They firstly used compliance as an explanatory variable in the clinical trials . however , compliance often ca n ' t be measured exactly . following the simple model put forward by them , we develop a new model that allow compliance with random errors and separate the effectiveness of treatment from that of the compliance 然而,依從性往往很難被精確測得,因此基于他們提出的簡單模型,我們提出一個(gè)新的模型允許其中的依從性變量帶有量測誤差,并把治療( treatment )的效應(yīng)和依從的效應(yīng)分離出來,從而得到一個(gè)更符合實(shí)際的對照臨床試驗(yàn)數(shù)據(jù)的ev模型。