xviii What Is Statistics? And how should we prepare the cups? Should we make 5 each way? Does it matter if we tell the woman that there are 5 prepared each way? Should we flip a coin to decide even if that means we might end up with 3 prepared one way and 7 the other way? Do any of these differences matter? (5) Make and record the measurements. Once we have the design figured out, we have to do the legwork of data collection. This can be a time-consuming and tedious process. In the case of the lady tasting tea, the scientists decided to present her with ten cups of tea which were quickly prepared. A study of public opinion may require many thousands of phone calls or personal interviews. In a laboratory setting, each measurement might be the result of a carefully performed laboratory experiment. (6) Organize the data. Once the data have been collected, it is often necessary or useful to orga- nize them. Data are typically stored in spreadsheets or in other formats that are convenient for processing with statistical packages. Very large data sets are often stored in databases. Part of the organization of the data may involve producing graphical and numerical summaries of the data. We will discuss some of the most important of these kinds of summaries in Chapter 1. These summaries may give us initial insights into our questions or help us detect errors that may have occurred to this point. (7) Draw conclusions from data. Once the data have been collected, organized, and analyzed, we need to reach a conclusion. Do we believe the woman’s claim? Or do we think she is merely guessing? How sure are we that this conclusion is correct? Eventually we will learn a number of important and frequently used meth- ods for drawing inferences from data. More importantly, we will learn the basic framework used for such procedures so that it should become easier and easier to learn new procedures as we become familiar with the framework. (8) Produce a report. Typically the results of a statistical study are reported in some manner. This may be as a refereed article in an academic journal, as an internal re- port to a company, or as a solution to a problem on a homework assignment. These reports may themselves be further distilled into press releases, newspa- per articles, advertisements, and the like. The mark of a good report is that it provides the essential information about each of the steps of the study. As we go along, we will learn some of the standard terminology and pro- cedures that you are likely to see in basic statistical reports and will gain a framework for learning more. At this point, you may be wondering who the innovative scientist was and what the results of the experiment were. The scientist was R. A. Fisher, who first described this situation as a pedagogical example in his 1925 book on statistical methodology [Fis25]. We’ll return to this example in Sections 2.4.1 and 2.7.3.
Previous Page Next Page