Economical Multifactor Designs for Human Factors Engineering Experiments.

Abstract : Experimental data collection plans are described that permit the study of from five to thirty experimental factors. These plans were selected from those employed in physical science research and were suitable for human factors engineering research. The maximum number of data collection points required for each basic design never exceeds three hundred and often is less. The method of employing these designs is a two-phase one. In the first phase, a large number of potentially critical factors are systematically screened quickly and economically in a way that identifies the more important ones. In the second phase, functions are obtained that relate the more important quantitative factors to operator performance. The need for this type of approach is justified on the basis of data obtained from an analysis of human factors engineering experiments published between 1958 and 1972 in the journal, Human Factors. Five principles that enable economical multifactor human factors experiments to be successfully conducted are stated and when assumptions are made concerning the application of these principles empirical data is provided in support. (Author)