A Multi-Institutional Critical Examination of Undergraduate Engineering Academic Probation and Suspension Policies

This dissertation provided insights for engineering programs to consider as they take responsibility for improving undergraduate retention and graduation outcomes, especially for programs working toward the goal of educational equity. My findings suggested that engineering programs and researchers should add environmental factors to their models for student outcomes of retention and graduation following academic probation and suspension. I chose the three-manuscript format for this dissertation for each paper to build on the findings of the previous one. This was because there was a lack of engineering-specific literature on factors associated with retention and graduation for students put on academic probation and suspension. My critical lens was informed by Critical Race Theory (CRT) experts, and not explicitly the five CRT tenants. These experts recommended a process for posing of research questions to examine institutional and not student deficits as well as suggested methods, analysis and discussion to interrogate who institutions retain and graduate best.
Manuscript 1 - Suspension Description
An exploratory descriptive manuscript, this co-authored paper examined trends of over and underrepresentation of first-time academically suspended engineers’ (FASE) outcomes at the intersection of ethnicity and sex. Outcomes included return to engineering from suspension and graduation rates. The rationale for this study was based on a gap in graduation rates by student ethnicity and to kick-start a future examination of how suspension policies and institutional environments play a role in inequitable outcomes. Our sample included undergraduates matriculated at two selective engineering programs, both Predominantly White Institutions (PWI). Students were admitted between Fall 2009 to Fall 2018 with term data from Fall 2009 to Spring 2019. There were 1,199 FASE students among the 20,043 undergraduates in our sample. The two institutions suspended six percent of students. We reported FASE outcomes in aggregate, namely institutions attracted 21 percent of FASE returners to engineering. Of FASE returners with enough semesters in the dataset, institutions graduated 73 percent. Consistent with literature, males and Black students were overrepresented within FASE students. By disaggregating by ethnicity and sex, we provided more nuanced trends. For example, while males were overrepresented among FASE students, White males were not. Where females were underrepresented among FASE students, Latino and Black females were not. Asian males were the only group overrepresented among FASE students and underrepresented among FASE returners. To build off these descriptive statistics of FASE undergraduates, we recommended future research and interventions based in creating greater equitable outcomes in engineering graduation rates.
Manuscript 2 – Probation Multilevel Modeling
A descriptive and inferential quantitative study, this paper examined which students engineering programs put on academic probation (AP) and their outcomes at the intersection of sex and ethnicity. Outcomes included voluntary return to engineering the subsequent semester and graduation from engineering. This research is important for engineering programs to know how their warning of low academic performance in the first semester is associated with who they retain and graduate. My sample included undergraduates admitted from 2001 to 2013 across 18 residential institutions who housed engineering programs. Out of the 49,095 students in their first semester in engineering, engineering programs put 6,025 (12.27%) on AP. For modeling the likelihood of return and graduating, I used Eccles’s Expectancy Value Theory with a critical lens, including environmental factors of student and faculty composition as well as policy standards. I found that 21.37% of the variance in returning after probation could be explained by knowing which institution the student attended. Student composition, as measured by how much a student differed from the student composition by sex and ethnicity, was a significant predictor for both retention and graduation. Engineering programs retained and graduated students who differed from the engineering student body significantly less than those who resembled the student body. Faculty composition was a significant predictor for graduation only. Engineering programs graduated students in engineering who differed from the engineering faculty significantly less. My findings point to the need for retention and graduation models to include environmental and policy factors and understand that factors may differ by retention and graduation. Engineering programs, professional organizations, and granting agencies should also act on this data to examine best practices across institutions to improve engineering graduation rates equitably.
Manuscript 3 – Suspension Multilevel Modeling
This quantitative paper statistically described and examined the likelihood of retention and graduation for students who engineering programs put on academic suspension (AS) at the intersection of sex and ethnicity. Literature is silent on the topic of how well engineering programs retain and graduate students following required removal from their programs and institutions. Institutions who aspire to be educationally equitable need this information to better understand who they serve best through academic policy, especially at the intersection of sex and ethnicity. My sample was limited to nine engineering programs with student record data spanning years from 1988 to 2013 with 94,391 unique undergraduates who started in engineering. My sample was also limited to students with enough term data to be eligible to graduate, totaling 9,665 (10.24%) academically suspended engineering students. I utilized a motivational framework, Eccles’s Expectancy Value Theory, with a critical lens built in, as it included environmental factors. I chose to measure environmental factors through student composition as well as added policy standards. With 12.85% of the variation of retention and 8.40% of graduation explained at the institutional level, I chose a multilevel model to examine any association with environmental factors.
Student composition, as measured by how much a student differed from the student composition by sex and ethnicity, was a not a significant predictor for retention and was for graduation. Engineering programs graduated students in engineering who differed from the engineering student body significantly less. My findings suggest that environmental factors can vary in significance depending on a retention or graduation outcome and should be considered for future models. For engineering administrators, policy makers, I recommend a self-reflective examination of institutional data and, to what extent programs are willing, to take responsibility to move engineering education outcomes like degree completion toward greater equity.