binaryregressionmodels二分类变量模型-sites@duke.pdfVIP

Binary Regression Models （（二分二分类变类变量量模型模型）） （（二分二分类类变变量量模型模型）） 2.1 Binary Outcomes – binary dependent variables have two values, typically coded as 0 for a negative outcome (i.e., the event did not occur) and 1 as a positive outcome (i.e., the event did occur) 2.2 Binary Regression Models: explore how each independent variable affects the probability of the event occurring – binary logit model （二分胜算对数模型） – binary probit model （二分概率单元模型） 2.3 The Statistical Model 2.3.1 A Latent Variable Model （隐性变量模型）：assume a latent or unobserved variable y * ranging from -∞ to ∞ -∞ ∞ τ = 0 y * 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 1 y – For example （女性参与就业市场） • y is observed in two states: a woman is in the labor force she is not in the labor force • The idea of a latent y* is that: There is an underlying propensity that generates the observed event We cannot directly observe y* , but at some point a change in y* results in a change in what we observe – As the number of young children in the family increases, a woman’s propensity to “work” would decrease. At some point the propensity crosses a threshold that results in a decision to leave the labor force 9 – Structural model y * = x β + ε i i i – Measurement model: the link between the observed binary y and the latent y * is made with a simple measurement equation: 1 if y * 0  i y i =  * 0 if y ≤ 0  i •