Knowledge evaluation about diabetes mellitus is necessary to self-manage diabetes effectively. We developed two parallel diabetes knowledge tests to meet a need for reliable knowledge assessment in diabetic patients. Materials and methods: The 75-items (59 items for general knowledge test, 16 items for insulin use subscale) were administered to 102 diabetic patients who visited Inha University Hospital. The items which had the appropriate difficulty (0.25~0.80) and good discrimination index

more »

... ve 0.25) were selected. However, the items which are thought to be an important item for education were also selected even though they did not meet the criteria of reliability and discrimination index. Two parallel diabetes knowledge tests were developed after matching the selected appropriate items for similar contents. Results: 102 patients fulfilled the tests and their mean age was 54.1 ± 11.5 years. Mean percentage of correct questionnaires was 60.9 ± 12.5% for general test and 45.9 ± 19.5% for insulin use subscale. There were significant differences of scores between patients with high and low education level, between patients with high income per household and low income level per household, between patients with the history of diabetes education and without history of diabetes education, and between the old (≥ 50 yrs) and the young (< 50 yrs) age group. However, there was no significant difference of scores according to diabetes duration and complication or not. The selected two tests had a similar score. And their Cronbach alpha was appropriate (> 0.70) in both tests. Conclusions: We developed two parallel diabetes knowledge tests. These tests can be used as an important means in evaluating the diabetes knowledge and effect of education in diabetic patients. (J Kor Diabetes Assoc 30:476~486, 2006) ━━━━━━━━━━━━━━━━━━