Many people wonder whether a sophisticated and intelligent person has a better chance of succeeding on an integrity test. Their assumption is that it may be easy for such people to know what is expected of them on the test, what the seemingly "correct answers" are, and thus more easily fake their responses.

In fact, the professional literature does show a link between the level of intelligence and achievements on some integrity tests.  Nevertheless, at Midot;s integrity test has been designed to neutralize such effects.  Specifically, in calculating test scores, only partial weight is given to the candidate's direct answers.  Instead, scores are primarily determined by analyzing measurements that are hidden from the candidate being tested.

Accordingly, it was interesting to examine whether, despite the hidden measurements, intelligent candidates would still manage to purposely improve their scores.

Our basic hypothesis at the beginning of the study was that a reasonable relationship exists between a person's years of education and his level of intelligence.  Accordingly, we set out to test whether more educated individuals receive higher integrity scores.

If it turns out that they do, it would mean that the Midot test in fact is affected by the (intelligent) subjects’ ability to direct their answers in a way that is best for them.  If it turns out that they don't, then the analytic methods used by Midot will have appeared to be unaffected by intelligence.

The study’s sample consisted of 1,026 subjects. Their level of education varied between just 5 years (lower education) and 20 years (higher education) with an average of 14 years.

The study examined the relationship between years of education and each of the four integrity scores: the tendency to report the truth ("Report"); the tendency to steal ("Theft"); the tendency to respect employment norms ("Norms"); and the tendency to accept bribes ("Bribery"). 

It was found that, in three of the four parameters evaluated, the relationship was near zero and insignificant.  A very low, close to zero, and insignificant relationship (0.08) was found for Norms, in favor of the educated subjects.  That is, in the majority of the parameters, including the Norms parameter, there was no statistically significant relationship between the subjects’ level of education and their test scores.
Thus, highly educated subjects were no more successful than others on these tests.

In order to support the findings, a comparision was made between the average years of education of all subjects who received a "Recommended" final evaluation and those who received a final evaluation of "Not Recommended".  No differences were found between these two groups.