Author:
Michael Karson - University of Denver
Author
Note:
Correspondence concerning this article should be addressed to Michael
Karson, University of Denver, Graduate School of Professional Psychology,
2450 S. Vine St., Denver, CO 80208.
E-mail: mkarson@du.edu.
Abstract:
The WAIS–III manual (Wechsler, 1997) cautions psychologists against interpreting differences between an individual’s
scores on two scales unless the difference is great enough to justify confidence that it does not merely reflect
measurement error. Similar considerations should apply to the Minnesota Multiphasic Personality Inventory–2
(MMPI–2) when scale combinations are interpreted or when code types are identified, because the identification of
a code type is really a way of saying that the scales in the code type are higher than the other scales in the profile.
They should also apply to the Rorschach when response counts on two variables are compared with each other. Key
Rorschach and MMPI–2 inferences made routinely in clinical practice may be based on measurement error rather
than on the subject’s true scores.
