소비자 지식의 측정: IRT 방법의 응용

Consumer knowledges play important roles in consumer choice and decision making processes. It is important to measure consumer knowledge accurately. There are two types of consumer knowledges: subjective knowledge and objective knowledge. Subjective knowledge is measured as the degree of confidence on themselves. On the other hand, objective knowledge is often measured as the number of choices consumer made or prior knowledges. But those measure does not guarantee the accuracy of knowledge. The better way is to use test items. Using test items, we can identify the abilities of consumers as well as the difficulty and discrimination of test items. Item response theory using Rasch model and 2-parameter model can handle and produce those results. In the paper, we use simple data and show how to use IRT using consumer knowledge aspects and brand aspects with housing knowledge. The model selection between Rasch model and 2PL model is done by likelihood ratio test, AIC, and BIC. Then we identify difficulties and discrimination capabilities of items. The estimation of abilities of consumers for the response patterns for several items are made. The application is very similar to recommendation system using item-based and individual based. Also, the application could be extended to the areas of consumer satisfaction, consumer choice, and brand equity.