A Study on Factors Affect the Individual Health Insurance Claim

• Main Authors: Tzu-Hsien Huang, 黃紫��
• Other Authors: Li-Hsien Chien
• Format: Others
• Language: zh-TW
• Online Access: http://ndltd.ncl.edu.tw/handle/76360558344215681943

碩士 === 國立中興大學 === 應用經濟學系所 === 97 === This study used the underwriting insured customers from some life insurance company from 2001 to 2004 as the research objects. The Binary Logistic Regression model with different critical values was adopted to measure and compare the prediction power of claim for each stage. Those factors influenced the claim of the personal health insurance were analyzed and tested. Meanwhile the Poisson Regression model was considered to probe into the important factors which affected the degree of risk for the specific insurance policy. Major statistical results were concluded as follows： 1.Three factors - insured amount, classification of premium paying, and job classification, were significantly negatively correlated with the claim. The premium, age of insured, and physique were the significantly positively correlated with the claim. 2.Under the different critical values, the Binary Logistic Regression model would attach to the type Ⅰ and type Ⅱ error at some degree. Under the minimum tolerance with type Ⅰ error, the critical condition with value at 0.13 presented the best performance of the claim to 2004 on 2001. 3.Term of paying premium, insured amount, the premium, age of insured, relation of applicant for insurance and insured, and business unit were significantly positively correlated with the degree of risk on the insurance policy. And classification of premium paying, physique, job classification, and the paying item of claim presented the opposite effection were the negative correlation and important influence for the degree of risk on the insurance policy.