In many disciplines of psychological assessment (educational psychology, clinical psychology, health psychology, and others) has been seized from the observation that the substitution of items on psychological tests, serious action will have on the scale values. However, these effects have hardly been investigated systematically, and attempts a theoretical explanation are only rare, although, for example, theories of personality psychology appropriate deductions allow.
In my dissertation, I conducted a preliminary study that the interchange of simulation-oriented item (s) with the identical construct-oriented items (K) was investigated. The derived from theories of personality psychology expectation was that the correlations between S and K is higher if the items are applied in the order in SK, as was the order of K p. This expectation is met with a significance level of 0.05.
This result can be measured theoretical conclusions, which taken together are here:
- The exchange of items or psychological tests have systematic effects on the the scale values
- changed the psychological state of the respondents in dealing with a psychological test / Items
- A psychological measurement theory, the non-interchangeability of tests / items considered
also practical conclusions are possible. Thus, with a suitable mathematical formalism shows that reliabilities are just perfect when the tests have no impact on the psychological state of the respondents (they must fulfill a so-called eigenvalue equation). Another observation concerns the test battery or Assessment Center: The order in which the appropriate tests administered be, can systematically affect the results obtained.
