Can I get "fail to reject the null" when testing with test statistic, and "reject the null" when testing with P-value at the same time in a same nonlinear regression model?
Asked
Active
Viewed 17 times
0
-
If so, how to conclude this? – David May 14 '21 at 03:21
-
1Hi: That's not possible because whether you use the t-statistic or the pvalue, the conclusion has to be the same. This is because the test statistic and the pvalue are a one to one mapping. You would need to provide more details in order for someone to explain what is being done incorrectly. Also, keep in mind that non-linear regression doesn't have the same kind testing properties as OLS unless asymptopia holds so there could be something going on there. – mlofton May 14 '21 at 03:35