Experiences Piloting a Program for Implementing High Impact Practices with Limited Resources

It is known that low-income, first-generation, and underrepresented students in engineering and computer science have rates of retention and graduation that lag behind their peers.A growing body of research has identified a range of high-impact practices and exemplar programs that have been successful in improving outcomes for these at-risk populations.Some areas that these practices seek to address include: financial need, academic preparation, sense of community, confidence, and professional identity.The challenge of emulating these successful examples is that their implementation can be costly and resource intensive.Over the past three years, a team at University of Detroit Mercy has initiated the Science & Engineering Equity Development (SEED) scholars program with 30 engineering and computer science students.This pilot program has leveraged existing on-campus resources and external industrial partners to implement a range of high-impact practices in a financially sustainable manner.Preliminary outcomes have shown promising results in that 87% of SEED program participants have been retained to engineering and computer science one year after entry to Detroit Mercy compared to a rate of 63% for our comparison group.Furthermore, SEED program participants have earned an average of 35.7 credits per year with an average cumulative GPA of 3.59 compared to 33.6 credits per year with a 3.27 GPA for the comparison students.While it is challenging to disaggregate the effect of the academic preparation of the students before their entry to the program, this paper investigates the contribution of program elements through an analysis of: (1) student attitudes and opinions, (2) student interaction with campus personnel and resources, and (3) student social networks.This evaluation shows SEED program participants have stronger connections to Detroit Mercy as evidenced by their interactions with the campus and the composition of their support networks.