Abstract
This report describes the development and refinement of a set of scales for use in research on predictors of colorectal cancer screening adherence. The study population included 2693 of 4490 eligible white male automotive employees who answered a mailed questionnaire (60% response rate) on beliefs and attitudes related to colorectal cancer and screening. Exploratory and confirmatory factor analyses and multitrait scaling analysis were used to evaluate the construct validity of a priori scales developed to measure salience and coherence, perceived susceptibility, worries about screening, screening efficacy, social influence, and intention. Analyses supported the construct validity of scales for salience and coherence, perceived susceptibility, and worries about screening. Four items originally assigned to the salience and coherence construct loaded on a separate factor that appeared to measure self-efficacy. There was no empirical support for scales measuring screening efficacy and social influence, and there was limited empirical support for a scale measuring intention. Confirmatory factor analysis of the scales measuring salience and coherence, self-efficacy, perceived susceptibility, and worries about screening showed a similar factor structure in white men with and without a personal history of polyps, indicating that the scales may be useful for studies of both colorectal cancer screening and surveillance. Multitrait scaling analysis showed some support for internal consistency reliability of those scales in women (n = 42) and in African-American men (n = 56), and there was some support for the factor structure in those two subgroups. Future studies should evaluate the psychometric properties of these and similar scales in diverse population subgroups.
Volume 6, Issue 10
