STATISTICS AND QUALITY IMPROVEMENT

Periods of intense practical activity in statistics have traditionally been times for reexamination of old ideas and development of new ones. Familiar examples are the ferment at Rothamsted experimental station in the 1920s and 1930s when, under the leadership of Fisher, the need for improved agriculture field trials resulted not only in the concept of designing experiments but also in the analysis of variance and many other fundamental ideas in statistical inference. Again, during World War II, the need for efficient inspection produced sequential tests. Also the need to screen many variables in small experiments led to the development of fractional designs and other orthogonal arrangements. This catalysis of progress by the practice half arrow pointing right over half arrow pointing left theory iteration runs throughout the history of statistics. The present quality revolution is now providing just such a stimulus for rethinking and re-evaluating many statistical ideas. Among these, discussed here, are the role of statistics as a catalyst to learning, the dual nature of statistical inference and some issues in experimental design.