MODELING STUDENT SUCCESS OF INTERNATIONAL UNDERGRADUATE ENGINEERS

Modeling student retention using entering secondary school academic performance metrics only is limited at best. Past research has shown that these variables can be somewhat informative, but are not the whole story. In order to expand our understanding of successful students, defined in this study as students who are retained and ultimately graduate with a degree in engineering, student retention and graduation modeling has been extended to include not only secondary school academic performance, but also self-reported affective and attitudinal measures. The Student Attitudinal Success Instrument (SASI), a 161-item survey assessing 13 specific noncognitive constructs, was developed based largely on existing instruments. The SASI is designed to provide data on noncognitive characteristics for incoming engineering students (a) prior to the onset of the first year and (b) for which higher education institutions may have an influence during students' first year. Data collected from this instrument have been found to be suitable for use in the development of predictive models of student retention and/or graduation, which is the definition of success in this model. The SASI is used to provide information about the academic preparation and affective characteristics of incoming first-year engineering students. Such systematically gathered information helps us assess the impact of University and programmatic decisions aimed at student recruitment, admission, retention, and ultimately the success of all students and, in particular, minority student populations. Though international students in engineering tend to have higher levels of overall retention and graduation versus any other majority or minority population, this study shows that the trend is in a concerning downward direction. In order to reverse this graduation trend, programs need to be expanded or created that are based on informed data decisions specific to student populations, such as international students. Understanding additional measures beyond admission metrics that lead to student success allows policy, programs or programmatic changes that increase overall student success. This study begins with a review of how this type of modeling was used to inform a change in admissions policy in the case of gender bias. Then, the techniques are expanded to international student success modeling. Though international and domestic students report similar levels of each success measure, the relative importance of each measure in predicting retention was different for these two student populations.