تأیید تحلیل عاملی شایستگی بررسی سیستم های حفاظتی هیئت مدیره از آزار جنسی در آمریکا : برازش یک مدل سه عاملی

Abstract Research over the past decade and a half exploring the psychological dimensions of the definition of sexual harassment has converged on a three-factor model: gender harassment, unwanted sexual attention, and sexual coercion. The current study used confirmatory factor analysis to determine whether the most widely used sexual harassment survey, the U.S. Merit Systems Protection Board (USMSPB; 1981) survey, measures this three-factor structure. Data from a sample of the USMSPB's 1988 survey of Federal employees (n= 1070) and from a midwestern university's survey of undergraduates (n= 572) and graduate students, faculty, and staff (n= 575) were analyzed. The three-factor model fit reasonably well in all samples, but two survey items, measuring sexual assault and sexual favors, were weakly related to their respective factors. This model, however, was only marginally stable across female and male subsamples, and gender harassment and unwanted sexual attention had low discriminant validity. Both practical and theoretical implications are discussed. ☆ We thank John Hetherington for his analytic assistance and review, and Jim Gruber and the SASH network for their last-minute rescue. ☆☆ Address correspondence and reprint requests to Margaret S. Stockdale, Department of Psychology, Southern Illinois University at Carbondale, Carbondale, IL 62901-6502. Copyright © 1997 Academic Press. All rights reserved.