Board 349: Predicting Persistence in Engineering via Framing Agency

Abstract With perennial interest in broadening participation in engineering, much focus has been given to predicting persistence. Persistence intentions related to degree completion and career are commonly connected to developing a strong sense of identity in the discipline and feelings of confidence (or self-efficacy) about disciplinary practices. While psychosocial factors like identity and self-efficacy are often studied in engineering, they are less often linked to specific learning experiences, such as design education. Specifically, we extend typical models of persistence intentions to examine the effects of engagement in a core engineering practice—design problem framing. We conjectured that framing agency, the capacity to make decisions consequential to design problem framing 1, relates to engineering identity and engineering design self-efficacy. We sought to answer a research question: To what extent do framing agency constructs predict first-year and senior students' design self-efficacy, engineering identity, and persistence intentions? The current study uses structural equation modeling of survey responses to investigate the relationships between students' perceptions of their agency, identity, self-efficacy, and persistence (N = 991). We used the framing agency survey 2,3 and collected a national sample: 59% (583) were first-years and 31% (305) seniors; a majority (69%, 685) were men and from racial and ethnic groups that are privileged in engineering (82%, 809). Overwhelmingly, students reported working in teams (96%). We found that students expressed a strong intention to persist to degree (Figure 1, M = 6.46; SD = 1.67, all constructs on scale 1-7), not surprising given the percent of seniors in the sample. They also expressed generally positive intentions to persist in engineering careers following graduation (M = 5.80; SD = 1.23). Students reported high design self-efficacy, (M = 5.64; SD = 1.00) and engineering identity (M = 5.55; SD = 1.17). In terms of the framing agency constructs, students reported very high shared consequentiality (M = 6.22; SD = 0.96); high individual consequentiality (M = 5.98; SD = 0.72); high learning consequentiality (M = 5.67; SD = 0.98); and moderately high constrainedness (M = 5.02; SD = 1.23); the scale on tentativeness, which probes students' certainty that their role is to solve the problem as if it were well-structured rather than an ill-structured design problem, is reversed (such that a high score is aligned with design practice), and students expressed neither certainty nor uncertainty about this (M = 4.09; SD = 1.29). Several framing agency subconstructs explained variance in engineering identity and engineering design self-efficacy, which in turn predicted persistence intentions to degree and beyond. In particular, individual consequentiality—the sense that one is responsible for making decisions about the design problem—predicted engineering identity and design self-efficacy. For first-year students, shared consequentiality also predicted engineering identity. For seniors, shared consequentiality and learning as consequentiality predicted self-efficacy and also post-graduation persistence. Our results underscore the importance of design education experiences that provide students opportunities to direct problem framing.