HOW I MIGHT LAY OUT A PROCEDURE NOTEBOOK

B. Gerstman
March 8, 1999

Although there is no single correct way to set up a procedure notebook, and its use is entirely optional, I would invest a good deal of effort in creating a brief well-organized three-ring binder with notebook dividers and perhaps an index that directs the user to each procedure he or she hopes to accomplish. For example, so far in the HS267, we have covered MEASUREMENT, STUDY DESIGN, EPI INFO BASICS, and "ANOVA and related techniques." Although this might be a good way to organize one's thinking, I might also consider combining the chapter on MEASUREMENT and STUDY DESIGN into a single divider section, since these are things one thinks about before collecting data. (Hey! That's a good title for the section -- Things One Thinks About Before Collecting Data.) This first section could including a checklist of what one needs to consider when planning a study, study as (a) how to create and remain focused a research question (it's OK to include things you've learned in other course), (b) how to define and measure things that need to be measured (consider variable types and measurement accuracy), and (c) how to consider study design basics (do I make my study observational or experimental, do I incorporate blinding, how do I select a representative sample, and so forth).

I would certainly have a brief section on Epi Info basics, such as how to set up a data base (QES file, REC file, data cleaning) and how to do some basic procedures in ANALYSIS (e.g., READ, LIST, BROWSE, UPDATE, DEFINE).

I might even include a page on "Analytic Approach," in which I remind myself that a data analysis must remain focused on the research question and that an analysis must include descriptive methods (summary measures, plotting, and so forth), estimation techniques, and formal tests. (Some students put basic definitions and principals they'd learned in the pre-requisite course here.)

The sections of the procedure on specific analytic techniques, in some ways, will be easiest to create. For each technique, address them, make certain you know the objective of the analysis, some background principals, how to compute or calculate the necessary statistics, their assumptions, and how to interpret them, and how to report their results. This type of thinking will become second nature over time.

The way you organize the procedure notebook, and what you put in it, is entirely up to you. This is the point! If I were to tell you what to put into the procedure notebook, or if you were simply to replicate the material in the Reader, the procedure notebook would not be very useful. By creating a procedure notebook of your own doing, you will be forced to do some necessary "soul searching" and to ask yourself "What do I hope to accomplish here?," "What is necessary and what is superfulous?" There is no substitute for thinking through the problem on one's own.