TOPIC:

The Use of the Center for Epidemiologic Studies Depression (CES-D) Scale in the Chinese Population: A Rasch Analysis

​

BIOGRAPHY:

Dr. Zhu, Jinxin is a post-doctoral fellow at the Assessment Research Centre at The Education University of Hong Kong. He has authored and co-authored a number of conference and peer-reviewed articles, and book chapters. Jinxin has severed as a lecturer for workshops on advanced research methods and a guest co-editor for Educational Psychology and Asia Pacific Journal of Educational Development. His major areas of research interest are self-regulated learning, parenting, and educational and psychological measurement.

​

ABSTRACT:

The Center for Epidemiologic Studies Depression (CES-D) Scale (Radloff, 1977) is widely used to screen depression among the general population. However, few studies examined the gender- and age-related Differential Item Functioning (DIF) of the scale, especially in a Chinese context. Using Winsteps® (Linacre, 2017), a partial credit Rasch model was conducted to examine the psychometric properties, as well as the age and gender DIF, of the CES-D scale on a large sample (N = 34,762; 50.6% females, 48.4% males) of Chinese general population covering a wide range of age between 10 and 99 years. Results showed a second dimension (positive-affect) against depression, one gender DIF item ("Cry") favoring females, and three age DIF items: the "Fear" item and the "Cry" item favoring younger respondents while the "Sleep" item favoring older respondents. The findings show that positive-affect might not be a good (negative) indicator of depression, at least for the Chinese population. Also, given the same level of depression, females cry more frequently compared to males and older people felt fearful and cried less frequently but had more sleep problem than younger people in China. As a result, the CES-D scale might need to be revised for better assessment of depression. Specifically, positive-affect items might need to be excluded from the scale and DIF items should also be excluded or estimated separately across different gender and age groups for the Chinese population.