The generalizability and construct validity of the modern sexism scale: Some cautionary notes

Our purposes are to explore the generalizability of the Modem Sexism Scale [J. T. Swim et al. (1995) "Sexual and Racism: Old-Fashioned and Modern Prejudices," Journal of Personality and Social Psychology, Vol. 68, pp. 199-214] beyond mostly white, traditionally aged undergraduates and to extend analyses of its construct validity. Fourteen African American and 32 white women firefighters nationwide, having previously completed an extensive mail survey, were followed up with the Modem Sexism Scale. A four-factor solution, low inter-item correlations, and a low coefficient alpha argue against combining the 8-items of the scale. Racial differences on one item and in response variances as well as correlations with age, years of service, and education caution against handily overgeneralizing scale usage beyond students. Coherent patterns of interrelationships among Modern Sexism items and other personality, attitudinal, work experiential, and gender identity scales suggest strong convergent validity.