The reliability and validity of short form health belief model scale for stroke patients

Objective To simplify the Chinese version of the champion health belief model scale (CHBMS). Methods The Chinese version of CHBMS was revised using item analysis and expert consultation methods. A total of 235 stroke patients were recruited and investigated with short form health belief model scale (SF-HBMS). SPSS 17.0 and Mplus 7.0 software were used to test the reliability and validity. Results The item-level content validity index(I-CVI)was 0.97. The Sale-CVI/ universal agreement (S-CVI/UA) was 0.85. The Sale-CVI/average (S-CVI/Ave) was 0.97. The Cronbach's α was 0.835. The split-half reliability was 0.779. The test-retest reliability was 0.811. The confirmatory factor analysis was performed to test the construct validity. The χ²/ df was 1.642. The comparative fit index (CFI) was 0.939. The tucker-lewis index (TLI) was 0.926. The standardized root mean square residual (SRMR) was 0.062; the root mean square error of approximation(RMSEA)was 0.052, P=0.361, which all demonstrated that six dimensions and their items were all consistent and fitted with its theoretical structure. The criterion-related validity was 0.956. It could distinguish the difference of health belief among patients with different educational level (t=-2.282，P=0.023) and had desired discriminative validity. Conclusions The SF-HBMS has satisfactory reliability and validity, and can be used to assess the level of health belief for stroke patients.