This paper presents ideas and observations about the infrequent use of factorial experimental designs; although the ideas presented are inspired by a case study, the paper is also conceptual in nature. The ideas build on the experimental practices at a large Swedish process industry manufacturer, where these practices were discussed with engineers using in-depth interviews. Factorial design is not used because, for example, statistical rigor is seldom required and because design of experiments methods are seen as too restrictive to accommodate and adjust to events occurring during long experimental campaigns. The interview results indicate that, while the scepticism of many respondents may be justified, lack of training, interest, and learning time are major obstacles to applying improved methodologies. Some ideas for overcoming these obstacles are suggested.
Validerad; 2015; Nivå 2; 20140526 (bjarne)