Data Analysis for Quality Improvement: A Supplementary Text for Statistics Courses

This text is based on the premise that concepts of quality improvement and data analysis can be woven into almost any business statistics course as an almost costless add-on. Building on the statistical concepts being taught in the statistics course, this is a guide to quality improvement and data analysis that students can read, learn from and apply, all with minimal guidance from the statistics instructor. Instructors are free to concentrate on the specific goals of their statistics courses. The business statistics course can be elementary or advanced, theoretical or practical. The book serves to establish a link- or an additional link - between statistical theory and useful areas of application of the theory. Quality improvement is one of the most important of these areas of application. The introduction of real-life statistical applications to quality improvement gives ample opportunity to develop skills in data analysis, skills that are hard to develop without lots of practice. Although the original concept was to supplement business statistics courses, the book can supplement non-business statistics courses as well. Quality improvement has usefulness far beyond business. It can be applied to scientific research, education, health care, government or even individual jobs and everyday personal life. Similarly, data analysis is essential in all applications of statistics, not just applications to business. Since this approach takes little of the instructors time, instructors are free to concentrate on the statistical content of the course. The aim is not to revolutionize the instructor's course or to change its basic content and coverage, but to make any existing statistics course more effective by showing students how to put to prompt use the statistical idea of the course. The book can help to improve teaching by stimulating student interest in statistics. The needed statistical tools are simple, as this book demonstrates. The instructor's chief role is to get students started on the applications. The instructor can focus classroom emphasis on the specific statistical topics of the particular course, while assigning projects - of small or large scope - in which students learn more on their own. The instructions in this book are self-contained.

Table of Contents:
Part 1 Introduction: the idea of the book; how quality improvement fits in; lightning data sets; practical data analysis; review/reinforcement of statistical concepts; different types of courses; why "adventures"?; how to use the book. Part 2 A primer on personal quality: quality management - a quick overview; personal quality; a key tool of personal quality - the personal quality checklist; how checklists work; tips on effective use of checklists; improvement projects in personal quality; waste elimination by simple personal improvement projects; further discussion of the nature of waste; a simple strategy for personal improvement projects; improvements of personal work flow; further information about personal quality. Appendices: fitness; "quality is personal for leaders"; illustration of the just-in-time principle; getting a fast start in personal quality. Part 3 Simple in-control data sets: timing the passage of ten seconds; guessing the passage of ten seconds with benefit of feedback; guessing ten seconds without benefit of feedback; NFL Playoff Games, 1933-1966; Sundowner 5K Race, women's data; personal fitness measurements, illustrated by blood pressure; from "hot hand" to cold calculation - statistical lessons of wide applicability from sports; a process in control but with a possible outlier. Part 4 Simple departures from statistical control: first complications; counting data - the Poisson distribution; data arising as time intervals; the exponential distribution; analysis of defect rates; control chart for log normally distributed data; a new wrinkle - trend; another new wrinkle - periodic effects; a final new wrinkle - intervention effects; data analysis and quality improvement; patient self-monitoring of blood pressure - weekly averages. Part 5 Regression modeling of simple departures from statistical control: fitting trends by regression - the running application; fitting periodic effects by regression - the Ishikawa data; fitting intervention effects by regression - the putting data; another illustration of trend fitting - an intensive care example; the hypertensive patient in regression perspective; first multiple regression - allowance for severity in the intensive care application. Part 6 Interventions and randomized experiments: stepwise regression and variable selection - medical audit; applications of intervention analysis; a randomized experiment; intervention, experimentation and statistical control. Part 7 Cross-sectional data analysis and regression. (Part contents)