Reliability and validity of SF-36 in health survey among empty-nest elders in countryside

Objective To access the reliability and validity of 36-Item Short Form Health Survey(SF-36)in the evaluation of quality of life among the empty-nest elders in countryside. Methods Totally 218 empty-nest elders were selected with stratified cluster random sampling for a face-to-face interview with SF-36.The reliability of SF-36 was evaluated by split-half reliability and internal consistency reliability and its validity was evaluated by discriminant validity, convergent validity,and structure validity. Results The split-half reliability was 0.84 and the internal consistency Cronbach'α coefficient was 0.89.The Cronbach's α coefficient for role-physical,physical functioning,bodily pain, general health,vitality,social funtioning,role-emotional,and mental health was 0.79,0.88,0.86,0.81,0.77,0.76,0.88, and 0.77,resectively.The convergent validity was 97.14% and discriminant validity was 98.21%.By confirmatory factor analysis,we got a model,of which the structure was consistent with the theoretical assumption,and two main statistics of root mean square error of approximation(0.12)and confirmatory fit index(0.91). Conclusion The results indicate that the SF-36 is a useful scale for the evaluation of the quality of life among the empty-nest elders in countryside.