coefficent is significant,but marginal effect is not significant

[Pengfei Chu]

When I used meologit to finish my regression, the coefficient of the variables was significant. But after I calculated the marginal effect , I found that the marginal effect is not significant, why is this?

[Clyde Schechter]

It would be better to show the actual command and output you got from -meologit- and the actual commands and output for calculating the marginal effect. One possibility is that you made a mistake somewhere. Another possibility is that you are taking a difference between p = 0.04 and p = 0.06 seriously--when it is in fact meaningless.

But even apart from those things, -meologit- is a nonlinear model. Which means that there is no such thing as the marginal effect of anything. marginal effects in non-linear models are always conditional on specific values of all variables in the model. (Or, people often calculate the average marginal effect, which looks at marginal effects at all observed values of the model variables and averages those results.) Depending on the particular marginal effect which was calculated, the results could be large or could be very close to zero.