Advice for Employers and Recruiters
Part 6: As the world interns: The impact of identity and social, economic, and cultural capital on college student internship engagement | Summary of Results, Implications for Practice and Policy & Future Directions for Research
This is the sixth and final article in this series, click here to go to the first article. If you’re searching for a remote internship, go to our search results page that lists all of the remote internships and other entry-level jobs advertised on College Recruiter and then drill down as you wish by adding your desired category, location, company, or job type.
CHAPTER FIVE: DISCUSSION
Research has established that internships are a credential correlated with valuable gains in college and employment (Aldas et al., 2010; Coco, 2000; Hurst et al., 2012; Kuh, 2008; Kuh et al., 2008; Sagen et al., 2000). Students who intern benefit from opportunities to reflect on their experiences, skills, and interests (Coco, 2000; Aldas et al., 2010). Internship experience is linked with better college retention and engagement, as well as a faster post-graduation job search and higher starting salary than non interning students (D’Abate, 2010; Coco, 2000). Impressively, those who have interned are measured to have more workplace engagement than those who did not intern, even many years after the internship is over (Gallup, 2014a). However, even with all of these benefits, little is known about who engages in these experiences and whether internships are available to all students who seek them (Allen et al., 2013; Frenette, 2013; Grugulis & Stoyanova, 2012; Keeton, 1977; O’Connor & Bodicoat, 2017; Perlin, 2012).
There were two purposes of this dissertation. The first was to compare the identities and experiences of students with internship experience with the identities and experiences of non-interning students. The second purpose was to determine whether particular identities or experiences predicted internship engagement. Using the framework of Bourdieu’s (1986) capital theory, I categorized 16 factors into larger themes of identity, social capital, economic capital, and cultural capital to see if these themes could predict internship participation. I used binary logistic regression and comparison of means through a t-test and chi-square analysis to test my hypotheses that:
- Economic, social, and cultural capital indicators have a significant, positive relationship with students’ participation in an internship.
- Dominant identities (male, able-bodied, heterosexual, and White students) have a significant positive relationship with participation in an internship.
- Students’ identities and economic, social, and cultural capital indicators significantly differ depending on whether the student participated in an internship or not.
This chapter provides a summary of results, explanation of the findings, and limitations and proposed changes for the current study. I follow with implications for practice and policy and directions for future research.
5.1. Summary of Results
My first research question was, “Do students’ identities and/or economic, social, and cultural capital indicators differ depending on participation in an internship?” A t-test and chi-square analysis revealed that several factors significantly differed between interning and non-interning students. Participants with internship experience were significantly more likely to be from outside of Vermont, to not identify as a firstgeneration student, have a higher cumulative GPA and median household income, and have higher scores on the NSSE engagement indicators Reflective and Integrative Learning, Collaborative Learning, Student-Faculty Interaction, and Supportive Environment. The results demonstrated support for my hypothesis that students’ identities and social, economic, and cultural capital indicators would differ depending on engagement in an internship.
My second research question asked, “What are the factors that significantly relate to undergraduate students’ participation in internships?” I used five separate regression analyses to measure the extent to which 16 chosen variables could predict internship engagement. The first regression used all 16 variables. The following four regressions split the 16 variables into categorical themes I developed based on Bourdieu’s (1986) capital theory. The themed regressions were based on identity, social capital, economic capital, and cultural capital.
Though all five regression models significantly predicted internship engagement, the four themed regressions had low statistical power. This led me to conclude that the four thematic areas I created were no better at predicting internship engagement than simply considering all of the 16 factors that I had selected to examine.
Although the themed regressions had only modest predictive power, several individual factors emerged as significant predictors of internship engagement only after I examined them within the smaller themed regression analyses. I did not find support for my hypothesis that dominant identities (male, able-bodied, heterosexual, and White students) had a significant positive relationship with internship participation. However, the variables of Quality of Interactions, Student-Faculty Interaction, residence, median household income, and GPA were all significant predictors of internship participation. GPA and residence had the most real-world significance in the current study, demonstrating that for every 1.0 increase in GPA a student was 2.74 times more likely to engage in an internship, and that students from out of state were 2.2 times more likely to intern compared to participants from the state of Vermont.
It is notable that GPA and state residence rose to significance in the smaller themed regressions yet were no longer significant in the larger 16 variable model. One reason for this may be due to Simpson’s paradox (Howell, 2010). Howell described Simson’s paradox as “the situation in which the relationship between two variables, seen at individual levels of a third variable, reverses direction when you collapse over the third variable” (p. 157). In other words, the full model may introduce one or more confounding factors that could obscure the significance of GPA and state residence on internship participation. Although I examined potential issues of multicollinearity (Table 2) and found none, a moderately correlated variable such as median household income with residence (a Pearson Correlation Coefficient of -.49, p<.01) or a weaker (yet significant) correlation of cumulative GPA and first-generation status (Pearson Correlation Coefficient of -.17, p<.01) may have been examples of confounding or mediating variables that obscured the significance of GPA and residence. In short, more research is needed to create a good-fit regression model that minimizes the effects of confounding factors.
5.2. Explanation of Findings
The most surprising result of the current study was the magnitude of influence of residence and GPA when predicting internship engagement. In the identity-themed logistic regression (Table 8), students from outside of Vermont were more than twice as likely to engage in an internship compared to students from within the state.
This dissertation focused solely on benchmarking internship engagement; it did not explore issues of student aspiration or potential barriers to participation. Thus, I did not investigate questions such as whether a student’s state residence affected their motivation for finding an internship. It could be the case that students from Vermont are simply less interested in this kind of experience. Or, perhaps students from Vermont have fewer opportunities to choose from or experience some other barrier to participation. In short, more research is needed to clarify the reasons for differences in internship engagement by geography and GPA. However, given that the information I have thus far, I have hypothesized reasons for these differences.
One possibility for internship disparity by state residence could be a difference in career aspiration. A relatively low rate of higher education enrollment in Vermont could serve as an analogous phenomenon to lower internship engagement for in-state students. Vermont performs extremely well with high school completion rates; the state graduated 89.1% of its high school students in 2016-2017, outpacing the national high school completion rate of 84.6% (National Center for Education Statistics, 2019). However,
compared to the national average, those from Vermont are much less likely to pursue post-secondary education. Only 52.3% of Vermont students enrolled in postsecondary opportunities immediately after high school in 2017, compared to a national rate of 66.7% (New England Secondary School Consortium, 2018). The Vermont Student Assistance Corporation (VSAC) (2015) published a special report on postsecondary enrollment, positing that although reasons for low enrollment rates are complex, geography and aspirations appear to be significant factors that influence students’ pursuit of further education. Although college-going rates are an entirely different research project, investigating the connection of rurality and education/career aspiration could shed light on differences in internship engagement by geography.
Allen and Hollingworth (2013) examined the intersection of young peoples’ geographic locations and career aspirations in the “knowledge economy” (that is, a creative and knowledge-based world of work). They described how “social class and place come together in powerful and complex ways to shape young people’s aspirations and capacity for mobility for and through work in the knowledge economy” (p. 513). Examining the interplay between aspiration and geography within the state of Vermont could provide additional context for internship participation rates.
Another possibility for lower engagement may be an issue of internship availability in rural areas. During the summer break, when many internships occur and students return to their hometowns, those from Vermont are bound for a much more rural setting compared to their out-of-state peers. Vermont is tied with Maine for being the most rural state in the United States, with 61% of the state population residing in a rural area (U.S. Census Bureau, 2019). Some research has already illustrated the “geographical
inequalities” (Grant-Smith & McDonald, 2017, p. 8) that can arise in internships, and how a student’s location can determine one’s access to internships (Allen et al., 2013; Allen & Hollingworth, 2013). For students hailing from a rural location, there simply may not be enough summer internship opportunities available to meet student demand, or a mismatch with a particular student’s major, interests, or skills.
Regarding GPA, there are several theories for why an increase in GPA predicts such a significance increase of internship likelihood. Knouse and Fontenot (2008) found that students with higher grade point averages were more likely to receive internship offers, though they acknowledged that it was unknown whether this was due to employers’ preferences, or because students with higher grades were more motivated or more capable workers. In contrast, Binder, Baguley, Crook, and Miller (2015) found that the opposite may also be true, and that internship experience positively impacts students’ grade point averages. In short, more research is needed to better understand the relationship between GPA and internship attainment.
Beyond GPA and residence results, it was surprising that there were a number of differences between the interning and non-interning participant pools (i.e., residence, first-generation status, GPA, median household income, and have higher scores on Reflective and Integrative Learning, Collaborative Learning, Student-Faculty Interaction, and Supportive Environment), but that most of those factors did not significantly predict engagement in an internship. More research on these issues, perhaps with a larger participant pool and fewer predictor variables, may provide more clarity on the factors that influence internship participation.
5.3. Limitations and Proposed Changes
All studies have their limitations, and mine was no exception. My study used data from the University of Vermont, limiting generalizability to other populations. Because I used data from only one year, the 2017 National Survey of Student Engagement, the study serves as more of a snapshot of a single moment, rather than a long-range understanding of internship participation over time. Examining data from all institutions taking the National Survey of Student Engagement or using multiple years of data would increase generalizability of the findings.
I also made the decision to exclude students in particular colleges from my participant pool. As detailed in Chapter Three, two colleges at UVM require internship or practicum participation, so I removed those colleges from the population to instead focus on students who had the option (and were not required) to engage in an internship. Again, this decision may also limit generalizability to other schools or programs.
My pool contained a relatively small number of participants (n=350) for the current study. A more robust study could include multiple universities or focus on a larger school to increase the size of the participant pool.
In this study, I converted two identity variables into binary variables instead of keeping all categorical options. These binarized variables included sexual orientation (LGBQ/straight) and race (Person of Color/White). I made this decision to ensure that I captured all experiences, rather than losing meaning from categories with a very small n. This decision collapsed many identities into generalized LGBQ and People of Color categories that did not represent the full breath of identity or experience. One recommendation to better capture and analyze a wide variety of identities would be to use Mayhew and Simonoff’s (2015) effect coding approach for examining self-identification categorical covariates. Mayhew and Simonoff’s (2015) coding procedure not only eliminates the need for essentializing one group’s experience over another’s, but also allows students to be analyzed based on their self-identified multi-raced selection. In addition, effect coding allows a more accurate assessment of the slopes representing any one racial group by including the partial estimates for this racial group when it is selected as part of a bi- or multi-raced category. (pp. 1-2)
While on the topic of coding identity, it would also be optimal to use gender rather than sex as a variable in future studies. I used sex as a research variable in my study because it was information that was easily accessible from the UVM Office of Institutional Research, and because there was no missing data for that variable. However, sex is reflected by one’s hormones and sex organs (Clayton & Tannenbaum, 2016), which are unlikely to be a significant influencer of internship participation. In contrast, gender comprises the social, environmental, cultural, and behavioral factors and choices that influence a person’s self-identity and health. Gender includes gender identity (how individuals and groups perceive and present themselves), gender norms (unspoken rules in the family, workplace, institutional, or global culture that influence individual attitudes and behaviors), and gender relations (the power relations between individuals of different gender identities). (p. 1)
Given the definition above, gender is a much more applicable variable to study for the purposes of educational research. Future studies could use gender as it is self-reported by participants on the National Survey of Student Engagement and also use Mayhew and Simonoff’s (2015) effect coding method to more comprehensively understand represented groups.
One noteworthy limitation of this study is the method I used to estimate participants’ income. Due to concerns about privacy and FERPA, I did not have access to information about students’ incomes or Pell eligibility, the latter of which is often used as a research indicator for low-income students. Thus, I used the median household income of each student’s zip code as a proxy for wealth. This is an imperfect system, as it does not represent that person’s individual financial reality, only the median income in their
whole zip code. If future research could utilize participants’ reported household income or eligibility for Pell grants, it would more accurately represent participants’ financial situations.
A future research study could also use multilevel exploratory factor analysis to create themes or commonalities in participants’ identities and experiences, rather than deciding on categorical themes a priori, as I did in this study. Conducted in SPSS statistical analysis software, factor analysis is used to summarize common themes, or latent constructs, amongst several larger variables (L. S. Meyers, Gamst, & Guarino, 2017). Latent constructs are what emerge during the factor analysis process (Reise, Ventura, Nuechterlein, & Kim, 2005), and the process of exploratory factor analysis could provide emergent themes within a set of variables. Multilevel factor analysis is a useful tool for examining a “series of building blocks that use sets of factor structures” (Goldstein & Browne, 2002). Goldstein and Brown described multiple levels of factor analysis as for “nested structures,” such as students within classrooms within schools (2002). The current study examines nested levels. There are students; their
socioeconomic backgrounds, academic capabilities, and social connections. Multilevel factor analysis could serve as alternate way to build and examine the latent factors that make up larger themes of Bourdieuan capital.
5.4. Implications for Practice and Policy
In this section, I outline approaches for practice and policy that address issues surfaced in the current study. The strategies I offer range widely in terms of the amount of funding and person-power required to implement them. My proposals are ordered from the most easily implementable to the more ambitious plans that require intensive resources.
When considering the conundrum of how to increase internship participation, a seemingly obvious answer would be to institute an internship requirement in academic programs. But, despite authoring a dissertation about the importance of engagement in and access to internships, I am reluctant to recommend an internship requirement. Hora (2018, 2019) outlined the numerous challenges with university internship requirements, noting that requiring students to intern could exacerbate financial distress for the many students who are already food-insecure, homeless, or struggling to pay rising tuition costs. Colleges’ limited ability to supervise and oversee internship opportunities means that an internship requirement would leave an increasing number of student interns vulnerable to exploitation and put many students in a difficult position of scrambling for an internship opportunity with minimal support (2018, 2019). Instead, Hora (2018) recommended focusing more on the quality of internship experiences, and slowly and intentionally growing the number of experiences to match student need.The amount of staff power to support internships is intensive; Hora noted that
Running a high-quality program takes a number of experienced employees who
can advise students, find appropriate placements, coordinate with employers, and
troubleshoot problems. For instance, it takes a full-time career-services director,
three employer-relations coordinators, and five part-time advisers to run the
internship program for the 2,550 undergraduates in the University of Wisconsin at
Madison’s business school. (para. 20)
Though it is not easy to develop high-quality, supported internship opportunities, several programs of this nature already exist at UVM. The College of Arts and Sciences has developed several “Communities of Practice” that offer local, highly-supervised, group internship opportunities within the fields of media, journalism, rural planning, legislation, and local research (University of Vermont, 2019b). The College of Arts and Sciences also partners with outside organizations to offer semester-long, credit-bearing
internship programs in Washington DC, New York City, and Boston (University of Vermont, 2019a). The Rubenstein School of the Environment and Natural Resources offers a Perennial Summer Internship Program, providing paid, intensive internships with environmentally-focused organizations (University of Vermont, 2019d). The Perennial Program Internship sites enter into a cost-share agreement with UVM, could serve as an incentive for the internship site to create a high-quality internship experience in exchange for accomplished and skilled interns and additional financial support to pay those students (University of Vermont, 2019d). Other organizations such as the Shepherd Higher Education Consortium on Poverty and the U360 Business Sustainability Internship Program collaborate with UVM offices to offer highly structured internship experiences to several students each year (University of Vermont, 2019e, 2019f). In short, intentional, organized, and high-quality internship experiences already exist at UVM. However, these experiences are offered through a wide array of UVM offices, with little coordination. No centralized office or website brings all of these experiences into one place. Creating an institutional structure to gather all of these programs would help consolidate resources and coordinate cohesive practices and communication.
Developing high-quality internships with Vermont employers is another recommendation that emerges from the study. However, in a similar way to how UVM struggles to bring internship programs together, Vermont-based workforce development groups are a complex, disparate web of programs, projects, and initiatives. A number of organizations are currently working on economic development issues in Vermont and considering internships as a tool to grow and develop a talented workforce. The following groups express a goal of developing workforce training, opportunities, and/or policy in the state of Vermont and include internship development as a strategy:
• Advance Vermont (formerly 70x2025VT)
• Brattleboro Development Credit Corporation Internship Program
• Careers CLiC (Connecting Learning in the Community)
• Institute for American Apprenticeships at Vermont HiTec
• State of Vermont Workforce Development Board
• VBSR (Vermont Business for Social Responsibility) Intern Program
• Vermont Community Foundation’s Pathways to Promising Careers
• Vermont Futures Project
• Vermont Internship Professionals Network
• Vermont Talent Pipeline Management
(Advance Vermont, 2019; Brattleboro Development Credit Corporation, 2019; Careers CLiC, 2019; Institute for American Apprenticeships, 2019; UVM Listserv System, 2019; VBSR, 2019; Vermont Community Foundation, 2019; Vermont Department of Labor, 2019; Vermont Futures Project, 2019; Vermont Talent Pipeline Management, 2019). Bringing these groups together in a consortium would allow them to communicate, share resources, and reduce duplication as they work on similar initiatives.
Several of the workforce development groups have researched economic policy issues in relation to internships and recommend growing them. However, none that I could find provide specific internship development tools for employers, who are the ones actually creating internships. Developing more internship opportunities within the state may also help in-state students find opportunities in or near their hometowns. A concerted outreach effort from units at UVM and/or the Vermont Department of Labor to Vermont-based organizations could grow the number of available internships. Gathering employers for a large internship best practices conference and providing an “internship toolkit” (UCLA Career Center, 2018) would also provide employers with specific steps to plan, promote, execute, and assess an internship program.
The most salient implication from the current study is that students from the state of Vermont need more encouragement and support to engage in internships to increase parity with their out-of-state peers. One strategy for increasing internship engagement would be to develop internship scholarship funding for students who hail from the state of Vermont. UVM currently offers internship scholarship funds to cover costs of living during a summer internship (University of Vermont, 2019c). Dedicating some of these funds specifically to in-state students could help support them in their internship endeavors.
Career counselors can also be powerful conduits for disseminating information to students about internship opportunities and accessibility. Before recommending internships, career counselors should listen carefully for potential obstacles to students’ engagement. For an example, it may help to return to Gloria and Marcus, the student amalgamations I presented in Chapter One. Gloria is an example of concerted cultivation; she is a student who understands that internships are a valuable opportunity that she can seize with her cultural, social, and economic capital. Gloria has the financial support from
her parents to engage in an international internship and the social capital to navigate professional networks to find other internship opportunities. In contrast, Marcus is a student for whom internships are an unfamiliar landscape. A career counselor working with Marcus could note the number of hours a week he works at Sodexo and Lowe’s to make ends meet and ask him about his need for payment in an internship. A career counselor might be able to recommend alternative paid internships or share information about UVM’s internship scholarships. Considering Marcus’ lower level of social capital, a career counselor might go beyond their typical supports and resources and offer to connect Marcus to industry professionals and the network to which he does not yet have access. To summarize, career counselors should be listening for signs of students’ real or perceived barriers to engagement and be ready with resources and supports that can deconstruct these barriers. To do that, career counselors need to familiarize themselves with high-quality, supportive, and paid internship opportunities and scholarships so that they can recommend them to students for whom unpaid work is not an option.
The results of this study also indicate that students with lower GPAs may need more support in their internship endeavors. It is difficult to know exactly why a lower GPA was so strongly associated with lower internship participation rates. It is possible that students with GPAs are simply less motivated or equipped to engage in an internship opportunity. Employers may also favor applicants with higher GPAs. However, though it is easy to assume that a higher GPA predicts internship engagement, it could also be the case that a lack of internship engagement has a detrimental impact on a student’s GPA. More research would illuminate possible causes for this correlation. Additionally, offering on-campus internship opportunities for students of all GPA levels could provide a first step in professional experience for students who may not otherwise be able to secure an internship.
If UVM were to build an on-campus internship, it would create more easily accessible internship opportunities with low barriers to entry. An on-campus internship program at Clemson (Nunamaker & Cawthon, 2018) could serve as a model, where the centralized Career Center developed placements within departments and offered matching dollars to fund the interns’ salaries. Offering priority to in-state and low-income students could help this program serve the needs of those who need it most.
A final recommendation would be to encourage and incentivize Vermont-based organizations to offer additional paid internship opportunities. Offering cost-matching for intern’s salaries, perhaps from the Department of Labor, could serve as an incentive for organizations to develop new internships. Rubenstein’s Perennial Internship Program could serve as a model for this kind of program, but with matching funds coming from a state-run program such as the state’s Workforce Education and Training Fund. Funding for interns’ salaries could possibly also come from an expansion of Federal Work-Study dollars.
5.5. Future Directions for Research
The theoretical framework for this dissertation was based on Bourdieu’s (1986) capital theory, Lareau’s (2011) observations of “concerted cultivation” in child-rearing, and Yosso’s (2005) community cultural wealth framework. Though the current study did not yield evidence to definitively state that social, economic, and cultural capital affected internship engagement, other methodological approaches and epistemologies may elucidate a better understanding of internship participation. Phenomenology could better capture participants’ lived experiences about expectations, experiences, and challenges around gaining internship experience. Rooted in philosophy, phenomenology is concerned with the subjects’ constructed consciousness and posits that there is no one objective reality (Creswell, 2013). Reality is the interpretation of the participants experiencing the phenomenon, and the researcher aims to describe the reality of the phenomenon from the participants’ perspective. Phenomenology has a main tenet of
respect for subjects, who are seen as co-researchers (Treager Huber, 2010). The person initiating the research is not an authority figure who is the expert on the phenomenon, rather, the participants are the experts on their own experiences and provide their narrative on the subject matter. Returning to the portraits I provided of Gloria and Marcus that I presented in Chapter One, I would like to interview the students whom I amalgamized for those depictions to more fully understand their lived experiences. I would especially like to investigate any influences of participants’ identities; perceived social, cultural, and economic capital; and possible experiences of parental concerted cultivation on participation in internships.
The current study examined trends in internship participation by identity and social, economic, and cultural capital. Although useful for discerning patterns in internship engagement, this study did not explore the reasons for why out-of-state students and those with higher GPAs were so much more likely to engage in internships. A study examining motivation and potential barriers to access would shed more light on the reasons for the disparity in internship participation. Specifically, it would be useful to ask about students’ desire to find an internship compared to how likely they were to secure an internship experience. If there is a population of students who are motivated to intern but cannot do so, it would be imperative to analyze the barriers to participation.
Another recommendation would be to examine other predictors in internship participation. I chose to study a selection of factors from the National Survey of Student Engagement, but there are still many others left to examine. It would be useful to analyze the impacts of hours spent working, providing care for others, time spent preparing for class, transfer status, the proximity of student residence to campus, membership in a fraternity/sorority, and status as an athlete or veteran. Analyzing trends in internship participation by major and industry of interest would also clarify which specific populations need additional support or opportunities. More research is also needed within professional majors such as education and nursing (which the current study excluded).
In general, it will be important for more researchers to investigate internships in the future. I mentioned in Chapters One and Two that the ambiguity around internship data collection, policy, and practices may be what allow a potentially inequitable system to persist and thrive. While there are ample data to demonstrate the successful outcomes of internships (see Chapter Two), there is still very limited information available about access to these opportunities. More clarity on internship access would help researchers, policy makers, employers, educators, and students make more educated decisions and address issues of inequity.
5.6. Summary
As internships become an increasingly common experience and credential, more examination of them is essential. Internships can provide powerful learning opportunities and important employment credentials for the students who engage in them. They serve as a cost-saving measure and a talent pipeline for employers. Colleges promote them as important experiences for career, academic, and personal growth. However, internships should not be viewed as a panacea for all problems.
The current study suggests that internships are not necessarily an experience that is attainable by all students. Analysis demonstrated that those from outside the state of Vermont were more than twice as likely to engage in internships compared to their instate person. For every 1.0 increase in GPA, students were also more than twice as likely to engage in an internship.
Though more research is needed to fully understand trends in internship participation, this study provides baseline data on the identities and experiences of students who participate in internships. My hope is that this dissertation also brings attention to an under-examined area of practice and sheds light on issues of educational equity.
— This is the sixth and final article in this series. Click here to go to the first article. This series of articles are courtesy of Amanda Chase. Amanda Chase is the director of strategic engagement for the collective impact organization Advance Vermont, where she works to increase access to postsecondary education. She also has a private consulting business, and previously worked as the internship coordinator for the University of Vermont. Amanda has worked with a wide variety of businesses to support their hiring goals, from one-person grassroots organizations to Fortune 500 companies. Her hundreds of individual career counseling clients have included high school students applying to first jobs, adults making significant career transitions, retirees seeking encore careers, and everything in between. She received a bachelor of arts in psychology from Hamilton College, a master’s degree in counseling from the University of Vermont, and an Ed.D. in educational leadership and policy studies from the University of Vermont. Her work has always centered on issues of equity and access in education and career development. To learn more about Amanda or to get in touch, visit her website.
