在醫學與其他研究領域中，常會出現成對的連續與離散資料，如一個人的血壓值與癌症的嚴重程度。由於這種相關性資料不容易找到合適的模型，因此在分析上較為困難。 本文主要的目的是在廣義線性模型下，利用強韌概似函數方法，來分析成對的混合型資料。我們建立參數之強韌概似函數，在不需要特別對於成對的反應變數間之相關性建立模型的假設下，仍可得到正確的統計推論。 ;Pairs of continuous and discrete data, such as a person′s blood pressure and the severity of cancer often appear in medicine and other fields of research. In general, likelihood inference for correlated data of mixed types is strenuous due to the scarcity of a suitable statistical model. In this thesis, we show that the robust likelihood method is suitable for paired mixed data. More specifically, we consider testing the treatment/covariate effects on paired responses. Simulations and real data analysis are used to demonstrate the merit of this parametric robust approach.
