced the transfer process in order to gain a sense of what to expect and to ensure a smooth transition. These conversations may create a shared experience or provide students with a sort of “insider” knowledge that students may not obtain from an advisor. This type of capital represented a potential support as students worked toward their STEM degrees.


At the same time, there were students who discussed setbacks toward becoming a scientist due to a lack of transfer capital. For instance, Clyde admitted that he made the mistake of taking classes that were not transferrable to a four-year institution. As a result, if he decided to transfer, he would need to take additional courses, subsequently extending his educational timeline:


now I’m not taking any transferring classes, if I need to be transferred, I’ll just have to retake some classes again and the classes will take more time. That’s the mistake that I made from the beginning. I didn’t really know because I did not talk to my advisor about that.


Thus, a lack of transfer capital through missing crucial conversations surrounding course options and credit transfer may subsequently create a barrier to students’ paths in STEM.


Employment Full-Time


When it came to students who worked full-time, the major theme that emerged was the scheduling constraints when trying to reconcile school and work. For Jim, he felt “boxed in” when it came to the time of the day to attend college and the types of courses that he could take: “I work a normal full-time job, so it’s—I’m pretty much boxed into later afternoon or online classes.” He went on to say, “I mean, I just couldn’t be there from eight to five, just a normal work day kind of thing.” Due to Jim’s full-time work schedule, it was a challenge for him to pursue a pathway in STEM when he was unable to take desired courses during normal work hours. Other students expressed similar issues when discussing school and work. For example, James mentioned how he was unable to do school full-time and ultimately changed his work schedule:


Because I work every day, so it was kind of hard for me. I can’t really do full-time school. Maybe a class or two . . . I actually work third shift. At that time, I was working second shift so I would come in, go and then I have to go to work, get ready for work now. So, I switched to third shift.


James was only able to take a few classes at a time because of work and was constantly in a rush trying to go between school and work. He was eventually able to change his work hours. Flexibility in both work and school can be crucial to help students stay in their STEM programs. As for reconciling multiple obligations, Gertrude also noted the stress of fitting school and work into one day:


Because I’ve got to take all my classes in the morning so it’s, work gets to be when we get busy and we start working 10 hours a day then it gets kind of hard because you’ve got to. . . . I work until midnight, then I sit up until 2:00, and then I’d have to get up at 7:00 and come in here, then go back to work, sometimes it gets kind of stressful.


Overall, students found it challenging to manage school and work, not only in terms of the courses that they could take and their progress, but also with respect to the physical and mental demands of doing both. As a result, full-time employment posed a potential barrier to students’ paths in STEM.


Major Decision


The decision regarding what to study in college was one that had myriad influences. We found that students’ decisions around their major field of study revolved around three themes: the overall cost of education, job prospects post-graduation, and the alignment of the major with students’ values and interests.


Overall cost of education. Students’ choice of major was influenced by how much money they would need to invest to pursue various fields of study. For example, Kwesi explained that it would have cost him too much money to go to law school, so he chose to major in IT instead:


I got my bachelor’s. That was in 2011 and I was going to go to law school, but I realized that the amount of money that I was spending...didn’t add up so I chose—that was one of the reasons that I didn’t want to—that I chose the IT field.


Other students also expressed a similar concern with respect to the costs of pursuing different disciplines. For example, although Hunter was interested in anesthesiology, he realized that the costs of medical school were too great. As a result, he shifted toward other majors and degrees that would not be as financially burdensome. He explained how he changed his mind about pursuing various majors:


I guess first what I wanted to do was be an anesthesiologist, and doing a transferring program to Centerville University, but I figured that might be a little bit more expensive with a lot of student loans. Then I thought about going into a two-year degree, trying to get something smaller and being able to pay off the loans afterwards, and then I just kind of played around with thinking about doing a biology degree or becoming a mechanic or computer IT and kind of jumped around a lot but, I think I’ve kind of settled on this idea [automotive technician] because it’s something that I would like to do, it’s a two-year degree and you make a good amount of money afterwards so I was figuring that would be a good job to take.


Thus, students sensitive to the finances associated with college and specific majors may be drawn away from following those pathways and decide instead to pursue other paths that require less financial investment and result in adequate economic returns post-graduation.


Conversely, some STEM majors appeal to students not only because of the relatively high salaries enjoyed by professionals in these fields, but also because of financial aid opportunities. Although some students may be hesitant to financially invest in a STEM pathway, grant or scholarship options can either encourage or reaffirm students’ decisions to follow a path in STEM. Jasmine discussed how the opportunity of having grant funding not only ensured that she could pursue IT, but opened up several doors of various specialties within IT:


Oh, I first heard about it last summer. I was meeting with an advisor, an academic advisor, and I was trying to enroll, and I said that I was interested in IT, and they started telling me about different IT that are offered here and they said, “Well, this one’s covered by a grant and this one’s covered by a grant.” So, I could either pick from Network Security or Help Desk, and so that’s how I found out about it is somebody here told me about it when I was looking at getting into IT.


Job prospects post-graduation. Other students discussed the importance of job prospects post-college when choosing a major. For example, Rain spoke explicitly of the job market in his field as being influential toward his decision to major in network security:


It’s all about the job. I’ll be honest. We all go to school to get a better paying job, make sure that our jobs are secure. I don’t think I’ve ever met a university student or college student whose goal was to just go learn. Their main goal will always be obtaining a job or getting a job in a field where they won’t have to worry about getting fired or laid off or anything. I mean, that’s what makes my choice in regards to what field I really want to go into. Network security. There’s a high demand for that right now.


Job placement was also a priority for other students when choosing a field of study. Kelly echoed this by describing how the job success rate in biotechnology supported her decision to go into that field:


I was looking through the catalog with my advisor again, and she told me about the [job placement] success rate of the biotech program and she taught me about, I guess she just like opened up the idea of going through tech, getting a great job, and the success rate is high.


Students may be more inclined to select fields of study that offer numerous job opportunities, and given the growing need for STEM professionals, more students may be inclined to select majors within STEM fields. Job security and placement can represent important assurances for students considering a pathway in STEM. This finding also highlights the importance of making such information explicit and transparent since it can help students not only decide on a major, but also stick with it as they progress through their program. Choosing a discipline that will result in success beyond graduation can be the motivation needed to persist.


Alignment of the major with students’ values and interests. A number of students sought to align their major choice with their own values and interests. For instance, Stella initially considered a major in nursing. After some experience in the field, she decided to go another route:


I got into health care. I thought that I was going to be a nurse. I’m actually working as a caretaker right now. I have my CNA. I decided to switch to the biotechnology because I thought I could do more good. It seemed like my patients weren’t getting any better. It’s kind of depressing. So, I think I would rather work on the research and development side of things than just watching people die. You know? That sort of thing.


Ultimately, Stella did not feel that she was making as much of an impact as she would have liked. She decided a change of major might help her do that. As such, her personal interest in helping people or “doing good” influenced her choice to go into biotechnology, putting her on a pathway toward becoming a scientist. Many STEM fields tend to be at the forefront of innovation, which can be a perfect choice for students who value progress and making a difference on a larger scale. Paul’s passion in STEM also played a role in his major decision: “I just have passion for technology. Because if you look at it, all over the world, or the world as a large, you cannot go without IT. Everything revolving around us is IT.” In these instances, the students’ interests and values led to major choices in STEM, and in turn sustained their paths toward becoming scientists.


Science Preparation


Behind this factor is the prevalent theme that science preparation at the high school level plays a key role in entry into STEM. For Hunter, taking advanced science courses in high school shaped his subsequent course-taking and consideration of a path in STEM. He discussed a particular experience, “I love biology, I took AP Biology when I was in high school, my junior year and I loved it. I decided to take two science classes the next year.” As a result, Hunter’s early, positive, and rigorous science preparation served as a support toward a pathway in STEM.


On the other hand, having little or no science foundation in high school may present a potential barrier for students. For example, Elizabeth did not feel that she was prepared for college-level science courses due to her high school science experiences. She mentioned her prior chemistry preparation, “we didn’t really get a good foundation in chemistry in high school because my teacher, she had maternity leave and after that it just went downhill.” Overall, an insufficient or lack of science foundation prior to college may represent a barrier in students’ STEM pathway. College-level science has the potential to help students overcome their prior preparation so that they are able to continue in STEM.


Income Upper/Middle Level


In regard to the ANN results pertaining to family income, the key theme from the qualitative interviews was that financial constraints pose a major barrier to students’ journeys toward becoming scientists, regardless of the level of support from their families. While few students explicitly spoke about their income in relation to college, those that did mention it discussed their reliance on their parents’ income to attend college. One of the participants, Kanda, wished to transfer and continue her journey in STEM. At the same time, there were financial constraints that were preventing her from doing so:


when I started at West Shore College I pretty much had a debit card that had two and a half grand in it and that was supposed to cover all four years, it wouldn’t cover my first semester unless I took out loans and that’s pretty much what I did. And pretty much now I’m out of money, so I have to borrow from my parents, and they want to retire so it’s like they really want me to just finish my degree and get a job as soon as possible.


Although the flexibility in borrowing from her parents allowed her to stay on her path in STEM at the two-year college, she felt pressured to finish school and find a job to ensure that she could pay her parents back without compromising their retirement plans, as well as make additional money to return to college.


Another participant, John, had a college fund, which supported his pathway to the extent that he was fairly confident about affording the expenses of college and graduating without debt:


Also, because I can afford it, my parents do have a decent amount in my college fund. Probably not enough for everything but, if I can save enough money I’m hoping that I can still get out debt-free. But it’s just kinda like I can afford this, and I want at least one good year of the college experience and I think that it’s gonna be worth it.


Other students pointed out that, despite the income their family earned, they did not see very much of it. To illustrate, Kevin mentioned how his father pointed out their income level, but Kevin felt that a lot of it went elsewhere:


With my family, I don’t know why, Dad says we’re upper middle class, but I don’t see it because most of our money is going everywhere else and not to us. And he said it was like 30% of our yearly income. Like really? Geez.


Kevin was also working in order to pay for college and felt that he would need to continue working in order to support his pathway in STEM. As such, these students demonstrated how income and finances might act as supports or barriers in their pursuit of STEM disciplines. Fewer financial obstacles tend to allow students to progress forward, whereas added financial obstacles can pose as a detriment. Yet, even in cases where students were able to rely on money from family in order to defray college costs, it usually did not cover the entire amount needed.


Transfer Efficacy


Self-efficacy refers to the belief in one’s ability to accomplish a particular task (Bandura, 1997). Our qualitative interviews pointed to the prevalent theme that self-efficacy surrounding transfer generally improves with increased knowledge about the process. For example, an understanding of whether or not credits transfer can increase students’ transfer efficacy. With regard to transfer, a number of students were explicit about their ability to transfer. Specifically, they knew what they were doing and expressed that they viewed transfer to be an easy process. For example, one of the participants, Nico, felt well-informed about transfer and “didn’t look at anything.” He went on to say that he “pretty much knew what [I] was doing.” Other than simply knowing how to transfer, other students, including Kooks, mentioned the process being easy, “the transfer process was very easy, and you know, luckily Lake Spirit Community College does have a co-op with, or works with Fieldcrest University so everything transferred, and it was, yeah. It was a very easy process.” Essentially, having enough information about, and explicit, clear institutional partnerships for transfer provided students with the efficacy needed to support their paths in STEM through upward transfer.


On the other hand, a number of students were less confident with respect to transfer, noting that they were nervous about their grades impacting their ability to transfer, or that they needed to know more about the transfer process. For instance, J.J. explained the grades he would need for transfer, “I was a little nervous when I heard you needed like basically a 4.0 to get in, I was like that’s probably not going to happen once I start taking these high-level courses.” Hearing about requirements for transfer to certain universities, J.J. became worried that his performance in advanced courses, if potentially less than satisfactory, may inhibit his chances of transferring to the institution of his choice. Grade expectations related to transfer access can influence students’ beliefs in whether or not they can transfer to specific institutions, or transfer in general. This kind of pressure can essentially make or break students’ paths in STEM. This finding also brings up questions around the various requirements being communicated to students as they gauge their ability to transfer, as well as any ways to overcome such challenges.


Another participant, Bubbles, also expressed that he had some concerns regarding his ability to transfer, and in order to feel completely comfortable in undergoing the transfer process, he would “need to find out more about how my transfer—like how my credits will transfer and all that stuff. That’s what I’m most concerned about.” As such, depending on students’ level of transfer efficacy, it may work for or against their pathways in STEM. Being confident in carrying out transfer has the potential to help students realize transfer and continue on their road to becoming scientists, whereas concerns regarding requirements or retaining credits may make students feel less confident in their ability to transfer, thus creating a barrier to persist in STEM. These are important matters that should be addressed and resolved with students to help them confidently and successfully transfer.


Intersections Among Factors


It is important to point out that intersections occurred across several of the factors, revealing the complex and nuanced ways in which these factors play out in students’ journeys toward becoming scientists. For instance, when Gwyneth talked about different biases that may impact attitudes toward science, she brought up gender, framing chemistry in particular as a “guy’s thing.” Although gender itself did not emerge as a top factor in the quantitative analysis, the interview data revealed how it intersects with other, more influential factors, illustrating some of the interwoven supports and barriers that impact the pathways that students take in STEM.


On a similar note, full-time employment is inherently linked with other factors, such as part-time enrollment in school and income levels. Gertrude illustrated this when discussing the reconciliation between work and school:


I’m only doing part-time so it’s like two classes per semester so with work and that they’re like, if you do school part-time you work full-time, if you do school full-time you work part-time, and I can’t afford to go part-time.


For Gertrude, either work or school had to take priority, not both. As such, if a student has to work full-time, they may need to attend school part-time. Life choices that students face shape their persistence and progression in STEM. For students who do not have the finances to go to college full-time, their road to becoming a scientist may be impeded. For those who have the finances to attend full-time, they may be able to progress more easily through their education and get a job more quickly and easily. As a whole, the dual and intersecting roles that these factors play can help or create obstacles for students’ STEM paths.


Another intersection worth noting is among attitudes toward math and science and major decisions. For example, Jim discussed his attitude toward science, specifically noting his joy for learning and tinkering with objects:


I just had a spark one day and I said, you know, I have this—a thing called an Arduino. And I had to program it in a C+ language. And I’m like, I know nothing about this, I’d love to find out, you know. So, then I’m looking through it and stuff and I’m looking through all these different opportunities for school and stuff and I said, wow, maybe I could make a job out of this.


Jim’s curiosity in learning how things work led to his passion and interest in choosing programming as a major field of study. In Jim’s case, his positive attitude toward science motivated his major choice, which together supported his pathway in STEM. This example further demonstrates the interconnected nature of the factors, which can serve as supports or barriers for students as they pursue STEM. The factors, regardless of their significance, can act as building blocks to help students become scientists or collectively dismantle the road to STEM.


Finally, it was interesting to see how certain factors, such as GPA, could shape transfer efficacy, which in turn would support or hinder students’ paths in STEM. Students not only felt that grades may define their STEM pathways positively or negatively as noted in a previous section, but that grades could also impact other factors. This again illustrates the complex, connected nature in which the factors operate and that students navigate en route to becoming scientists.


DISCUSSION AND IMPLICATIONS


Our study set out to disentangle the potential supports and barriers community college students experienced as they navigated the road to becoming a scientist. Using ANN, we were able to reveal what factors carried significant weight in shaping students’ intent to transfer. In turn, students’ voices from our qualitative interviews shed light on how these factors operated within students’ STEM paths. It is clear that these factors were exerted in meaningful and intricate ways. Moving beyond simply naming the mechanism of how students’ interests, motivations, experiences, and life circumstances impact their persistence in a STEM pathway toward a career, this study allowed us to begin uncovering how and why these factors served as barriers and supports.


CULTIVATING POSITIVE ATTITUDES TOWARD SCIENCE AND MATH


Our findings underscore the significance of positive math and science attitudes to persist in STEM. Such attitudes or dispositions are linked with STEM achievement (Ethington & Wolfle, 1986) and are often found in STEM professionals already in the field (Christensen, Knezek, & Tyler-Wood, 2014). Therefore, cultivating initial positive attitudes toward math and science may help students realize their path to becoming a scientist, even in the face of challenges or obstacles. Coupling this with thinking like a STEM professional may help students foster various components of their identities, especially in the science or STEM domains. Developing a strong science identity can empower students and act as a vehicle toward success in STEM, particularly among women and students of color (Carlone & Johnson, 2007; Starobin & Laanan, 2005).


Since a deep understanding of science and math subjects is essential for career development in STEM, providing classroom experiences in which students explore meaningful and validating approaches to studying science and math is warranted. Students in our study mentioned that developing an understanding of the math and science itself and why these subjects are important shaped their attitudes in positive ways. Accordingly, the use of innovative teaching approaches or learning opportunities, such as active learning, contextualization, or research programs, among others, prove promising in helping students make immediate, concrete connections between the concepts they learn and what they may encounter down the road. Active learning can improve student performance in STEM versus more traditional lecture formats (Freeman et al., 2014). Contextualization provides explicit connections between academic subject matter and application, which are essential for knowledge transfer, academic achievement, and career readiness (Bransford & Schwartz, 1999), and play a role in students’ learning experiences and self-efficacy around their abilities and later success (Wang, Sun, & Wickersham, 2017). These approaches have great potential to transform initial negative attitudes into positive ones, thus dismantling a potential barrier to students’ persistence in STEM.


Strong high school science preparation may also contribute to cultivating and maintaining positive science and math attitudes along the STEM pathway. As our findings indicated, some students attributed their positive perceptions of science to their enjoyable learning experiences during high school. As a result, these students seriously deliberated on several STEM-related fields. On the other hand, a number of students recalled less positive experiences in high school, leaving them apprehensive about taking science courses or majoring in STEM in college. Although institutions and instructors cannot alter students’ prior high school preparation and experiences with regard to science, and math for that matter, they can consider integrating the promising instructional approaches discussed above to help students either maintain existing positive attitudes or possibly transform previously held negative perceptions of STEM subjects.


REVISITING ACADEMIC PERFORMANCE MEASURES TO ENHANCE STEM PATHWAYS


Regardless of being an imperfect measure of academic performance, GPA can shape students’ decisions whether to continue their STEM pathway. Our interview participants were acutely aware that they “might not be able to get into that four-year college” if their grades were not high enough, internalizing grades as a key deciding factor on their path to transfer in STEM. Indeed, grades, both in the beginning and cumulatively over the course of students’ studies, can determine their later attainment in STEM (Crisp, Nora, & Taggart, 2009; Whalen & Shelley, 2010). Although doing well in STEM courses is crucial toward progressing and successfully completing a STEM baccalaureate, particularly among community college students, the odds of doing so are not favorable for these students (Wang, 2015). As a result, there can be immense pressure to perform well from the onset in order to successfully realize a pathway in STEM. This high stakes mindset can be detrimental and discourage students from transferring in STEM.


Notwithstanding the value and importance of having a solid academic foundation, our findings suggest the inclusion of assessments that reward class participation and demonstration of knowledge in more informal ways. For instance, meaningful participation represented one piece of Tom’s final grade, which in turn reinforced his decision to stay on a STEM career path. Teaching and learning in STEM is ripe for assessment modalities that allow students to demonstrate competencies in multiple ways, such as proposing and performing experiments, using math in applied ways, and designing and testing systems. One appeal that community colleges have is greater opportunities for hands-on coursework, which tends to enhance student success in STEM coursework and beyond (Wang et al., 2017). The students in the Biotechnology program from our study were able to manipulate stem cells, test cells and tissues for genetic changes, and understand how bacteria can be used to study diseases and manufacture drugs. Providing students with more authentic and applied research experiences aid in students’ career development (Phelps & Prevost, 2012) and align their knowledge and skills with STEM workforce expectations. In a sense, using hands-on activities and participation as part of a student’s evaluation may offer more meaningful experiences and training than solely relying on grades as a proxy for engagement and ability.


RECONCILING WORK AND SCHOOL: MAXIMIZING SUCCESS TO PERSIST IN STEM


The impact of working full-time on college student retention is well known (e.g., Astin, 1999; Bers & Smith, 1991; Pascarella & Terenzini, 1998) and rang true in our study. Community colleges offer flexibility in course schedules and availability (Cohen et al., 2014), which at first glance may benefit community college students who often work full-time (Bailey, Jenkins, & Leinbach, 2005; Cohen et al., 2014). Our findings, however, revealed something quite different. Students employed full-time found it challenging to pursue a STEM pathway when STEM courses were primarily offered during the day. These limited course options created a conflict with the students’ work schedules. Jim, Gertrude, and James all wrestled with how to fit courses into their existing, demanding schedules. In the end, Gertrude and Jim decided that they might have to concede and either alter their work schedules or take time off. These can be extremely challenging decisions to make, especially when community college students not only work full-time, but also have other responsibilities, including supporting their families and covering other living expenses (Bryant, 2001; Cohen et al., 2014).


These issues raise questions around how community colleges can help students stay on the road to becoming scientists considering their potentially restrictive schedules, multiple responsibilities, and in many cases, lack of financial means. Online courses or providing classes at night or on the weekend present several viable ways in which to accommodate students with packed schedules (Perna, 2010). At the same time, there are learners who need or desire in-person interaction. Therefore, it is important that community colleges and instructors make the most of class time for those with limited opportunities to be on campus. For community college students, social and academic involvement or integration tends to happen in a more concentrated manner within classrooms (Deil-Amen, 2011), which tends to be positively related to student success (Deil-Amen, 2011; Wang, 2013a). Although working full-time may initially appear to be a barrier to students’ paths in STEM, this can be minimized through intentional support and interaction at the classroom level.


Furthermore, to address barriers to pursuing STEM due to employment and finances, community colleges should consider resources in the form of institutional or federally funded scholarships or financial assistance opportunities aimed at students in STEM (e.g., National Science Foundation). These types of aid helped students in our study choose or reaffirm their STEM paths. As a result, it is important to recognize and take deliberate action to help working and low-income students find and explore various financial sources, as they can create and support a valuable, diverse, and untapped STEM talent pool.


NAVIGATING THE STEM PATH: CULTIVATING TRANSFER CAPITAL TOWARD TRANSFER EFFICACY


Resonating with prior scholarship (e.g., Laanan, 2007), our study highlights the significance of transfer capital in ensuring a successful transfer process for community college students. Furthermore, transfer efficacy emerged as an influential factor and revealed that students’ transfer efficacy was rooted in accumulated transfer capital—what students came to know about the transfer process. For instance, the students in the study noted transfer capital they gained in the form of knowledge about articulation agreements and credit transfer between institutions. Through this knowledge and assurance of credit transfer, the students were able to find confidence and ease in navigating the transfer process. At the same time, transfer capital, in the form of information alone, could impact students’ transfer efficacy in an adverse way. As an example, J.J. accumulated knowledge about grade requirements for transfer; however, after learning of the stringent grades needed, he felt much less confident about his ability to transfer to his desired institution. Transfer capital represents an intricate construct that can work toward or against students’ paths in STEM.


Although increased transfer capital might lead to higher transfer efficacy, more information does not necessarily help students feel more confident about transfer. This raises questions around how to cultivate transfer capital in a way that works toward higher transfer efficacy. Forming relationships at the community college has the potential to benefit students’ self-efficacy (Moser, 2013), which may arguably extend to transfer efficacy. It is important that advisors help cultivate students’ transfer efficacy via holistic and contextualized transfer capital, not merely presenting hard facts. In addition to presenting transfer information, Jim’s advisor encouraged regular conversation before and as he approached the point of transfer. As a result, he felt his advisor had his best interests in mind, which helps students feel assured and confident in realizing transfer. In cases where students doubt their ability to transfer due to transfer requirements, advisors can help map out several transfer options that lead to desired destinations and improved transfer efficacy.


STEM PERCEPTIONS AND MINDSETS: OVERCOMING ADVERSITY AND FOSTERING SUCCESS


Often times, STEM fields are seen as stressful, with competitive or exclusive learning environments that can discourage women (Packard et al., 2012; Seymour & Hewitt, 1997) and students of color (Espinosa, 2011) who might otherwise pursue STEM. Gwyneth from our study experienced moments in which people told her that chemistry would be too difficult and that it was a “guy’s thing.” To create a scientific community that is more diverse and supportive, instructors and other institutional actors must be mindful of this diversity and make every attempt to not only accommodate, but also actively encourage these differences (Jain, Herrera, Bernal, & Solorzano, 2011). Moreover, the “scientific personality” or professional identity that individuals develop during their education and career (Jackson & Laanan, 2015) can be disrupted (Carlone & Johnson, 2007), reinforcing the need for increased support for all students, especially those underrepresented in STEM. For women and students of color, this means moving away from gendered and racialized perceptions and cultivating a strong science identity early on, potentially through female role models within and beyond the classroom, which significantly and positively impacts recruiting and retaining women in STEM (Drury, Siy, & Cheryan, 2011; Stout, Dasgupta, Hunsinger, & McManus, 2011). In doing so, these students can build strong self-perceptions and identities that will allow them to maintain a smoother path in STEM.


Although institutions and staff are responsible in facilitating persistence and success in STEM, students also survive or thrive in these disciplines depending on how they react to the adversity that is often inherent in taking a difficult academic pathway. Several women in our study were determined to continue in the sciences despite the potential difficulty and gendered biases. While these obstacles might threaten to derail some women’s STEM pursuits, others see them as opportunities to rise to the challenge. In light of this finding, instructors, advisors, and other institutional staff need to frame the rigor of STEM fields as an intellectually stimulating part of the educational journey in these fields (Morganson, Major, Streets, Litano, & Myers, 2015), while they tap into, cultivate, and celebrate students’ agency and resilience. By doing so, students can move beyond the fixation on STEM being too difficult (Morganson et al., 2015) and toward embracing the challenges as opportunities for personal growth and progress.


CONCLUSION


Our mixed methods study illuminated significant factors shaping the road to becoming a scientist, as well as how those factors exerted their influences within the contexts of students’ educational journeys. Combining ANN and qualitative interviews expands the methodological repertoire for studying STEM education in the community college by deeply exploring and reflecting the topic’s complexity. Through this approach, we were able to establish the significance of factors without presuming directionality and leveraged the interview data to disentangle how these factors functioned independently and together in sophisticated and nuanced ways. Future research on this topic would benefit from further inquiry to distill these factors across various student groups. Moreover, additional exploration of the intersections across factors would help researchers, policymakers, and practitioners work toward efficient supports and services to holistically address the factors to benefit student success in STEM.


Transfer from community colleges into STEM programs at four-year institutions represents a crucial point of entry and access to STEM careers requiring a baccalaureate. However, the democratizing potential of community colleges in the STEM context is not straightforward. Our study brings forth a deeper understanding of community college students’ STEM pathways, including the many plot twists and processes involved to overcome challenges and maintain progress. Although the road to becoming a scientist is not an easy one, students and institutions can work together to dismantle potential obstacles and forge a clear, supportive path to success.


Note


1. A neural network generally consists of three layers: input, hidden, and output layers. An input layer contains a set of independent variables and an output layer consists of dependent variables. A layer located in between the input and output layers is named a hidden layer because it is not directly observable. Each element in each layer in a neural network is generally referred to as a node or neuron. For ease of interpretation, in this paper, we use the terms independent and dependent variables for the elements in the input and output layers.


Acknowledgement


This study is based on work supported by the National Science Foundation under Grant No. DUE-1430642. The authors thank Falon French and Ning Sun for assistance with research.  


References


Adelman, C. (2005). Moving into town—and moving on: The community college in the lives of traditional-

age students. Washington, DC: U.S. Department of Education.


Adelman, C. (2006). The toolbox revisited: Paths to degree completion from high school through college.

Washington, DC: U.S. Department of Education.


American Association of Community Colleges. (2016). 2016 fact sheet. Retrieved from

https://www.aacc.nche.edu/wp-content/uploads/2019/05/AACC2019FactSheet_rev.pdf


Astin, A. W. (1999). Student involvement: A developmental theory for higher education. Journal of

College Student Development, 40(5), 518–529.


Bahr, P. R. (2013). The deconstructive approach to understanding community college students pathways

and outcomes. Community College Review, 41(2), 137–153.


Bailey, T. R., Jenkins, D., & Leinbach, T. (2005). Graduation rates, student goals, and measuring

community college effectiveness (CCRC Brief No. 28). New York: Community College Research Center,

Teachers College, Columbia University.


Bandura, A. (1997). Self-efficacy: The exercise of control. New York: W.H. Freeman and Company.


Bers, T. H., & Smith, K. E. (1991). Persistence of community college students: The influence of student

intent and academic and social integration. Research in Higher Education, 32(5), 539–556.


Bransford, J. D., & Schwartz, D. L. (1999). Rethinking transfer: A simple proposal with multiple

implications. Review of Research in Education, 24(1), 61–100.


Bryant, A. N. (2001). ERIC review: Community college students: Recent findings and trends. Community

College Review, 29(3), 77–93.


Carlone, H. B., & Johnson, A. (2007). Understanding the science experiences of successful women of

color: Science identity as an analytic lens. Journal of Research in Science Teaching, 44(8), 1187–1218.

doi:10.1002/tea.20237


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(2014),

173–186.


Clark, B. R. (1960). The “cooling-out” function in higher education. American Journal of Sociology, 65(6),

569–576. doi:10.1086/222787


Clark, B. R. (1980). The “cooling out” function revisited. New Directions for Community Colleges,

1980(32), 15–31.


Cohen, A. M., Brawer, F. B., & Kisker, C. B. (2014). The American community college (6th ed.). San

Francisco, CA: Jossey-Bass.


Cohen, J., Cohen, P., West, S.G., & Aiken, L. S. (2003). Applied multiple regression/correlation analysis

for the behavioral sciences (3th ed.). Mahwah, NJ: Lawrence Erlbaum Associates.


Creswell, J. W. (2013). Qualitative inquiry and research design: Choosing among five approaches (3rd

ed.). Thousand Oaks, CA: Sage Publications, Inc.


Crisp, G., Nora, A., & Taggart, A. (2009). Student characteristics, pre-college, college, and environmental

factors as predictors of majoring in and earning a STEM degree: An analysis of students attending a

Hispanic serving institution. American Educational Research Journal, 46(4), 924–942.

doi:10.3102/0002831209349460


Deil-Amen, R. (2011). Socio-academic integrative moments: Rethinking academic and social integration

among two-year college students in career-related programs. The Journal of Higher Education, 82(1), 54–

91.


Dillman, D. A., Smyth, J. D., & Christian, L. M. (2009). Internet, mail, and mixed-mode surveys: The

tailored design method (3rd ed.). Hoboken, NJ: John Wiley & Sons, Inc.


Dowd, A. C. (2012). Developing supportive STEM community college to four-year college and university

transfer ecosystems. In S. Olson & J. B. Labov (Eds.), Community colleges in the evolving STEM

education landscape: Summary of a summit (pp. 107-–134). Washington, DC: The National Academies

Press.


Doyle, W. R. (2009). The effect of community college enrollment on bachelor’s degree completion.

Economics of Education Review, 28(2), 199–206.


Drury, B. J., Siy, J. O., & Cheryan, S. (2011). When do female role models benefit women? The

importance of differentiating recruitment from retention in STEM. Psychological Inquiry, 22(4), 265–269.

doi:10.1080/1047840X.2011.620935


Espinosa, L. (2011). Pipelines and pathways: Women of color in undergraduate STEM majors and the

college experiences that contribute to persistence. Harvard Educational Review, 81(2), 209–241.

doi:10.17763/haer.81.2.92315ww157656k3u


Ethington, C. A, & Wolfle, L. M. (1986). A structural model of mathematics achievement for men and

women. American Educational Research Journal, 23(1), 65–75.


Freeman, S., Eddy, S. L., McDonough, M., Smith, M. K., Okoroafor, N., Jordt, H., & Wenderoth, M. P.

(2014). Active learning increases student performance in science, engineering, and mathematics.

Proceedings of the National Academy of Sciences, 111(23), 8410–8415.


Garson, G. D. (1991). Interpreting neural-network connection weights. AI Expert, 6(4), 47–51.


Garson, G. D. (1998). Neural networks: An introductory guide for social scientists. Thousand Oaks, CA:

Sage Publications, Inc.


González, J. M. B., & DesJardins, S. L. (2002). Artificial neural networks: A new approach to predicting

application behavior. Research in Higher Education, 43(2), 235–259.


Hagedorn, L. S., Cypers, S., & Lester, J. (2008). Looking in the review mirror: Factors affecting transfer

for urban community college students. Community College Journal of Research and Practice, 32(9), 643–

664.


Hagedorn, L. S., & DuBray, D. (2010). Math and science success and nonsuccess: Journeys within the

community college. Journal of Women and Minorities in Science and Engineering, 16(1), 31–50.


Hagedorn, L. S., & Purnamasari, A. V. (2012). A realistic look at STEM and the role of community

colleges. Community College Review, 40(2), 145–164. doi:10.1177/0091552112443701


Hoffman, E., Starobin, S. S., Laanan, F. S., & Rivera, M. (2010). Role of community colleges in STEM

education: Thoughts on implications for policy, practice, and future research. Journal of Women and

Minorities in Science and Engineering, 16(1), 85–96. doi:10.1615/JWomenMinorScienEng.v16.i1.60


Jackson, D. L. (2010). Transfer students in STEM majors: Gender differences in the socialization factors

that influence academic and social adjustment (Doctoral dissertation). Retrieved from Iowa State

University Graduate Theses and Dissertations Database.


Jackson, D. L., & Laanan, F. S. (2011). The role of community colleges in educating women in science

and engineering. New Directions for Institutional Research, 2011(152), 39–49. doi:10.1002/ir.407


Jackson, D. L., & Laanan, F.S. (2015). Desiring to fit: Fostering the success of community college transfer

students in STEM. Community College Journal of Research and Practice, 39(2), 132–149.

doi:10.1080/10668926.2012.762565


Jackson, D. L., Starobin, S. S., & Laanan, F. S. (2013). The shared experiences: Facilitating successful

transfer of women and underrepresented minorities in STEM fields. New Directions for Higher Education,

2013(162), 69–76. doi:10.1002/he.20058


Jain, D., Herrera, A., Bernal, S., & Solorzano, D. (2011). Critical race theory and the transfer function:

Introducing a transfer receptive culture. Community College Journal of Research and Practice, 35(3),

252–266. doi:10.1080/10668926.2011.526525


Kantardzic, M. (2011). Data mining: Concepts, models, methods, and algorithms (2nd ed.). Hoboken, NJ:

John Wiley & Sons, Inc.


Karabel, J. (1972). Community colleges and social stratification. Harvard Educational Review42(4), 521–

562.


Kuhn, M. (2008). Building predictive models in R using the caret package. Journal of Statistical Software,

28(5), 1–26.


Laanan, F. S. (2007). Studying transfer students: Part II: Dimensions of transfer students’ adjustment.

Community College Journal of Research and Practice, 31(1), 37–59.


Laanan, F. S., Starobin, S. S., & Eggleston, L. E. (2010). Adjustment of community college students at a

four-year university: Role and relevance of transfer student capital for student retention. Journal of

College Student Retention: Research, Theory & Practice, 12(2), 175–209. doi:10.2190/CS.12.2.d


Leigh, D. E., & Gill, A. M. (2003). Do community colleges really divert students from earning bachelor’s

degrees? Economics of Education Review, 22(1), 23–30.


Leigh, D. E., & Gill, A. M. (2004). The effect of community colleges on changing students’ educational

aspirations. Economics of Education Review, 23(1) 95–102.


Luan, J. (2002). Data mining and its application in higher education. New Directions for Institutional

Research, 2002(113), 17–36.


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.


Lent, R. W., Brown, S. D., & Hackett, G. (2000). Contextual supports and barriers to career choice: A

social cognitive analysis. Journal of Counseling Psychology, 47(1), 36–49.


Mills, J., & Birks, M. (2014). Qualitative methodology: A practical guide. Thousand Oaks, CA: Sage

Publications, Inc.


Mooney, G. M., & Foley, D. J. (2011). Community colleges: Playing an important role in the education of

science, engineering, and health graduates (Info Brief NSF 11-317). Arlington, VA: National Science

Foundation.


Morganson, V. J., Major, D. A., Streets, V. N., Litano, M. L., & Myers, D. P. (2015). Using embeddedness

theory to understand and promote persistence in STEM majors. The Career Development Quarterly,

63(4), 348–362.


Moser, K. (2013). Exploring the impact of transfer capital on community college transfer students. Journal

of The First-Year Experience & Students in Transition, 25(2), 53–75.


Myers, B. (2013). A study of the influence of student engagement on community college students’

intentions to transfer and STEM aspiration (Doctoral dissertation). Retrieved from Iowa State University

Graduate Theses and Dissertations Database.


National Economic Council and Office of Science and Technology Policy. (2015). A strategy for American

innovation. Washington, DC: Authors.


National Science Board. (2010). Preparing the next generation of STEM innovators: Identifying and

developing our nation’s human capital. Arlington, VA: Author.


National Science Board. (2015). Revisiting the STEM workforce: A companion to science and engineering

indicators 2014. Arlington, VA: Author.


National Science Board. (2016). Science and engineering indicators 2016 (NSB-2016-1). Arlington, VA:

Author.


Packard, B. W., Gagnon, J. L., LaBelle, O., Jeffers, K., & Lynn, E. (2011). Women’s experiences in the

STEM community college transfer pathway. Journal of Women and Minorities in Science and Engineering,

17(2), 129–147. doi:10.1615/JWomenMinorScienEng.2011002470


Packard, B. W., Gagnon, J. L., & Senas, A. (2012). Navigating community college transfer in science,

technical, engineering and mathematics fields. Community College Journal of Research and Practice,

36(9), 1–14. doi:10.1080/10668926.2010.495570


Paliwal, M., & Kumar, U. A. (2009). Neural networks and statistical techniques: A review of applications.

Expert Systems with Applications, 36(1), 2–17.


Patton, M. Q. (2002). Two decades of developments in qualitative inquiry: A personal, experiential

perspective. Qualitative Social Work, 1(3), 261–283.


Pascarella, E. T., & Terenzini, P. T. (1998). Studying college students in the 21st century: Meeting new

challenges. The Review of Higher Education, 21(2), 151–165.


Perna, L. W. (2010). Understanding the working college student. Academe, 96(4), 30–32.


Phelps, L. A., & Prevost, A. (2012). Community college–research university collaboration: Emerging

student research and transfer partnerships. New Directions for Community Colleges, 2012(157), 97–110.


Plano Clark, V. L. & Ivankova, N. V. (2016). Mixed methods research: A guide to the field. Thousand

Oaks, CA: Sage.


R Core Team. (2015). R: A language and environment for statistical computing. R Foundation for

Statistical Computing. Vienna, Austria.


Reyes, M. (2011). Unique challenges for women of color in STEM transferring from community colleges

to universities. Harvard Educational Review, 81(2), 241–263. doi:10.17763/haer.81.2.324m5t1535026g76


Rouse, C. E. (1995). Democratization or diversion? The effect of community colleges on educational

attainment. Journal of Business & Economic Statistics, 13(2), 217–224.


Seymour, E., & Hewitt, N. M. (1997). Talking about leaving: Why undergraduates leave the sciences.

Boulder, CO: Westview Press.


Starobin, S. S., & Laanan, F. S. (2005). Influence of precollege experience on self-concept among

community college students in science, mathematics, and engineering. Journal of Women and Minorities

in Science and Engineering, 11(3), 209–229.


Starobin, S. S., & Laanan, F. S. (2008). Broadening female participation in science, technology,

engineering, and mathematics: Experiences at community colleges. New Directions for Community

Colleges, 2008(142), 37–46. doi:10.1002/cc.323


Stout, J. G., Dasgupta, N., Hunsinger, M., & McManus, M. (2011). STEMing the tide: Using ingroup

experts to inoculate women's self-concept and professional goals in science, technology, engineering,

and mathematics (STEM). Journal of Personality and Social Psychology, 100(2), 255–270.


Teddlie, C. B., & Tashakkori, A. M. (2009). Foundations of mixed methods research: Integrating

quantitative and qualitative approaches in the social and behavioral sciences. Thousand Oaks, CA: Sage

Publications, Inc.


Tsapogas, J. (2004). The role of community colleges in the education of recent science and engineering

graduates. Washington, DC: National Science Foundation.


Wang, X. (2013a). Baccalaureate expectations of community college students: Socio-demographic,

motivational, and contextual influences. Teachers College Record, 115(4), 1–39.


Wang, X. (2013b). Modeling entrance into STEM fields of study among students beginning at community

colleges and four-year institutions. Research in Higher Education, 54(6), 664–692. doi:10.1007/s11162-

013-9291-x


Wang, X. (2015). Pathway to a baccalaureate in STEM fields: Are community colleges a viable route and

does early STEM momentum matter? Educational Evaluation and Policy Analysis, 37(3), 376–393.


Wang, X. (2016). Upward transfer in STEM fields of study: A new conceptual framework and survey

instrument for institutional research. New Directions for Institutional Research, 2016(170), 49–60.

doi:10.1002/ir.20184.


Wang, X., Sun, N., & Wickersham, K. (2017). Turning math remediation into “homeroom”:

Contextualization as a motivational environment for community college students in remedial math. The

Review of Higher Education, 40(3), 427–464. doi:10.1353/rhe.2017.0014


Wang, X. & Lee, S. (2019). Investigating the psychometric properties of a new survey instrument

measuring factors related to upward transfer in STEM fields. The Review of Higher Education, 42(2),

339–384.


Whalen, D. F., & Shelley, M. C., II. (2010). Academic success for STEM and non-STEM majors. Journal

of STEM Education, 11(1&2), 45–60


Yeager, D. S., & Dweck, C. S. (2012). Mindsets that promote resilience: When students believe that

personal characteristics can be developed. Educational Psychologist, 47(4), 302–314.

doi:10.1080/00461520.2012.722805




APPENDIX A

INTERVIEW PROTOCOL

1.

To begin, could you tell me a bit about your experiences as a student at [name of institution], since last fall?


Possible follow-up questions


a.

From your survey responses, it looks like last fall you were taking courses or enrolled in a program in science, technology, engineering, and mathematics, or STEM fields. How did you like that experience?

b.

Can you walk me through what it was like to be in one of your STEM classes?

c.

What courses are you taking this spring? How have you found them so far? What stands out to you? Can you explain?


2.

As things are now, what are your future plans for school?


Possible follow-up questions


a.

Have you thought about transferring from this college to a four-year school?

i.

What would you like to study at a four-year school?

ii.

Has anyone spoken to you about transferring?

iii.

What programs or services have you looked into to understand the process to transfer?

b.

(For those indicating not interested in transferring to a four-year school). Have you thought about attending other two-year colleges?

c.

As you think about or make plans related to school, who do you talk to and what resources do you use to help make these decisions? Could you give an example or two?

d.

What or who would you say is the biggest influence on your choices and decisions related to your education?

e.

Do you see any barriers to realizing your educational plans? If so, what are those barriers? How could they be addressed?

f.

What do you see as helping you achieve the plans you have with regard to your education?


1.

What kind of work or jobs would you like to do after you are done with school? [two-year or four-year, or just course-taking if that’s the plan]


Possible follow-up questions


a.

Do you have any work experience in the field that you’re interested in already? Do you know others who have jobs like the one/s you’re thinking of exploring?

b.

As you think about or make plans related to your future career, who helps you  make decisions? Do you have any resources you use or people that you go to? Could you give an example or two?

c.

What is the biggest influence on your career choice or plans today?

a.

Do you see any barriers to realizing your career goals? If so, what are those barriers? What are some ways that you might address them?

b.

What do you see was helping you in your career goals?


1.

Is there anything else you would like to share with me related to what we talked about, or any other aspects of your experiences as a college student?





APPENDIX B

SAMPLE OF DATA CODING AND THEME FORMATION

Factor

Raw Interview Data

Codes

Explanation of Codes

Theme

Attitude toward science

. . . we started learning about this DNA, and all of a sudden everything just clicked. . . . There’s this, awesome, and ever since then, I’ve just . . . every class that I’ve taken, I just liked it. I think it’s just once you find that thing that you actually really like and really want to just know everything about . . .

Everything just clicked

Liked every class

Finding an interest

Want to know everything

Student was able to make sense of science content and enjoyed subsequent classes, making them curious to learn more

Curiosity to learn

I’m like a compulsive tinkerer. I love taking things apart and trying to find out how things work. And I’m always working on some kind of project . . . And I’m like, I know nothing about this? I’d love to find out, you know.

Compulsive tinkerer

Taking things apart

How things work

Always working on project

Love to find out

Student likes investigating things and is motivated to continuously learn

I like to learn stuff, and I like learning new things. I read a lot. I love to read, and so just—as long as I’m learning something or doing something interesting I’m okay with it.

Like to learn stuff

Love to read

Learning something

Doing something interesting

Student likes learning and reading and is content to learn












Transfer capital

I went to a college career expo. They had a bunch of colleges representing their, their programs they offered, and I asked a lot of them if they can transfer.

College resource

Career information

Transfer

Student seeks information on what programs transfer

The need to seek out resources

Lately I’ve just been going right on to the webpage for the [4-year college] Zoology program and they have a required course list there. It may not be completely updated or accurate, but it gives me kind of a general knowledge on what it is I need to do.

Web-based resources

General knowledge

Transfer process

Course information

Student accesses resources to plan for transfer – what are the required courses for the degree?

[advisor name] pretty much knows the schedule for everything and some courses that are required they only offer certain semesters like in the spring semester so she’s really specific with my schedule. She’s like stick to the schedule otherwise you’re not going to make that deadline. So, I think without that advice I would probably just be lost. And I don’t know, they’re just really supportive.

Advisor supportive

Advisor knows everything

Advice important

Students seek information on transfer and rely on advisor for accurate information

Encouragement and support from advisors

I actually had a lot of teachers, I mean sorry, counselors. I had them, they helped me look for classes over at the university that would you know they helped kind of set up my schedule so when I got there, I mean besides the counselors over there helping me choose what classes will, the counselors here helped me line up my first year so that I could have a very easy path to go onto when I did transfer to the university.

Help determine classes

Set up schedule

Easy path with help from advisor





Cite This Article as: Teachers College Record Volume 122 Number 2, 2020, p. 1-50
https://www.tcrecord.org ID Number: 23000, Date Accessed: 1/22/2022 10:03:02 PM

Purchase Reprint Rights for this article or review
 
Article Tools
Related Articles

Related Discussion
 
Post a Comment | Read All

About the Author
  • Xueli Wang
    University of Wisconsin-Madison
    E-mail Author
    XUELI WANG is a professor in the Department of Educational Leadership and Policy Analysis at the University of Wisconsin-Madison. She studies college students’ learning, pathways, and success, with a particular focus on community colleges and STEM education. Her recent work includes “Toward a Holistic Theoretical Model of Momentum for Community College Student Success,” published in Higher Education: Handbook of Theory and Research and a co-authored piece, “Does Active Learning Contribute to Transfer Intent Among 2-Year College Students Beginning in STEM?” published in The Journal of Higher Education.
  • Kelly Wickersham
    University of Wisconsin-Madison
    E-mail Author
    KELLY WICKERSHAM is a post-doctoral research associate at the Wisconsin Center for Education Research at the University of Wisconsin-Madison. Her research addresses community college student pathways and success, including student pathways and learning in STEM. Her recent publications include co-authored pieces, “Exploring Sources and Influences of Social Capital on Community College Students’ First-Year Success: Does Age Make a Difference?” published in Teachers College Record and “Exploring the Relationship Between Longitudinal Course-Taking Patterns and In-State Transfer into STEM Fields of Study” in The Journal of Higher Education.
  • Seo Young Lee
    Prometric
    E-mail Author
    SEO YOUNG LEE is a psychometrician at Prometric. Her research interests include the application of measurement models and statistical methods to address issues in education. Her recent publications include co-authored work, “The Role of Aspirational Experiences and Behaviors in Cultivating Momentum for Transfer Access in STEM: Variations Across Gender and Race” in Community College Review and “Does Active Learning Contribute to Transfer Intent Among 2-Year College Students Beginning in STEM?” in The Journal of Higher Education.
  • Na Lor
    University of Wisconsin-Madison
    E-mail Author
    NA BEDOLLA LOR is a doctoral student in the Department of Educational Leadership and Policy Analysis at the University of Wisconsin-Madison. She is an Interdisciplinary Training Program Pre-Doctoral Fellow with the Institute for Education Sciences and a research associate at Wisconsin’s Equity and Inclusion Laboratory. Her research involves the use of mixed methods to examine the role of culture in higher education as it relates to student outcomes and higher education policy and practice.
  • Ashley Gaskew
    University of Wisconsin-Madison
    E-mail Author
    ASHLEY GASKEW is a doctoral student pursuing a joint PhD in the Departments of Curriculum and Instruction and Educational Leadership and Policy Analysis at the University of Wisconsin-Madison. Ashley’s research interests involve studying the long- and short-term impacts of for-profit education on students and faculty of color. Her most recent publications include a co-edited book, Critical Theory and Qualitative Data Analysis in Education and a co-written book chapter, “Cultivating Aspirational Capital Among Black Men in Community Colleges” in Engaging African American Males in Community Colleges.
  • Amy Prevost
    University of Wisconsin-Madison
    E-mail Author
    AMY PREVOST is an associate researcher at the University of Wisconsin-Madison. Her work focuses on educational pathways in STEM programs, student outcomes at the post-secondary level, including access to careers, and experiences that contribute to students’ abilities to transfer knowledge. Her recent work includes co-authored articles, “The Role of Aspirational Experiences and Behaviors in Cultivating Momentum for Transfer Access in STEM: Variations Across Gender and Race” published in Community College Review and “A Researcher-Practitioner Partnership on Remedial Math Contextualization in career and Technical Education Programs” in New Directions for Community Colleges.
 
Member Center
In Print
This Month's Issue

Submit
EMAIL

Twitter

RSS