Multiple Choice
The influence of a variable x_(i) on the the binary dependent variable D_(i) should be
investigated. For this purpose the linear probability model D_(i)=eta _(0)+eta _(1)x_(i)+epsi _(i) is
estimated. In addition, the relationship is also estimated via a logit model, which for
x_(i) has the estimated coefficient hat(alpha )_(1). For hat(alpha )_(1) and hat(eta )_(1) then holds, ...
a) dotshat(eta )_(1) and hat(alpha )_(1) measure respectively the marginal effect of an increase of x_(i) by one
unit on the probability that D_(i)=1.
b) dots if hat(alpha )_(1)<0, then the probability of D_(i)=1 decreases with increasing values of
x_(i).
c) dotshat(alpha )_(1)
d) dotshat(alpha )_(1)
2. Multiple Choice The influence of a variable X; on the the binary dependent variable D; should be
estimated. In addition, the relationship is also estimated via a logit model, which for X; has the estimated coefficient &. For and i then holds, : :
a) .. . 1 and a1 measure respectively the marginal effect of an increase of X, by one unit on the probability that D; = 1.
b) ... if a < 0, then the probability of D; = 1 decreases with increasing values of Xi. c)...a1 ~ 31. d)...a1+31=1.