A virtual factory teaching system in support of manufacturing education

Journal of Engineering Education, Oct 1998 by Dessouky, Maged M, Bailey, Diane E, Verma, Sushil, Adiga, Sadashiv, Et al

Results of the comparison are presented in table 2. There were three significant, uncontrollable differences between the two classes: the 1997 class was half the size of the 1996 class, the 1996 class was nearly two years younger on average, and the 1997 class was less strong academically, with an incoming average GPA over 0.3 lower on a 4-point scale than the 1996 class. These factors might work against each other, as a smaller class size and older population should benefit the 1997 class, but their lower GPA would predict lower performance. Other demographic factors were consistent across the two classes: most members were male, spoke English as a second language, had little manufacturing work experience, and were first semester seniors. There was also a difference in academic preparation between the two classes. More students in the 1996 class had taken the linear programming course than in the 1997 class, while more students in the 1997 class had taken the quality control class.

In terms of average study patterns, the 1997 class read the textbook more frequently and spent nearly twice as much time studying for the midterm than the 1996 class. Other study factors were similar between the two classes. Members in both classes paid few visits to the professor or TA, and spent between seven and eight hours on homework each week About 70% of the students in each class regularly worked on homework in groups.

In rating their own abilities or learning outcomes, the 1996 students felt more confident than the 1997 cohort in the area of demand forecasting, while the 1997 class felt more confident in the area of factory scheduling. The latter is perhaps a positive indicator of the benefit of the VFTS. In all other areas, there are no significant differences in self-ratings.

Every attempt was made to maintain grading consistency. When the 1996 finals were graded, a sheet was maintained of how to deduct points on each problem. The 1996 exams were reviewed problem-by-problem as the 1997 exams were graded, to ensure that similar errors received the same penalty. There are two significant differences among the quiz scores, one on machine scheduling, where the 1997 class outperformed the 1996 class, and one on Gantt charts (also used in scheduling), where the outcomes were reversed. Thus, while the former bodes well for the VFTS, the latter's results are perplexing. However, the VFTS does not include Gantt charts in its report generation window, and therefore should add very little to the student's understanding of them. The more striking difference comes in the comparison of the midterm and final grades. On the midterm, which included no material on factory scheduling, the 1996 class outperformed the 1997 class by over 17%. On the final, which included factory scheduling, the gap was lowered to only 4%. The dramatic improvement of the 1997 class in the second half of the course bodes well for the benefits of the VFTS. The final course grade was also significant, with the 1996 class scoring better; this result derives from the high weight collectively assigned to homework, quizzes, and the midterm in comparison to that of the final (35%).