Psychometric and substantive issues in scale construction and validation

Abstract

Presents a procedure for assessing the success of a scale in measuring an underlying construct. This procedure (a) provides a quantitative index of the relation between a scale and its underlying factor, (b) provides guidance in interpreting reliability coefficients, (c) indicates 1 case where the multitrait-multimethod procedure may lead researchers astray, and (d) aids development of scales suggested by clearly articulated theory. The 5 Job Descriptive Index scales were examined with this procedure in Exp I with 1,046 retail employees. Each scale was found to relate highly with a single underlying construct. In contrast, results for the Job Characteristic Inventory scales completed by 851 National Guard members in Exp II are less compelling and suggest that 3 of these scales (task identity, task variety, and dealing with others) may benefit from further refinement. (25 ref) (PsycInfo Database Record (c) 2025 APA, all rights reserved)