An Approach to the Validation of the Szondi Test through a Systematic Study of Unreliability
Research work with projective techniques is teaching us more sophisticated and basically valid conceptions of reliability. In the earlier days of psychological testing, we thought that a test was reliable if it (or its component parts, or equivalent forms) always gave us nearly the same scores. As soon as one starts dealing with more variable features of personality than total intelligence, using tools which are sensitive enough for the purpose, he discovers that it is unreasonable to require a test always to give the same results. Such a requirement would make sense only if we could properly conceive of personality as a static, fixed entity, instead of the dynamic (that is, moving and changing) process that it is. Would a pilot in the Rockies want a “reliable” altimeter that always gave the same reading? Certainly not. In such a case the instrument’s validity would depend on its giving constantly changing values, as long as the plane was in motion. Of course, when the actual distance between plane and ground was constant, the reading should be also. A barometric altimeter, which varies with humidity as well as altitude, would then be less reliable than one which worked by radar.
KeywordsSexual Experience Projective Technique Psychological Meaning Resistance Item Total Choice
Unable to display preview. Download preview PDF.