Researchers in quantitative social sciences have two resources to work with: methodology, and data. Far too often, they fall in love with the former and ignore the latter. Without good data, quantitative methodology yields little but "statistical artifacts." Computer programmers have a pithy saying for it: "garbage in, garbage out."

In this study of the relative tax burdens of the SBT, the limitations of the data are significant. An honest researcher will report those limitations and qualify the conclusions accordingly.

The data have three limitations due to the sample period, and one due to their aggregation. The sample period limitations prevent an accurate prediction of later relative tax burdens from the results in the sample period. The aggregation limitation will prevent applying the relative tax burden of a category to all the Members of that category.

However, the study still illuminates. A snapshot may not be as good as a motion picture, but it is better than nothing. Moreover, the few other "snapshots" available provide some check on the results from the sample period. Furthermore, the findings that depend on the structure of the SBT hold without regard to the sample period.

Ultimately, the methodology developed here provides a basis for evaluating future data. An accurate "motion picture" of the relative tax burdens awaits only a few more frames of data.

The 1985 Treasury report indicates that similar analyses will be available for years after 1981 "as soon as SBT data are available." The proper methodology to analyze the data is developed in the next section.