Premium
ITEM GENERATION PROCEDURES AND BACKGROUND DATA SCALES: IMPLICATIONS FOR CONSTRUCT AND CRITERION‐RELATED VALIDITY
Author(s) -
MUMFORD MICHAEL D.,
COSTANZA DAVID P.,
CONNELLY MARY SHANE,
JOHNSON JULIE F.
Publication year - 1996
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.1744-6570.1996.tb01804.x
Subject(s) - construct (python library) , construct validity , psychology , content validity , reliability (semiconductor) , set (abstract data type) , variety (cybernetics) , criterion validity , test validity , validity , scale (ratio) , psychometrics , social psychology , statistics , computer science , developmental psychology , mathematics , power (physics) , physics , quantum mechanics , programming language
Background data measures are one of the best predictors of job performance. Nonetheless, questions have been raised about their content and construct validity. The present effort describes a set of procedures for developing construct and content valid background data items. Data gathered in seven field studies and six laboratory studies are presented bearing on the reliability and validity of the measures constructed using these item generation procedures. Findings in these studies indicate that construct‐based item generation procedures yield reliable scales evidencing both content and construct validity. Furthermore, these scales are capable of predicting performance in a variety of settings. Theoretical and practical implications of these findings for the development and validation of background data measures are discussed.