Objective: To evaluate the reliability and validity of the Chinese version of Pittsburgh sleep quality index (PSQI) among the medical students.
Methods: A total of 603 undergraduate students were selected from a medical school in Beijing using a stratified-cluster random-sampling strategy, who were assessed by using the Chinese version of PSQI. Then 171 subjects were randomly selected to retest about 1 month later, using a variety of psychometric properties to evaluate reliability and validity of the scale.
Results: The average of global PSQI scores was 5.38±2.34, the overall Cronbach's α coefficient of all the scale items was 0.734, the half-split reliability coefficient was 0.655; the test-retest correlation coefficient was 0.530 for global PSQI score, and there was no statistically significant difference before and after (P>0.05). As for the construct validity, confirmatory factor analysis results indicated that the two factor model (containing sleep efficiency factor and sleep quality factor) was the optimal model (goodness-of-fit index 0.949, root of the mean square residual 0.083, comparative fit index 0.699), and the factor loading for "hypnotic drug use" was the lowest. Using subjective sleep quality of the students as the criteria, the global PSQI score was statistically different between the criteria-based groups (P<0.001). And the area under the receiver operating characteristics (ROC) curve which was plotted using the global score of PSQI was 0.908 (95%CI: 0.880-0.936).
Conclusion: The Chinese version of PSQI has been verified, having good reliability and validity among the medical students, and is an effective tool to investigate and screen the sleep quality of medical students in China.