Open Book Vs. Closed Book Testing: An Experimental Comparison

This research adds to the ongoing dispute on what is the better method of assessing college students during examinations: openbook or closedbook. The open book assessment method is considered by many to be a realistic method that resembles the actual professional setting of demonstrating acquired knowledge in the field. On the other hand, the closed book assessment method has been used for centuries in traditional institutions as a rigorous method for knowledge assessment. In this research, engineering and business students from a large university in the southwestern United States participated in an experimental comparison designed to determine whether open-book or closed-book is the better approach to access academic knowledge during examinations. The Latin Square experimental design was used to block the variation due to the order in which students received the open-book and closed book treatments, as well as differences in material content tested on the exam. The research study produced mixed results for engineering students that were tested in three different classes: Statics, Mechanics of Materials and Quality Assurance. After adjusting for material content and treatment order differences, in two out of the three classes, the engineering students attained higher scores, a possible indication of achieving a higher learning level, when they were tested in the closed book approach. For the business side, the results indicate that the students attained higher scores, indicating a possible higher learning level, using the open book approach. The implications for this research can be extended to today’s online testing and certification environments, which are typically “open-book”. The open-book nature of online testing is viewed by some as a necessary evil that poses a validity threat, and by others as a simulation of the professional environment. As a direction for future research, this study could be followed up with future experiments that will attempt to reproduce the results in an online environment.