| Summary: | 碩士 === 義守大學 === 管理學院碩士班 === 97 === In Taiwan, subject to the policy of Bureau of National Health Insurance, doctors are mainly the decision makers of prescription medicine. They have hence been targets of most research focusing on decisive factors of using prescription medicine and marketing strategies of medication. Nonetheless, once patients decide to purchase self-paid medicine, what will be the factors they consider and are those the same with the doctor’s? The purpose of this thesis is to analyze the factors of self-paid medicine targeting on the purchase of Glucosamine through the statistical methodology of DEA-DA (Data Envelopment Analysis—Discriminant Analysis) by which a predictive model is set up.
There are 664 valid questionnaires collected from nine different hospitals and clinics and randomly divided into 2 groups. One is examined with the DEA-DA method to gain data of main factors of the purchase of self-paid medicine; while the other is employed to analyze and prove the first model. In addition, the differences among demographics of the factors of self-purchasing will be analyzed
According to the survey, there are six decisive factors which affect self-paid medicine purchasers. The order of the discriminant parameter is as following: doctor’s recommendations, relative and friend’s recommendations, the media, advertisements, health care and medication efficacy. The results are fairly different from those of doctor’s consideration based on the medicine’s safety, efficacy, convenience, economy and the pharmacist’s service and reliability.
According to the survey, responders who are military, civil and teaching staff or service providers aged below sixty with a bachelor’s degree or with monthly income over NT$30,000 are mainly affected by doctor’s recommendations and the relative and friend’s recommendations. Those who are below 60 with higher education and higher income are mostly affected by the media and advertisements. Responders aged over 60 put more emphasis on medication efficacy.
Utilizing the statistical method of DEA-DA (Data Envelopment Analysis— Discriminant Analysis) enables the research to build up a predictive model with higher hit rate. In the future, this research model will be worth applying on the medication domain.
|
|---|
