Initial validation of the Yin-Yang Assessment Questionnaire for persons with diabetes mellitus

To initially test for the content validity, comprehensibility, test-retest reliability and internal consistency reliability of the Yin-Yang Assessment Questionnaire (YY-AQ).

The process of initial validity and reliability test covered: (1) content validation from the findings of 18 multiple-case studies, validated Yin- and Yang-deficiency assessment questionnaires, relevant literatures and registered Chinese medicine practitioners; (2) comprehension with the levels of comprehensibility for each item categorized on a 3-point scale (not comprehensible; moderately comprehensible; highly comprehensible). A minimum of three respondents selecting for each item of moderately or highly comprehensible were regarded as comprehensive; (3) test-retest reliability conducted with a 2-wk interval. The intraclass correlation coefficients (ICCs) and their 95%CIs were calculated using a two-way random effects model. Wilcoxon Signed Rank test for related samples was adopted to compare the medians of test-retest scores. An ICC value of 0.85 or higher together with P > 0.05, was considered acceptable; and (4) internal consistency of the total items was measured and evaluated by Cronbach's coefficient alpha (α). A Cronbach's α of 0.7 or higher was considered to represent good internal consistency.

Eighteen Yin-deficiency and 14 Yang-deficiency presentation items were finalized from content validation. Five participants with type 2 diabetes mellitus (T2DM) performed the comprehensibility and test-retest reliability tests. Comprehensibility score level of each presentation item was found to be moderate or high in three out of the five participants. Test-retest reliability showed that the single measure ICC of the total Yin-deficiency presentation items was 0.99 (95%CI: 0.89-0.99) and the median scores on the first and 14(th) days were 17 (IQR 6.5-27) and 21 (IQR 6-29) (P = 0.144) respectively. The single measure ICC of the total Yang-deficiency presentation items was 0.88 (95%CI: 0.79-0.99) and the median scores on the first and 14(th) days were 10 (IQR 6-18) and 14 (IQR 7-23) (P = 0.144) respectively. The results of a descriptive correlation study on 140 survey participants with T2DM using the YY-AQ showed that internal consistency of the total Yin-deficiency and Yang-deficiency presentation items was satisfactory, with Cronbach's α of 0.79 and 0.78 respectively.

The YY-AQ will be tested further for comprehensibility, test-retest and internal consistency reliabilities, scoring system validity, construct validity, convergent and discriminant validities, responsiveness and predictive validity.