Development and validation of a novel survey instrument regarding stress in doctoral engineering students’ workplaces

Mirabelli, Joseph Francis

Permalink

https://hdl.handle.net/2142/121332 Copy

Description

Title

Development and validation of a novel survey instrument regarding stress in doctoral engineering students’ workplaces

Author(s)

Mirabelli, Joseph Francis

Issue Date

2023-07-09

Director of Research (if dissertation) or Advisor (if thesis)

Cromley, Jennifer G

Doctoral Committee Chair(s)

Cromley, Jennifer G

Committee Member(s)

• Corr, Catherine
• Xia, Yan
• Jensen, Karin J

Department of Study

Educational Psychology

Discipline

Educational Psychology

Degree Granting Institution

University of Illinois at Urbana-Champaign

Degree Name

Ph.D.

Degree Level

Dissertation

Keyword(s)

• Mixed methods
• stress
• mental health
• mental health and wellness
• doctoral students
• engineering education

Abstract

It is well known that doctoral students face high rates of attrition and prevalent mental health issues, contextual to the larger, more general crises surrounding mental health challenges and overburdened counseling services observed within academic spaces. These issues have been reported to have become more widespread within recent years, particularly in response to the COVID-19 pandemic. In engineering disciplines, issues related to mental health occur at rates comparable to other disciplines, however engineering students less commonly engage in help-seeking behaviors, particularly ones using university resources. While doctoral retention rates are higher in engineering compared to many disciplines, doctoral degree completion rates overall are concerningly low in the United States. Even for engineering disciplines, as many as one third of all PhD students are estimated to not complete their degrees. Attrition rates, counseling service use, and diagnoses of mental health disorders during doctoral studies have also been reported to be worse for individuals of minoritized identities. Specific, additional stressors beyond those inherently present in doctoral studies also exist for populations, such as for international students, who may navigate visa issues and who may face language and cultural barriers. Finally, reports suggest that experiences of high stress are related to both students’ attrition rates and to the rate and intensity of psychiatric symptoms and disorders, suggesting that explorations into the role of stress are crucial for addressing and understanding these crises. Prior research on doctoral student wellness, stress, and retention has focused on identifying and characterizing stressors and the consequences of stressors, but studies within this literature typically address single sources of stress, such as by exploring student-advisor relationships, or address particular populations of students, such as the experiences of Black students or of students within a single academic discipline. While this work which deeply explores single categories of stressors is crucial, few authors have contributed to work which identifies and addresses the broad range of stressors experienced by graduate students across identities, settings, disciplines, and at various programmatic phases. A full understanding of these stressors requires acknowledging a nexus of factors, including the social proximity of the source of stressors experienced by doctoral students, the timing of these stressors, these students’ perceived degree of control over stressors, the role of identity in determining the prevalence and severity of stressors, and understanding the doctoral student experience as both a workplace experience and learning experience. To overview my dissertation, in Chapter 1, I first motivate the importance of studying stressors experienced by doctoral students as an important and modern issue. This argument includes characterizing doctoral student environments as workplaces, akin to traditional jobs in many ways. I then explore a landscape of stressors impacting or contextual to doctoral students using a literature review which is organized by the proximity of stressors to students’ social contexts, which is described in Chapter 2. Following these literature-grounded chapters, I present my own investigations into the landscape of stressors experienced by doctoral engineers and the effects of stressors using two sequentially designed mixed methods studies. An overview of the full research method is provided in Chapter 3. The first study, described in Chapters 4 and 5, is a longitudinal, concurrently designed mixed methods study in which I interviewed and surveyed a cohort of engineering PhD student participants about their experiences with stress and stressors multiple times during an academic year. Following a mixed methods analysis, I contextualized the findings from this study with the broad literature review of previously reported doctoral student stressors described in Chapter 2. The product of that literature review and the results of Study 1 is a list of commonly reported engineering doctoral student stressors and a taxonomy of characteristics by which these stressors can be described. Next, in Chapters 6 and 7, I present Study 2, in which I describe the development of the Stressors for Doctoral Students Questionnaire in Engineering (SDSQ-E), a novel survey which measures the frequency and severity of stressors experienced by engineering doctoral students. In the Study 2 chapters, I will include results of a pilot of this survey for the purpose of validation using exploratory factor analysis, classical test theory, item response theory, and triangulation across the multiple mixed methods data sources within the two studies. In Chapter 8, I provide final thoughts, including the implications of these studies to theory, practice, and to researchers interested in replicating or expanding upon this work. I outline planned future validation and implementation work of this survey including a multi-institutional delivery of the survey in its engineering and its more general form. Future tests of the predictive and correlative power of the survey will compare it with measures including engineering culture, mental health challenges, quality of life, and intention to persist in doctoral programs. To more explicitly describe the research, in Study 1, a cohort of N = 55 doctoral students in engineering were interviewed four times during an academic year and were administered surveys eight times during the same year. This cohort of participants included students from each engineering discipline at the focal institution, who were at various stages of completion of their PhDs, from first year students to participants who defended their dissertations during the study. The cohort included a large population of international students and was representative of the focal institution’s students by gender, racial, and ethnic identities. The cohort exhibited excellent longitudinal retention of > 95% participation. Following a mixed methods analysis of the corpus of data from this cohort’s participation, a list of 11 categories of stressors consistently identified as most impactful for phase one participants was generated by compiling Study 1 results. This list of stressors is described, including by exploring the inherency of stress to a typical PhD experience, the social proximity of stressors to individual students, and how the stressors manifest as challenges or hindrances and the degree of support or control that students perceive regarding these stressors. Then this list of stressors was supplemented by a review of the stressors presented in Chapter 2. From this list, in Study 2, I developed the Stressors for Doctoral Students Questionnaire in Engineering (SDSQ-E), a 65-item, novel survey describing common sources of stress, and then administered the survey to doctoral students during two timepoints, resulting in N = 104 participants in Fall 2022 and N = 83 in Spring 2023. While this pilot study proves to be underpowered for full validation and fairness considerations, the SDSQ-E subscales exhibit high reliability and the findings triangulate closely with Study 1 results. Results of Study 1 included the identification of top stressors expressed by student participants at the focal institution and the development of a taxonomic system for classifying these stressors. Notably, these findings suggest that local context is important for determining which stressors are most salient in a doctoral program. Top stressors included research-related stressors such as dealing with experimental failure and making consistent research progress; advisor-related stressors such as communication with advisors and friction with advisors related to advising style; class-related stressors such as taking exams and navigating curricula not perfectly aligned to research interests; teaching assistant-related stressors such as grading consistently and fielding student questions; stressors related to time management, balance, and prioritization; campus life stressors such as finances, the importance of having a car to navigate the local community, and demands of family or partners; and stressors related to racial- or gender-based microaggressions, which were feared more often than experienced by the sample. I also observed these stressors to be time dependent with respect to both the time course of a semester, an academic year, a calendar year, and time in program. For example, international-student-related and family-related stressors increased in frequency in months nearer to school breaks, and classroom stressors were typically worse near final exams at the end of the semester. Additionally, I noticed fewer reported financial and microaggression stressors compared with other literature on doctoral student stress, perhaps the political and economic conditions surrounding the focal institution’s community. Implications of these local- and time-sensitive differences in stressors suggest that measurement at a variety of institutions and longitudinal study designs are necessary to fully explore the landscape of stressors faced by doctoral students. In Study 2, I provide evidence for the validity of the novel SDSQ-E survey including triangulation of the mixed methods data sources, classical test theory analysis of reliability and correlations between items and factors, an exploratory factor analysis and further exploratory analyses of subgroups, and item response theory tests including fairness considerations. In particular, the survey scales exhibit high reliability, and despite being underpowered, the data are suitably factorable. Further, latent factors within the survey indeed matched the categories of stressors identified qualitatively in Study 1. This dissertation describes a proposal for future validation work for the SDSQ-E, including a forthcoming confirmatory analysis. This study contributes to the literature on doctoral student retention, well-being, and workplace culture. The classifications of stress sources developed by this research can be used by researchers as a common taxonomy for discussing graduate student stress. Considerations of workplace models and social proximity models of stress can inform department- and university-level interventions and workplace locus of control considerations can inform practices for university counseling services. Departments or colleges can use the novel SDSQ-E survey measure produced in this study, or subscales within the SDSQ-E, to observe the severity of stressors within their programs’ local contexts. Departments can use the SDSQ-E to design interventions targeting specific student populations or aspects of doctoral program cultures. A general version of the survey has been developed and will be tested and validated for doctoral students across different disciplines.

Graduation Semester

2023-08

Type of Resource

Thesis

Copyright and License Information

Copyright 2023 Joseph Francis Mirabelli

Owning Collections