How Did We End up Together? Evaluating Success Levels of Student-formed vs. Instructor-formed Capstone Teams

Effective team functioning is one of the key ABET criteria and is also essential for successful capstone design work. Existing teamwork enhancement practices focus on key factors such as contributions of team members, balancing skills and personality types, fostering a constructive team climate, and response to conflict; however, the best method for forming successful teams is still the subject of debate. In the Senior Capstone Design course at Northeastern University, no explicit instruction in team functioning is provided at present. Teams are typically studentformed when possible; however, the course coordinator needs to ensure that team formation outcomes align with the course constraints –such as of number of projects, number of faculty advisors, and team size of 4-5 students– and must form teams when students are unable to. In terms of project assignments, students rank possible projects, and although an attempt is made to give them one of their top choices, this is not always possible. In this work, the two types of teams, student-formed and instructor-formed, are examined to see if there are any differences in terms of design quality, project completeness/implementation, and final prototype grade. Assessment tools include the validated prototype scoring system previously developed by one of the authors. Teams were categorized based on how the teams were formed –instructor versus student, topic preference –students initial ranking of their assigned project, what percentage of the team members were actively chosen by other team members, and whether they were chosen based on skill or work style, and/or because of friendship, and the degree to which their assigned advisor was among their top choice. The team’s passion and commitment to their project was also assessed, using an operational definition of this engagement factor. Results show that teams who select team members themselves with an eye to the skills and work style of their team members have high final scores and also a high level of commitment to and passion for the project. Additionally, it seems that neither the actual topic of the project nor the team advisor necessarily affects the outcomes of the project. Findings will inform guidelines for capstone team formation and future coaching of the students in general and in capstone teams once formed. Results will help determine what type of team-formation protocol is recommended and the coaching intervention may improve the performance of potentially low-functioning teams. Introduction and Overview of Previous Work Capstone design teams have been studied by a number of authors. Dutson et al. in their review of capstone design courses noted that a majority of capstone courses use teams, as the ability to work in teams is seen as an essential skill for engineers. They observed that team formation schemes vary across programs from students being randomly assigned to groups, to students selecting their own team members, to instructor-formed teams. Instructors who assign students to teams may do so simply on the basis of interest in project topics, or may strive to consider skills, personalities, academic strengths, nationality, and gender among other characteristics. There have been a number of efforts to assign students to teams based on various student characteristics. Brickell et al. compared groups that were formed based on GPA and project interest to groups that were self-selected by the students. Both Brickell and other authors have observed that self-selected teams often perform worse than intentionally formed teams that take into account specific skills and academic ability. Other researchers have found that there was no one good way to form teams, and that having teams carefully crafted by the instructor was no guarantee of team success. The Meyers-Briggs personality test has also been used in various institutions to attempt to form groups that would potentially function more smoothly, although the link between personality and team performance is not always clear. Tools such as the CATME web based system for team formation and other institutionally developed tools have been used to allow instructors to match students to their most desired projects while controlling the mixture of skills and abilities in the team members. Some of these tools have specifically been used to develop high performance teams, with a reasonable rate of success in terms of the final result. However, in this case, the high performance teams that resulted were not formed by the instructor, but rather by students using the web-based tool. It is clear from an overview of previous research into capstone team formation that there is no one preferred and foolproof method for forming teams. Regardless of how teams are formed, there is some consensus about what characteristics make an effective team. Teams that are successful need to have individual responsibility and effective work habits. The team needs to have a mixture of abilities and skills among the team members and they need to have a clear goal with clear metrics and individual tasks and goals for each team to accomplish. High performance teams have similar characteristics, but with strong commitment to the project as well as to each other. The current work will show that carefully considered student-formed groups can foster a mixture of complementary skills and the deep commitment to the project necessary for success as well or better than instructor formed groups. Capstone Design at Northeastern University Team Formation and Project Assignment. In the Mechanical and Industrial Engineering Department at Northeastern University, the Capstone Design course is a required two-semester sequence for senior students. Projects can be proposed by industry sponsors, faculty members, or by the students themselves, subject to approval by the capstone administration. All approved projects are presented to the students in the first week of class. After the projects are presented, students are asked to submit a form ranking all the projects in terms of their preferences. Students may submit these ranking forms as a team of 4-5 students if they know who they want to work with. Students can also submit the forms as individuals, or groups of 2 or 3, with the understanding that the instructor will place the students into teams. The instructor makes every attempt to give each team their highest-ranked project. Individuals and small groups are combined based on common interests. Each project has an assigned advisor, with each advisor working with 1-3 teams. Generally, teams work within their discipline (industrial or mechanical engineering) but there are typically 1-3 projects per term that are interdisciplinary. The goal at the end of the capstone design sequence is to produce a working prototype –or process– that has been tested and validated. For the mechanical engineering students, this typically a physical prototype of their solution, such as a consumer product or a piece of lab equipment. This is then tested experimentally to determine whether specifications were met. For the industrial engineering students, their prototype may be a database, a facilities layout, or a new organizational scheme. These are validated by simulation in some cases or by data collection or observation in others, ideally with implementation. Prototype Score. At a point three weeks prior to the end of the course, teams must produce an Executive Summary that is sent out to the alumni jury that judges the final presentations. It is understood that the projects are not necessarily going to be completed at this point. Therefore, assessing the prototypes for completeness and extent of testing completed at this point has been found to be a good predictor of team effectiveness. The prototypes are assigned a score of 1-5 for completeness, where 5 indicates a completed prototype and 1 indicates that no prototype is likely by the end of the course. Testing was also rated on a 1-5 scale, where 5 indicates that prototype verification testing is complete and 1 indicates that testing was not planned or discussed. This validated rubric is presented in Table 1. In addition to scoring the executive summary, grades are determined based on oral and written reports and whether the final projects met the specifications laid out by the team and their advisor at the beginning of Capstone. Table 1: Capstone Executive Summary Scoring Rubric, also called the Prototype Score. Solution Score Verification score 5 = Fully developed solution 5 = Fully verified and validated 4 = Solution partially developed 4 = Verification substantially done 3 = In progress, solution expected by course end 3 = Verification planned and in progress 2 = Solution in progress, unlikely to meet all specifications by end of course 2 = Verification planned, unlikely to be complete by end of course or not started 1 = Working solution unlikely by course end 1 = Verification not discussed or planned