A psychometric evaluation and validation of the Preferences Scale.

The Preferences Scale (PS) is a new measure of morningness and therefore requires intensive investigation to establish its measurement properties and validity. A confirmatory factor analysis (CFA) of Smith and colleagues (2002) posited model structure was not optimal in both a student (n = 731) and working sample (n = 218). However, significant differences (p < 0.01) were found for times to sleep and wake between morning and evening types. Principal components and reliability analyses were used to develop a 6-item model comprising 2 factors using a student sample (n = 368). Cronbach's alpha for the PS was 0.73 and the factors explained 61% of the variance. The revised model was replicated via CFA in a separate student sample (n = 363). A subsequent CFA confirmed the model structure in the working sample. Cronbach's alpha was 0.74 and the factors explained 64% of the variance. Significant differences (p < 0.01) in self-reported alertness ratings between morning and evening types were obtained by time-of-day. These results provide preliminary support for the PS. The predictive efficacy of the PS requires further validation against a number of health and work-related variables.