Personal Relevance and the Choice of Constructs for the Repertory Grid Technique

Abstract

A number of studies have shown with psychological testing that the form a particular response takes will depend on the personal relevance to the subject of the stimulus, and that the more personally relevant the stimulus the less the effects of other response determinants. For example, Dahlstrom (1962) has listed six “contextual” and five “mediating” variables which he feels account for responses to such test items as those composing the M.M.P.I. These variables include the experimental setting, the examiner and the specific test items (contextual variables), and the veridical facts, personality styles, test instructions (mediating variables). Similar arguments have been advanced with regard to projective testing by Hutt (1951, 1954) and Wertheimer (1957), among others. Caine (1967) has shown that response suppression operates in a sentence building test only when the test material is of limited personal significance to the subject, and that the consistency of response between different testing levels along the overt/covert dimension pertains only when the stimulus material is of psycho-pathological significance. With the T.A.T., Smail (1966) has argued that if meaningful results are to be obtained responses must be interpreted in the light of the patient's particular situation much more than is the general practice, and that responses which are clinically typical of a given diagnostic group may only appear where the stimulus card accurately reflects the situation in which the patient finds himself.