ReliabilityInternal consistency and item–total correlation analysis Item means, standard deviations, item–total correlations and Cronbach’s alpha coefficients, if deleted for each item, are presented in Table 3. In order to identify poorly functioning items, the following criteria were used: an increase of more than 10 in total scale reliability when that item was deleted, or a correlation of <0Æ30 between an item and the total subscale score (Champion 1999, Go ¨zu ¨m & Aksayan 2003). All items met the criteria for inclusion, except one from the Benefits and Health Motivation Subscale (Benefit 18). This item – ‘If I have a smear test regularly and the result is good, I don’t need to worry too much about cervical cancer’ – was deleted because of low correlation between the item and subscale scores (r =0 Æ17). The reliability measure for theBenefits Subscale and Health Motivation Subscale was repeated twice to test the difference in reliability when the one item was deleted. Cronbach’s alpha coefficient increased from 0Æ82 to 0Æ86 after removing this item from the Benefits and Health Motivation Subscale. Cronbach’s alpha coefficients for the five subscales ranged from 0Æ62 to 0Æ86 (Table 4).Stability The stability of the measures was examined by test–retest reliability with 93 women. The interval between the first and second tests was 3–4 weeks. Test–retest reliability intra-class coefficients ranged from 0Æ79 to 0Æ88 (Table 4).
การแปล กรุณารอสักครู่..
