Premium
EMPIRICALLY DERIVED JOB FAMILIES AS A FOUNDATION FOR THE STUDY OF VALIDITY GENERALIZATION
Author(s) -
COLBERT G. A.,
TAYLOR L. R.
Publication year - 1978
Publication title -
personnel psychology
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 6.076
H-Index - 142
eISSN - 1744-6570
pISSN - 0031-5826
DOI - 10.1111/j.1744-6570.1978.tb00452.x
Subject(s) - psychology , generalization , social psychology , regression analysis , job performance , predictive validity , job analysis , regression , econometrics , statistics , job satisfaction , mathematics , clinical psychology , psychoanalysis , mathematical analysis
Studies I and II of this series (Taylor, 1978; Taylor and Colbert,1978) were concerned with the empirical development of homogeneous families of insurance company jobs based on data from the Position Analysis Questionnaire (PAQ). The current study, Study III, involved studies of validity generalization within the job families which resulted from the previous research. In general, it was hypothesized and found that regression equations yielded statistically significant cross‐validated multiple correlations within job families and that different predictors were found valid for different job families. Cross‐validated multiple correlation coefficients appeared to be somewhat larger when validation samples were drawn from the same job family than for cases in which validation samples were drawn from different job families. Absolute errors of prediction were found to be greater for regression equations applied across job families than for regression equations applied within job families. Implications of the job analysis and validity generalization studies are discussed.