Development of a reliable and valid psychosocial measure of self-perception of health in type 2 diabetes.

OBJECTIVES The aim was to develop a reliable and valid psychosocial measure of self-perception of health in type 2 diabetes. METHODS Item analysis and factor analyses were done in order to stabilize and establish principal components of the questionnaire. Reliability (internal consistency aspect) was established using Chronbach's alpha method. Concurrent and discriminant validities were established using conventional methods. RESULTS Factor analysis yielded 12 significant factors (eigen value > or =1), but first six components were retained based on Scree test. These six principal varimax factors explain 49.7% of variance of psychosocial measure of self-perception of health. CONCLUSIONS The final version after all these psychometric procedures had 27 items with six principal components. They were appropriately named as follow: (I) Positive self-feeling; (II) Sociality; (III) Attention seeking; (IV) Feel healthy; (V) Worry about health; (VI) Dependence. Total variance explained is 49.7%.