Online education is fast growing due to its accessibility and scalability, but engineering has fallen behind other fields in adopting and researching the online educational format. Student course-level attrition is a significant issue in online courses. The goal of this dissertation is to better understand the factors that impact course level persistence decisions among online undergraduate engineering students. Three different research methodologies were employed for this study: a systematic literature review (SLR), learning analytics and data mining, and multi-level modeling.The SLR focuses on understanding the temporal trends and findings from research in online engineering education. A total of thirty-nine articles published between 2011 to 2020 met inclusion criteria, and the synthesis of these articles revealed five themes: content design and delivery, student engagement and interactions, assessment, feedback, and challenges in online engineering. Theoretical, methodological, and publication trends across the forty articles were also summarized. Data for the second study was compiled from 81 courses contained within three online, ABET-accredited undergraduate engineering degree programs at a large, public institution in the southwestern United States. The students' learning management system (LMS) interaction data was utilized to create features that represent the amount of time students spent on different course activities and how those times differed from “typical” interaction patterns among students in the same course. Association rule mining was used to develop rules that describe the behavior of students who completed the course (i.e., completers) and those who opted to withdraw (i.e., leavers). The best measure of student engagement was determined to be the mathematical difference between the percentages of completer and leaver rules met by each student. Finally, multi-level modeling was used to examine the impact of interpersonal interactions on online undergraduate engineering students' course-level persistence intentions. The data for this study was gathered from online courses during the 2019-2020 academic year. Students completed questionnaires about their course and related persistence intentions twelve times during their 7.5-week online course. Students’ perceptions of the course LMS dialog, instructor practices, and peer support were found to significantly predict their course persistence intentions.

Although knowledge about effective teaching and learning exists, and theories of change strategies are considered, the lack of the understanding of the behavior of engineering faculty during curricular change remains a major contributor against robust efforts for change. In this work, faculty adaptability is conceptualized as self-regulation during curricular change. Faculty participants were recruited from two divergent curricular change contexts: one that is prescribed with interdependence while the other is emergent with uncertainty. In this study, attitude toward context’s strength is conceptualized along the four dimensions of clarity, consistency, constraints, and consequences of the context, while faculty’s self-efficacy and willingness for adaptability are conceptualized along the three dimensions of planning, reflecting, and adjusting. This study uses a mixed method, quantitative-qualitative, sequential explanatory research design. The quantitative phase addresses the question of “How does faculty group in the first context differ from faculty group in the second context in terms of self-efficacy and willingness for planning, adjusting, and reflecting?” The qualitative phase addresses the question of “How do faculty respond to curricular change, as exhibited in their activities of planning, adjusting, and reflecting during change?” Findings point to differences in patterns of correlations between attitude toward context with both self-efficacy and willingness across the two contexts, even though analysis showed no significant differences between attitude toward context, self-efficacy, and willingness across the two contexts. Moreover, faculty participants’ willingness for adjusting, in both contexts, was not correlated with neither attitude toward context’s clarity nor constraints. Furthermore, in the prescribed context, Group A faculty (self-identified as Lecturers, Senior Lecturers, or Adjunct Faculty) showed higher willingness for planning, adjusting, and reflecting activities, compared to Group B faculty (self-identified as Assistant, Associate or Full Professors). Also, in the prescribed context, Group A faculty showed no overall significant correlation with attitude toward context. This study has implications on the way change is conceived of, designed, and implemented, when special attention is given to faculty as key change agents. Without the comprehensive understanding of the adaptability of faculty as key change agents in the educational system, the effective enacting of curricular change initiatives will remain unfulfilled.