In binary discrete regression models like logit or probit the omis-sion of a relevant regressor (even if it is orthogonal) depresses the re-maining b coefficients towards zero. For the probit model, Wooldridge(2002) has shown that this bias does not carry over to the effect ofthe regressor on...

In binary discrete regression models like logit or probit the omis-sion of a relevant regressor (even if it is orthogonal) depresses the re-maining b coefficients towards zero. For the probit model, Wooldridge(2002) has shown that this bias does not carry over to the effect ofthe regressor on...

In binary discrete regression models like logit or probit the omis-sion of a relevant regressor (even if it is orthogonal) depresses the re-maining b coefficients towards zero. For the probit model, Wooldridge(2002) has shown that this bias does not carry over to the effect ofthe regressor on...

In binary discrete regression models like logit or probit the omis- sion of a relevant regressor (even if it is orthogonal) depresses the re- maining <font face="Symbol">b</font> coefficients towards zero. For the probit model, Wooldridge (2002) has shown that this bias does not carry over to the effect of the regressor...