Objectives: Most sleep scales assess sleep quantity (e.g., sleep duration and daytime sleepiness) or sleep quality (e.g., sleep latency and maintenance); the Pittsburgh Sleep Quality Index (PSQI) is an exceptional example. However, the prevalence of 24-hour operations presents the need for a scale that can also measure sleep phase (e.g., sleep onset and offset). Furthermore, we have to assess the phase, quality and quantity respectively to understand which of them has a problem. Thus, the 3 Dimensional Sleep Scale (3DSS) - day workers version - was developed to assess each of them related to sleep, and this study attempted to verify its reliability and validity.
Methods: Subjects were 635 day workers (461 men, 174 women; average age = 40.5 years) from the manufacturing and service industries. A scale was created based on a pre-study and discussions with specialists. The scale consisted of 17 sleep-related items. The skew of the data was assessed, and the construct validity and reliability were verified using exploratory and confirmatory factor analysis and Cronbach's alpha, respectively. The scale was scored and G-P analysis was performed. The items measuring phase, quality, and quantity of sleep were selected from the PSQI and SDS, and their correlation with the three scales of 3DSS were measured to verify the convergent and discriminant validity. In addition, the total scores obtained on the PSQI were compared with each scale of the 3DSS.
Results: No skew was found in the data. Exploratory factor analysis revealed a three-factor structure--quality, quantity, and phase. Each factor consisted of five items, therefore two items were excluded. The fitness of the 15-item model was better than that of the 17-item model according to confirmatory factor analysis. Cronbach's alpha for phase, quality and quantity score were 0.685, 0.768 and 0.716, respectively. The hypothesis tests were almost accepted, therefore convergent and discriminant validity were sufficiently established.
Conclusions: The present study established the reliability and validity of the 3DSS; however, further studies using larger samples are needed to standardize the test and to establish a cut-off value.