Do Honesty Tests Really Measure Honesty?
We now need to present some technical concepts from the fields of psychology and statistics. Some readers will already be familiar with these concepts, and they can merely glance through the next section. Others, however, may lack this familiarity. If that is the case, our intention is to provide an opportunity to learn. The idea is to be sure that all readers are able to "start on the same page" as we approach the studies dealing with whether honesty tests do measure honesty.
These technical concepts involve the logic of research design for selection purposes and the conduct of statistical analyses. They are important for two reasons. First, you need to understand the studies that constitute the bulk of this chapter, and why the research was conducted as it was. Second, these psychological and statistical concepts influence the language that test publishers and their representatives use in talking and writing about their tests. One cannot evaluate a test without understanding what they are saying. Furthermore, it is our experience that on occasion representatives of test publishers misuse these concepts and reveal their own ignorance regarding their products. We do not believe this type of ignorance is widespread, but it is well to be able to recognize it where it exists.
The objective in selection is to identify and hire those individuals most likely to perform well on the job over a substantial period of time. To discover which selection procedures will accomplish this end, analyses should be carried out relating pre-employment or pre-placement data on candidates for a position to indexes of performance effectiveness. These measures of performance are typ-