Abstract
Coaching effects on the psychometric properties of tests. In this research the effect of two different coaching programs on the psychometric properties of tests was studied. The first program was designed to coach tests in which academic know ledge is the construct assessed. In this program the training was focused on the academic contents to be assessed. The second program is focused in the strat egies and test wiseness used by the subjects in order to answer multiple choice items. A sample of 152 university students was used for the first program, and a sample of 120 workers from the public administration was used for the second program. Test reliability is not affected by the use of coaching programs. Diffe rential Item Functioning was not detected after the application of the coaching programs. However, slightly differences in const ructvalidity (assessed through factor analysis) we re observed as a consequence of the implementation of coaching programs. The oretical and practical implications of the results obtained are discussed.
