2023 ASEE Annual Conference & Exposition

In multidisciplinary engineering capstone courses, students of a variety of disciplines work in teams to complete design projects. This study stems from a well-established industry-sponsored capstone design program where students work in multidisciplinary teams for two semesters in designing, building and testing projects. The process of placing students on teams is critical to ensuring students are successful in the program and project results meet sponsor expectations. Students are placed on teams by the program staff based on a project ranking survey they must submit after attending presentations about each project. These presentations are given by the faculty who will coach each project. The survey asks students to rank the projects based on their preferences and how their skills best serve the project objectives. Students must also state their commitments outside the program, GPA, academic major, minor and certificates, participation in student organizations and military experience, details on previous work experience, and describe any relevant software, hardware, or other technical experience and skills. Students are told that they have a high probability of being placed in one of their top three choices, and they are generally comfortable with their project placement. However, the process can be improved with respect to helping sponsors and coaches attract students through their presentations, and helping the students complete the survey to increase their opportunity to join their preferred project. Therefore, this study will evaluate the project descriptions, the coaches’ presentations, and the students’ survey results from the past four years to explore the following questions: (1) How much effort do students place on project placement? (2) How does student effort relate to their project choices during project ranking for placement? (3) Which elements of the project have a high impact on student project ranking for placement? (4) What makes a project engaging based on students’ written justifications for placement? This study will also consider the teaching record of the coach, the demographics of the students and coaches, and the project source (industry, academic or community service). As a mixed-methods analysis, this study will explore creating a predictive model that will utilize informative attributes automatically extracted from natural language, such as topic modeling with Latent Dirichlet Allocation (LDA), as well as qualitative attributes, such as the effort student placed in writing responses. These attributes will be organized in a structured way to train interpretable models such as XGBoost for project ranking predictions. These models will allow us to quantify which attributes are the most informative for making accurate predictions in student rankings, allowing us to determine which project attributes must be emphasized in order to maximize student interests. Awareness of student interests would help sponsors and coaches propose more compelling projects and provide insight to improving academic-industry collaborations. The results of this study will also lead to improving the guidelines for students to complete the survey to increase their odds of being placed on the team of their choice.