Background: A better understanding of the workplace difficulties experienced by people with multiple sclerosis (PwMS) may be critical to developing appropriate vocational and rehabilitative programs. Objective: We aimed to assess the factor structure, internal consistency and validity of the new Multiple Sclerosis Work Difficulties Questionnaire (MSWDQ). Methods: Work difficulty items were developed and reviewed by a panel of experts. Using the MSWDQ, cross-sectional self-report data of work difficulties were obtained in addition to employment status and MS disease information, in a community-based sample of 189 PwMS. Results: Exploratory Maximum Likelihood Factor Analysis on the draft questionnaire yielded 50 items measuring 12 factors. Subscale internal consistencies ranged from 0.74 to 0.92, indicating adequate to excellent internal consistency reliability. The MSWDQ explained 40% of the variance in reduced work hours since diagnosis, 40% of the variance in expectations about withdrawing from work, 34% of the variance in expectations about reducing work hours, and 39% of the variance in expectations about changing type of work due to MS. Conclusion: The MSWDQ is a valid and internally reliable measure of workplace difficulties in PwMS. Physical difficulties, as well as cognitive and psychological difficulties were important predictors of workplace outcomes and expectations about future employment.

This study assessed the factor structure, internal consistency, and concurrent validity of the Smoking Expectancies Scale for Adolescents (SESA) using 717 Australian adolescents (87% nonsmokers, 11% current smokers, and 2% ex-smokers). Exploratory factor analysis of SESA yielded 8 factors. Confirmatory factor analysis indicated that the 8-factor model, and also a 2nd-order cost– benefit model, fit the data significantly better than 4 alternatives. Validation analyses revealed the 8-factor model explained 26% to 32% of the variance in adolescent cigarette use, smoking intentions, smoking subjective norms, and peer smoking. The 2nd-order model explained 12% to 17% of the variance in these same variables.