In This Chapter
Throughout Part III, I discuss Excel capabilities for testing hypotheses. The general process is that you formulate a hypothesis, gather data, and use a statistical technique to test the hypothesis.
Famed statistician John Tukey referred to hypothesis testing as confirmatory data analysis. He stressed, however, the importance of another process — exploratory data analysis (EDA). As the name suggests, in this process you explore data to find and illustrate features of interest, and perhaps generate hypotheses to investigate further.
Tukey catalogued a wide array of techniques for exploring sets of data and for presenting data sets in ways that make them easy to understand. In this Appendix, I show you how to use Excel to implement three of those techniques.
Whenever I return an exam, naturally I tell the class the mean and the standard deviation, and what a score means in terms of a letter grade. If you read Chapter 6, you know I do that in terms of z-scores.
What's missing, though, is a sense of how the distribution of exam scores looks. Even with a frequency distribution (Chapter 2), an individual doesn't know exactly where his or her score sits in the overall distribution.
Tukey's stem-and-leaf display is a way around this. It organizes data so that an entire distribution of scores is quickly and easily comprehensible.
The display breaks each ...