Karlin, Ernest W.; Procopio, A. John (1986, ASQC) Karlin Associates, Westport, CT; ITT Corporation, New York, NY
Within the past three years there has been a great re-awakening in the United States as to the value and importance of applying statistical methods in virtually every area of business and industrial activity. The catalyst for this movement has been the widespread publicity concerning the successful use of Applied Statistical Methods by the Japanese in:- analysis of marketing data- product design and development- design and control of industrial processes- problem solving proceduresThis paper will focus on the use of the personal computer in one of the most important of these areas -- the analysis of data from statistically designed experiments; and in particular, the use of two-level Facotrials (and fractions of such designs) as powerful screening tools for "zeroing-in" on significant experimental factors.There are three important stages in the experimental process:1. Planning the strategy of the experiment - taking into cognizance: balance, randomization, replication2. Performing the analysis in a valid manner - to reveal the significant effects.3. Interpreting the analysis to arrive at the optimal conclusions and recommendations.This paper will pay particular attention to stage (2), which is often the most difficult hurdle to overcome when the people involved do not have extensive statistical training. The use of the computer - with powerful new software provides the means to achieve this stage successfully.