Multiple stakeholders are interested in measuring undergraduate student success in college across academic fields. Different metrics might appeal to different stakeholders. Some metrics such as the fraction of first-time, full-time students who start in the fall who graduate within six years, the graduation rate, are federally mandated by the U.S. Department of Education, Integrated Postsecondary Education Data System (IPEDS). We argue that this calculation of graduation rate is inherently problematic because it excludes up to 60% of students who transfer into an institution, enroll part-time, or enroll in terms other than the fall. By expanding the starters definition, we propose a graduation rate definition that includes conventionally excluded students and provides information on progression in a specific program. Stickiness is an even more-inclusive alternative, measuring a program's success in graduating all undergraduates ever enrolled in the program. In this work, programs are grouped into six academic fields: Arts and Humanities, Business, Engineering, Other, Social Sciences, and STM (Science, Technology, and Mathematics. Stickiness is the percentage of students who ever enroll in an academic field that graduate in the same field. We use the Multiple Institution Dataset for Investigating Engineering Longitudinal Development (MIDFIELD) 2023 which contains unit-record data for over 2 million individual students at 19 institutions. For the academic fields studied, Engineering has the highest graduation rate and third highest stickiness. Social Sciences and Business also have higher graduation rates and stickiness than the other fields. We also track the relative fraction of students migrating to and from each academic field. This paper continues our work to derive better metrics for understanding student success.

This research full paper explores the academic outcomes of undergraduate engineering students who leave their native countries and pursue their education in the USA. These “international students” are defined as students who were not citizens or permanent residents of the USA who enrolled in engineering programs in the USA. In this paper, we quantitatively analyze metrics for international and domestic students pursuing undergraduate degrees in one of the five most popular engineering disciplines in the USA: chemical, civil, electrical, industrial/systems, and mechanical using the Multiple-Institution Database for Investigating Engineering Longitudinal Development (MIDFIELD). MIDFIELD includes institutional records from over 1.7 million undergraduate, degree-seeking students in 21 programs at nineteen universities in the USA with over 85,000 men and 21,000 women enrolled in one of the most popular five engineering disciplines in the USA. Metrics used include representation at the start of university studies, initial engineering major choice, six-year graduation rate, and stickiness (the number of students graduating in a major divided by the number of students who ever enrolled in that major). Results are disaggregated by origin (domestic or international), sex (female and male), and major (chemical, civil, electrical, industrial/systems, and mechanical). Results show that there are more men than women in these disciplines and this is more pronounced for international students. In these disciplines, international students graduate at higher rates and have higher stickiness than domestic students. International females have the highest graduation rates. Industrial/Systems Engineering has the highest graduation rates and stickiness for all populations. Insights from this work can inform student services personnel and others committed to international student success.

The Multiple Institution Database for Investigating Engineering Longitudinal Development (MIDFIELD) is a unique research resource offering student record data at an unprecedented scale. This special session aims to introduce participants to MIDFIELD including the data that it contains and some key results from research using MIDFIELD, explore how to conduct research with such a resource, and explain how participants can access MIDFIELD. Understanding what data is available in MIDFIELD and how to access it will help researchers decide if this is a useful resource for their own research.

In this Full Research Paper, we propose a new definition of overpersistence in an engineering discipline and investigate its implications at one institution. Precisely defining overpersistence in both a conceptual and operational sense is a critical step in predicting overpersistence and identifying indicators that will allow for personalized guidance for students at risk of overpersisting. We have previously identified our population of interest as students who enroll at the institution as first-time-in-college students for at least one year, attend full time, have had six years to graduate, and have enrolled in only one degree-granting program. Within this group, we operationalized overpersistence by identifying students as overpersisters if they either (i) left the university without a degree or (ii) enrolled in the same major for six years and did not graduate. In this work, we revisit our definition of overpersistence using more recent data by reconsidering two groups of students in particular - those who spend only a short time in the discipline before leaving the institution (formerly classified as overpersisters), and those who spend a long time in the discipline but eventually switch majors (formerly excluded from the initial population). We conclude that the most appropriate definition of overpersistence at an institution with a first-year engineering program is when a student spends three or more semesters in their first discipline-specific major and does not graduate in that major within six years of matriculation to the institution. These results will be useful for researchers and practitioners seeking to identify alternative paths for success for students who are at risk of overpersisting in a major.

In this Research Full Paper we examine the reported experiences of Black students who are majoring in or switched from electrical (EE), computer (CPE), or mechanical (ME) engineering. Prior work has shown different persistence trajectories for Black students in these majors relative to White students, as well as differences between Black men and Black women. We surveyed 79 students at four institutions in the USA, three Predominantly White Institutions and 1 Historically Black University. In all, 33 students who had ever majored in ME, 27 in CPE, and 19 in EE completed a pre-interview survey that asked about aspects of the learning environment, faculty and peer relationships, and perception of belonging. Fifty-six students persisted in these majors while 23 switched to other majors. Compared to switchers, persisters are more likely to feel that the quality of instruction is higher, feel more encouraged by professors and peers to continue, and feel a greater sense of belonging in their departments. ME students are much more likely to experience group learning in their classes than either EE or CPE students and their ME peers are more likely to encourage them to persist. The difference in persistence between EE and CPE may be explained in part by the attraction of the computer science major as an alternative option for computer engineering majors; half of our CPE switchers switched to computer science. However, teaching quality may be an additional factor as CPE students perceived teaching quality to be lower than EE students did. Future research will explore these findings in the context of our in-depth interviews with these students.

This work-in-progress presents an overview of current research with the Multiple-Institution Database for Investigating Longitudinal Development (MIDFIELD) and strategies for inviting and supporting other researchers to use this database in their own research. MIDFIELD is a resource for the study of students that includes longitudinal, de-identified, whole population data for multiple institutions. This enables researchers to examine student characteristics (such as race/ethnicity, sex, socioeconomic indicators) and curricular pathways (including coursework) by institution and over time. Because the dataset contains records of all students matriculating over a period of time, researchers can study students across all disciplines, not just engineering. The MIDFIELD team aims to educate the broader research community, expand the network of researchers capable of conducting this research, and share innovative research methods in addition to the actual data. We have offered workshops at several conferences as well as the MIDFIELD Institute. The inaugural MIDFIELD Institute brought together researchers from across the USA for two days in 2019. In an evaluation at the end of the Institute, participants were pleased with the Institute and rated it highly. There was high agreement about the value of the content and delivery. Most participants reported that the instituted lived up to their expectations and was a good way to learn about how to use MIDFIELD data. Areas of improvement suggested by participants included making it longer and more challenging. Several of the MIDFIELD Institute attendees are presenting works-in-progress together at FIE 2020 to provide attendees with a range of examples of research that is possible with MIDFIELD.

This Research Work in Progress paper reports on Year 1 of a three-year mixed-methods study to identify institutional factors that promote increased retention of Black students in engineering curricula and to determine causes of their attrition. The paper outlines preliminary findings from 14 interviews of students who left computer, electrical, or mechanical engineering curricula. The students cited poor advising, academic challenges for which they were unprepared, lack of knowledge about their chosen major, financial and personal difficulties, perceived poor fit in the major, and issues around being a first-generation college student as important factors in deciding to change their major. They identified their race as an important factor in their selection of a university and in their sense of belonging at the institution, but not in their choice of a major or their decision to change majors.

The primary purpose of this Student Research Poster Paper is to discuss two grouping methodologies: cluste analysis and the Q-Methodology. Each of these statistical methodologies quantitatively group similar individuals but do so in two separate ways. In cluster analysis, individuals are grouped by optimizing proximity measures. For example, the single link clustering algorithm groups individuals together that have the smallest Euclidean distance from each other. In Q-Methodology, individuals are grouped by evaluating person-to-person correlations. For example, if two individuals have a correlation of 0.78, they are likely to be grouped together whereas individuals with a correlation of 0.09 are likely to be in different groups. In this paper, we outline multiple clustering approaches, the grouping mechanism in Q-Methodology, and discuss the differences between these two approaches when grouping participants. During this discussion, we will use an example from our own engineering education research to compare grouping results from the same data set. This paper contributes to the research field by describing and utilizing a relatively unknown methodology in engineering education. It will also add to our knowledge of cluster analysis techniques and compare those algorithms to another robust grouping method.

This Full Research Paper discusses ongoing work to develop a survey instrument to reliably assess undergraduate engineering student self-regulated decision-making. This work focuses on a second round of item expansion and refinement to the Decision-Making Competency Inventory (DMCI) to develop items related to learning from past decisions. The refined instrument was distributed to first-year engineering students enrolled in a large, public, land-grant institution located in the southeastern United States in the Fall of 2018. Of the approximately 1,200 students in first-year engineering courses, 883 valid surveys were randomly split into two separate samples for exploratory factor analysis (EFA) and confirmatory factor analysis (CFA). EFA results indicated a viable four-factor solution, which was explored with the CFA. The CFA results also indicated a four-factor model was appropriate. Improving this instrument will help researchers document and understand students' decision-making skills and how they relate to observed decisions like initial choice of major or change of major. A decision-making instrument will also be valuable in evaluating the effectiveness of interventions to help students build their decision-making competency and make adaptive choices.

This workshop, similar to our 2018 FIE workshop but revised to reflect participant feedback, introduces data and tools for investigating undergraduate persistence metrics using R. Student record data are from MIDFIELD, a database of registrars’ data from US institutions. The stratified data sample includes demographic, term, course, and degree information for 98,000 students from 1987 to 2016. The midfieldr package provides functions for determining persistence metrics such as graduation rates and for grouping findings by institution, program, sex, and race/ethnicity. The goal of the workshop is to share our data, methods, and metrics for intersectional research in student persistence. The workshop is designed for R beginners.

This workshop introduces data and tools for investigating undergraduate persistence metrics using R. Student record data are from MIDFIELD, a database of registrars' data from US institutions. The stratified data sample includes demographic, term, course, and degree information for 98,000 students from 1987 to 2016. The midfieldr package provides functions for determining persistence metrics such as graduation rates or program stickiness and for grouping findings by institution, program, sex, and race/ethnicity. The goal of the workshop is to share our data, methods, and metrics for intersectional research in student persistence. The workshop is designed for R beginners.

This full research paper presents a pilot study exploring how mid-year engineering students' perceptions of their future careers shift over one academic year. Perceptions of future careers determine many academic decisions students make, affecting student recruitment, persistence, and performance in engineering. Understanding how students' perceptions of their future careers change will give researchers, practitioners, and academic advisors insight into what influences students' choice of engineering major and what they choose to focus on in their courses. Students in sophomore engineering courses (n=71) took a survey measuring the clarity of and attitude towards their future career goals at two time-points, 2-3 semesters apart. Participants' perceptions shifted between four distinct ways of thinking about the future. Most participants shifted to either a well-defined and positive perception of their future possible career or an ill-defined less positive perception of their future possible career.

The Multiple Institution Database for Investigating Engineering Longitudinal Development (MIDFIELD) is expanding from 14 to about 100 institutions across the USA. This special session aims to introduce participants to MIDFIELD, explore how to conduct research with such a resource, and explain how participants can access MIDFIELD.

The Multiple Institution Database for Investigating Engineering Longitudinal Development (MIDFIELD) is expanding to include 113 institutions and is being redesigned and archived to be more accessible to researchers. This special session will describe how researchers can better use or gain access to MIDFIELD. At the conclusion of the session participants should be able to: describe MIDFIELD including common data elements, discuss how new variables can be derived from MIDFIELD, understand what is necessary to access the data on the Interuniversity Consortium for Political and Social Research, and define quantitative and qualitative data types and structures and outline research questions and methods of personal interest to them.

How are engineering students different from other students? This study determined which demographic and background variables influenced graduation for engineering students, as well as education, business, psychology, health, and social science students. Once determined, a comparison between disciplines was made to discover if or how engineering students differ from other students. To perform the analysis, logistic regression with backwards elimination was used. The results showed that high school grade point average was a significant influence with respect to graduation for each discipline except psychology and business. A student's American College Test (ACT) English score was a significant and positive variable for both business and engineering majors while ACT math scores were only statistically significant for engineering students. These findings could potentially be used for recruiting as well as retention.

Many studies have shown the importance of student self-efficacy in engineering retention. This work intends to contribute to the engineering education literature by exploring which student learning experiences are most likely to increase student confidence in engineering skills and if there are differences in these relationships among students of different gender, and achievement level, as measured by high school GPA and ACT Math scores. The study takes place at Louisiana Tech University through the Living WITH the Lab (LWTL) first-year engineering curriculum that serves over 500 students each year. Results indicate that frequency of hands-on activities is important for males and females in building confidence in their engineering skills. For females, however, the relationship is more complex.

Sixty-one sophomores were interviewed at six large public institutions to learn why they chose their institution and their engineering major. The institutions were categorized as either requiring a first-year engineering (FYE) program or allowing students to matriculate directly into a major. At these institutions, the first-year experience either required a common introduction to engineering course, required introduction to engineering courses that were not common to all majors or included an optional introduction to engineering course. The impact of the matriculation mode on selection of the institution and the presence or absence of a required first year course are studied. We find that cost of attendance is far more important than matriculation mode for most students choosing their institutions. Required and optional first-year courses, when taken, do tend to help students either affirm their prior choice of major or select an engineering major that suits their interests.

Using longitudinal data from eleven institutions in the U.S., this study explores the persistence of students in the two largest engineering disciplines: Electrical (EE) and Mechanical (ME). These programs have large enrollments of students but small percentages of women. Despite these similarities, enrollment and persistence in these majors is qualitatively different. In this research, we adopt an intersectional framework and consider both race/ethnicity and gender. Our results show that ME attracts more White students while EE attracts more Black and Asian students. Hispanic men and women are attracted in similar numbers to EE and ME. Overall, ME has higher graduation rates than EE and women have higher rates than men in both disciplines. Transfer students of nearly all race/gender groups are more likely to persist to graduation than starters in the same disciplines. Black and Hispanic female transfer students are particularly successful in EE and ME, which suggests enhancing the transfer pathway as a strategy to improve diversity. The success of ME starters causes a shift in the demographic profile between starters and graduates. ME could learn from EE how to diversify its enrollment and EE could learn from ME strategies to retain its diverse students. These findings suggest that program factors affect each race-gender group differently. Therefore, the success of recruitment and retention strategies may depend on considering both the target population and the discipline.

A new metric, “stickiness,” is proposed, tracking longitudinally all students who have contact with a discipline to determine the likelihood those students will “stick” to that discipline and graduate in it. This metric has the versatility to be relevant for students making contact with engineering through a variety of pathways. Stickiness exhibits significant disciplinary differentiation. Whereas earlier work has shown that Industrial Engineering is the most successful at attracting and retaining students, the disciplinary distribution of stickiness shows that Industrial Engineering is exceptional. Disaggregating by race/ethnicity and gender, much larger variations in stickiness are observed (as much as 48 percent), and positive and negative outcomes are identified where students in particular subpopulations are more or less likely to stick than expected. Aggregated by race/ethnicity and gender, the stickiness of transfer students ranks the disciplines in the same order as the stickiness of first-time-in-college students, but transfer stickiness exhibits less disciplinary variation and transfer students in all disciplines exhibit higher stickiness than first-time-in-college students.

Longitudinal data from ten U.S. institutions are used to characterize outcomes of three matriculation models: Direct Matriculation to a specific major (DM), First-Year Engineering programs (FYE), and Post-General Education Programs (PGE). Both DM and FYE programs show high persistence rates, but FYE programs are less likely to attract transfer students and switchers. FYE graduates are the most likely to stick with their first choice of major (after completing FYE requirements), followed by DM graduates who begin in undesignated engineering (taking extra time to decide), then DM graduates who choose their major as part of the matriculation process, and then PGE graduates. FYE students also have the shortest time to graduation. We conclude that encouraging students to associate with engineering or an engineering discipline from the start, yet maintaining the curricular flexibility to allow alternate entry points onto the engineering path improves persistence, accessibility, effectiveness of major choice, and time to graduation.

Transfer students make up a significant share of engineering college graduates, yet their persistence is seldom studied, largely because of the lack of longitudinal data. This analysis used longitudinal data from 11 universities enrolling large numbers of engineering students to investigate the demographic characteristics and educational outcomes of transfer students in engineering relative to non-transfers. We find that students who transfer to four-year engineering programs are more likely to come from under-represented minority groups (URMs) and less likely to be women, although both groups are over-represented at two-year colleges. The findings confirm existing research indicating that, on average, non-transfers outperform transfer students, and non-URMs outperform URMs. But we also find that URM transfers, and especially Black transfers, are no less successful than nontransfer students - indicating that the transfer pathway is an effective bridge to a four-year degree. This is partly true for women transfers who do as well as men but are outperformed by women non-transfers. Finally, we find significant variation in outcomes between full- and part-time students, which may be driving the observed differences by transfer status. Our results should inform debates regarding the efficacy of the transfer pathway in engineering, particularly for women and URMs.

The purpose of this study is to examine differences in student performance among engineering disciplines, as measured by term GPA's. Results indicate that: 1) Women outperform men in most engineering disciplines; 2) Student performance starts low at the freshman level, drops slightly at the sophomore level, and then increases over the junior and senior levels (without controlling for mortality); 3) Significant differences in GPA's remain between majors after controlling for relative SAT score, academic class level, race, and gender; 4) After controlling for major, relative SAT score, academic class level, and race, the gender gap in performance grows even larger.