Validating the coping scale for Chinese athletes using multidimensional Rasch analysis

Objective. The Coping Scale for Chinese Athletes (CSCA) was developed and validated using classic testing theory in 2004 (Chung, Si, Lee, & Liu, 2004). This study aimed to validate CSCA using multidimensional Rasch analysis with the ConQuest software programme. Method. The sample in this study comprised 367 athletes from mainland China. A Multidimensional Rating Scale model was applied to investigate the validity of the four-dimension scale. Standard fit statistics (Infit and Outfit MNSQ) and Differential item functioning (DIF) were computed to examine the model-data fit. Test reliability and category functioning were also checked. Results. The item difficulty and the athletes’ trait level of coping were calibrated along the same latent trait scale. Three items were removed from the scale due to misfit with the Rasch model. No DIF across gender was found for the remaining 21 items. Test reliabilities for the four subscales ranged from 0.66 to 0.76. The results also indicated that the orginal 5-category rating scale structure did not function well. Conclusion. The multidimensional Rasch analysis supported that the 21-item CSCA measures four latent traits of coping of Chinese athletes as expected. The results also domstrated advantages of multidimensional Rasch analysis over unidimensional Rasch analysis as well as traditional approach in examing the quality of multidimensional scale in sport settings. Copyright © 2011 Elsevier Ltd. All rights reserved.