This paper presents an experimental study aimed at investigating interaction effects affecting personnel in manual assembly. The main experiment with 36 subjects used a mixed method design which included a quantitative study, including time and errors as dependent measures, and a qualitative study, including workload ratings and a questionnaire. The overall task in the experiment was to assemble components on a pedal car. The main factors involved were assembly information (text & component numbers or photographs), material presentation (using structured kits, unstructured kits and material racks) and component variation (situations with and without component variation). It was found that performance, measured in assembly time, was best when combining photographs with no component variants and when using an unstructured kit.