Aggregation or aggravation? The relative merits of a broad withdrawal construct

✍ Scribed by Gary Johns

• Publisher: John Wiley and Sons
• Year: 1998
• Tongue: English
• Weight: 135 KB
• Volume: 19
• Category: Article
• ISSN: 0894-3796
• DOI: 10.1002/(sici)1099-1379(199809)19:5<453::aid-job2899>3.0.co;2-9

✦ Synopsis

Composite variables, then, are so common in psychological measurement that we need to examine their nature and speci®c characteristics . . . In some instances we can obtain a better prediction of scores on one variable from a composite of scores on a series of predictor tests than we can from scores on any one of the predictor tests (Ghiselli, Campbell and Zedeck, 1981, p. 155).

The claim that an index has validity as a measure of some construct carries a considerable burden of proof. There is little reason to believe and much empirical reason to disbelieve the contention that some arbitrarily weighted function of two [or more] variables will properly de®ne a construct. More often, the pro®table strategy is to use the variables separately in the analysis to allow for complex relationships (Cronbach, Gleser, Nanda and Rajaratnam, 1972, p. 339).