Generally, it can be concluded that the items were rather too difficult for the sample, as indicated by a lower proportion of correct answers (see Fig. 1).
Accordingly, DIF methods designed to help monitor fairness in CAT must function in diverse testing environments and, often, when there are large number of items in the bank but the number of individual examinees who respond to each one of these items is rather small.
The theoretically predicted result that approximately 6 additional three-option items might be needed to compensate for the reliability reduction from four-option items is rather inconclusive.
Subsequently, we excluded 17 less relevant items: 12 items were rather identical to other items in the same questionnaire, and 5 items were related more to aspects of parental child rearing than to child behaviour.
For example, in the present study, over 80% of the participants denied that others disliked them, therefore the distribution of this item was rather skewed.
Since the response rate of No.14 item was rather low in Chinese population [ 13], we chose item A8 from NEI-VFQ39 to serve as the appendix of No.14 item according to the instruction of the NEI-VFQ25 manual.
First, the measure of bullying based on a single-item is rather limiting but other studies have also had a similar problem.
Alternating buttock pain is often suggested as a possible contributor 15 since, although not very sensitive, this item is rather specific and is now shown in a primary care setting.
Item difficulty was rather low, ranging between.16 and.40.
Of the 39 items, we excluded another nine items that are rather directly related to the lung (cough, asthmatic breathing); the heart (fearfulness); the spleen (loose stools, bowel incontinence); cold invasion (stiffness in the low back or knees); excess heat (constipation, high libido); excessive water intake (clear urine).
