A fractional factorial screening experiment to determine factors affecting discrete part process control

Abstract

While fractional factorial experiments are common in quality assurance practice, they are rarely used for human factors experiments. This study used a between-participants 2^6–1 fractional factorial design to examine how people control a discrete manufacturing process. The process simulation produced batches of parts that could be sampled for quality with changes made to the process based on the sample. Three factors reflecting process parameters, two representing individual differences and one representing an environmental variable were tested. Despite only using one participant per condition, many significant main effects and two-way interactions were found. Increasing task difficulty lowered performance, while individual experience and training in control techniques improved performance. Fractional factorial designs should take their place alongside full factorial designs for efficient human factors experimentation.

Keywords:

• fractional factorial
• human factors
• process control
• manufacturing