Blackwell Reference Online

Extract

Synthetic validity is “the inferring of validity in a specific situation from a logical analysis of jobs into their elements, and a combination of those elemental validities into a whole” ( Balma, 1959: 395 ). The process of synthetic validation is related to validity generalization , since the validity of a test battery is inferred rather than demonstrated for a specific job and setting. This inference is based on the development of a matrix of (1) job elements common to a family of jobs ( see job family ) and (2) selection tests that are predictive of job performance on those individual job elements. A basic assumption of synthetic validation is that different jobs involving the same kinds of behavior should also require the same knowledge, skills, abilities, and other characteristics ( see ksaos ). Synthetic validity subsequently assumes that if a test is valid for a particular job element, then it will be valid for use with any job involving that same element ( McCormick, DeNisi, and Shaw, 1979: 51 ). Using the concept of synthetic validity, a test battery can be assembled for a particular job by identifying the elements involved, and then selecting tests that have previously been shown to predict performance on those elements. This approach presumably eliminates the need to validate the new test battery, since each of the elemental validities is already known and should ... log in or subscribe to read full text