Critical Findings in the Development of the Community-engaged Educational Ecosystem

Abstract The Bowman Creek Educational Ecosystem (BCE2), a pilot project on which the Community-Engaged Educational Ecosystem was developed, is a community-university, cross-institutional partnership to attract and retain underrepresented groups in engineering and science, improve the quality of low-income neighborhoods, and build STEM literacy across the regional workforce. In its final year of an NSF Improving Undergraduate STEM Education (IUSE) planning grant, BCE2 has had outcomes across its domains of interest – from neighborhoods to student regional and STEM retention. In its final year, the Community-Engaged Educational Ecosystem expanded educational programming and partnerships into another city in the region as a prelude to scaled replication. Circling back to the mature pilot at the close of the grant, BCE2, researchers examined demographic differences of the influence of the programming. Held by the College of Engineering at the University of Notre Dame, the BCE2 theory of change drew from edge work on project-based active learning, community-engaged learning, and innovation ecosystem environments, with the intention of examining how multi-dimensional diversity would influence, positively or negatively, known outcomes (such as persistence) from these approaches. Early findings, however, indicated that aside from improved engagement with STEM there were also indications of increasing retention in the region – regardless of where the student originally was from [1]. As a workforce development grant in a legacy industrial landscape, this finding proved important to explore. Using mixed-methods data collection, including surveys and interviews, researchers specifically examined the differences in impacts of the programming between lower/higher socio-economic status (SES). The focus was on impacts of the educational ecosystem experience to self-efficacy with disciplinary content, intention to remain in STEM, and collaboration skills. Descriptive and inferential statistics, including a paired-samples t-test, are used to examine the differences from intervention on the construct(s) of interest. For certain analyses, researchers aggregated cohort data from the final two internship periods (2018, 2019) to increase power. Given the attitudinal quality of many of the constructs, researchers will also discuss checking for response shift bias and modifying the measurement approach accordingly [2, 3]. In addition, factors influencing attachment to place were collected through exploratory interviews; drawing from both the literature [4-7]and emergent factors from reflections, probable factors were integrated into the semi-structured interview template. For this paper and poster presentation, researchers will present findings from the analysis of the final cohort(s) of the original pilot program on aforementioned constructs, as well as an exploration of factors in place attachment for alumni.