Factors Impacting Retention and Success of Undergraduate Engineering Students
Research shows that engineering schools in the US are hemorrhaging qualified students: 45% switch to non-STEM majors before graduating, or do not complete a 4-year degree. Moreover, current approaches to reform undergraduate STEM education show little evidence of diffusion across programs. To insure persistence and success in engineering, a coherent model of the factors influencing success must be developed to retain and support engineering students, and STEM majors more broadly.
The goals of the project are: 1) to determine the impact of faculty attitudes and practices in foundational STEM courses on student attitude, learning, and persistence; 2) to determine the impact of support programs targeting first and second year engineering students, on attitude, learning, and persistence; 3) to determine the impact of departmental culture on student outcomes for all engineering programs as well as Mathematics, Physics, and Chemistry departments teaching foundational courses for engineers; and 6) to develop a systems model for characterizing the impact of the factors related to evidence-based instruction on student attitude, learning, and persistence.
Our approach employs multiple methods: 1) Surveys of faculty and student attitudes and experiences; 2) Observation of faculty practice; 3) Interviews of faculty and departmental administration; 4) Social Network Analysis of faculty relationships within and between departments; and 5) Historical analysis of institutional data including student course enrollment, grades, major, and engagement patterns across support programs. Faculty from the schools of engineering, arts and sciences, and education collaborate on data collection, analysis, and interpretation to insure findings are robust and representative of departmental culture. Key deliverables include identification of course and grade predictors of student persistence; the spread of evidence-based, effective teaching practices across faculty networks; the degree to which faculty relationships help or hinder reform at scale, and identification of key faculty clusters showing potential for development.
Key outcomes of the project include identifying factors influencing student success and persistence: 1) Courses (Multinomial regression shows that the first mathematics course taken in engineering programs is critical for retention: Students who begin engineering in courses below calculus, or who pass Calculus I with a grade of C or below have < 20 percent chance of graduating with an engineering major); 2) Time in program (85% of all leavers occur in the first two years); 3) Student support programs (Interrupted Time Series analysis shows persistence rates increased following implementation of support programs); 4) Faculty learner-centered attitudes and practices (Social Network Analysis shows that departmental affiliation, and faculty connectedness are associated with more learner-centered practices). This data is being used to design a faculty development program that connects participants within and across departments.
We are active within our University STEM community, designing a multi-department PD program for faculty. Workshops and assessment tools will be made available electronically for colleagues around the nation. In addition, we have published reports for ASEE, FIE, and other conference proceedings, and have presented our findings at meetings in the United States and internationally. These reports are being developed for publication in refereed journals.
Accessing institutional data, and combining datasets across multiple years.
Middleton, J. A., Krause, S., Beeley, K., Judson, E., Ernzen, J., & Culbertson, R. (2015). Examining the Relationship between Faculty Teaching Practice and Interconnectivity in a Social Network. In 2015 IEEE Frontiers in Education Conference Proceedings (pp. 2285-2291). El Paso, TX: IEEE.
Ernzen, J., Judson, E., Krause, S., Middleton, J. A., Culbertson, R. Beeley, K., & Chen, Y. (2015). What is the Effect of Establishing Programs that Address Sense of Belonging on Undergraduate Engineering Retention? In 2015 IEEE Frontiers in Education Conference Proceedings (pp. 2285-2291). El Paso, TX: IEEE.
Middleton, J. A., Krause, S., Beeley, K., Judson, E., Ernzen, J., & Chen, Y. C. 2015). Examining Relationships and Patterns in Pedagogical Beliefs, Attitudes and Classroom Practices for Faculty of Undergraduate Engineering, Math and Science Foundational Courses. To be published in Proceedings of the 2015 ASEE Annual Conference. Seattle, WA: ASEE.
Krause, S., Middleton, J. A., Judson, E., Ernzen, J., Beeley, K., & Chen, Y. C. (2015). Factors Impacting Retention and Success of Undergraduate Engineering Students. To be published in Proceedings of the 2015 ASEE Annual Conference. Seattle, WA: ASEE.
Middleton, J. A., Krause, S., Maas, S., Beeley, K., Collofello, J., & Culbertson, R. (2014). Early Course and Grade Predictors of Persistence in Undergraduate Engineering Majors. In 2014 IEEE Frontiers in Education Conference Proceedings (pp. 2285-2291). Madrid, Spain: IEEE. Article Available online [https://www.dropbox.com/s/36iu8ozq7nogi3z/FIE2014_Proceedings.pdf?dl=0]
Social Network Analysis of Faculty Connectivity and Its Relationship to Learner-Centered Attitudes and Practices
STEM Faculty who participate in pedagogical professional development report using learner-centered strategies, but in reality, few actually do. One reason for this may be lack of social support for facultyﾒs changing beliefs, classroom routines, and activities (Ebert-May, 2011; Borrego, 2010). McKenna, et al., (2009) found that greater collaborative reflection between engineering faculty and learning scientists, yielded greater shifts toward student-centered practices. Therefore, support in the form of collaborative interactions among faculty appears to be a promising strategy for diffusion of learner-centered practices within and across STEM departments.
This study reports the results of a Social Network Analysis (SNA) of faculty teaching courses in the first two years of undergraduate engineering programs at a major research university. The research question was: To what extent is the degree of social connectedness among faculty within and among departments related to attitudes about learner-centered instruction, and to subsequent practice? The results of this study are being used to design a professional development program that emphasizes developing communities of practice within and among STEM departments.
SNA measures the connectivityﾗthe relationships among people. It uses standard graph theory as a model. Participants were from Physics, Chemistry, Mathematics, and Education, and from six departments in Engineering. The sample began with 21 randomly-selected faculty from each department engaged in the instruction of first- and second-year engineering students. 81 additional colleagues were identified by the first level as people they utilize as resources for improving their instruction. Faculty were presented with the names of each colleague, and asked to state their relationship pertaining to learning of instructional strategies, assessment practices, and instructional technology. Attitudes were measured using the Approaches to Teaching Inventory (Trigwell & Prosser, 2004). The Reformed Teaching Observation Protocol (Piburn & Sawada, 2003) was used to measure implementation of Learner-Centered practices.
Results show that connectivity was positively associated with Learner-Centered attitudes and practices. In particular, faculty who had higher outdegree in the network showed highest Learner-Centered attitudes, lowest Teacher-Centered attitudes, and highest implementation of Learner-Centered practices. Differences were found regarding degree of connectedness within departments and across departments: Mathematics, had a mean indegree 0.45; First Year Engineering showed the highest mean indegree at 4.25. As a community, First Year Engineering faculty meet regularly, sharing tasks, strategies, assessments, and technological tools. This is not true among most other departments. Since few faculty across the studied University had many connections at all (mean indegree = 1.2 connections, SD = 1.2; mean outdegree = 1.8 connections, SD = 1.9), a prime strategy for increasing Learner-Centered attitudes and behaviors is to promote communities of practice within and among departments.
The results of this research is supporting the design of a multi-department PD program for faculty. Initial participants and leaders and promising practices identified are key strategic outcomes. We have published reports for ASEE, FIE, and other conference proceedings, and have presented our findings at meetings in the United States and internationally. These reports are being developed for publication in refereed journals.