Selecting teams: The potential of “the Grid”

Abstract

This paper investigates the use of a Computer-Based Team Selection Algorithm (CBTSA) in a capstone project course involving Information Systems, Human Resources, and Supply Chain/Operations undergraduate students. Data from three semesters of the course before this genetic algorithm was put into use are compared to data from three semesters after it was adopted. More than 50 % of the teams formed by the CBTSA were able to meet in part or in whole at times arranged by the CBTSA. Students in the CBTSA-formed teams experienced considerably less general time, procrastination and motivation problems, or specific scheduling conflicts. The percentage of students reporting that the team had to change its meeting schedule to accommodate them dropped to zero percent for the CBTSA-formed teams. Fewer CBTSA-formed teams seemed to break down with one or more low performing members. However, 42 % of the CBTSA-formed teams were created such that there were few really open meeting slots for every member. Some improvements to the genetic algorithm used in the CBTSA are investigated, but regardless of improvements, it is difficult to provide every team with rich meeting times and still achieve parity for the teams on a number of other criteria.