This thesis research tested the utility of a quantitative measure of idea progression (B₁) in a computer-based training environment. Anderson (1971) only studied the different levels of the metric by simply rearranging sentences. In this experiment, however, careful rephrasing of a text created three statistically different training programs with B₁ averages of 0.31, 0.39, and 0.48 from instructional design, human-computer interaction, and human information processing guidelines. Three groups of industrial subjects (η = 8) were trained and tested to see which level maximizes worker comprehension of the chemical hazard concepts found in the Occupational Safety and Health Administration’s (OSHA) Material Safety Data Sheet (MSDS) Form 174. It was expected that subjects who received medium levels of complexity would show the highest post-test scores; however, no significant differences were detected (p>0.05). The negative results and corresponding small effect size are discussed along with future research suggestions. / Master of Science