Hello This is my first post here. I'm running a mixed logit model for transport mode choice with 4 variables in Stata 17 with cmmixlogit. The coefficient for travel time variable is defined random. As part of estimation result the LR test vs fixed parameters is reported. The prob>=chibar2 is 0.4442. How do we interpret this prob? Can we reject randomness and accept fixed parameters? I read "sg160: On boundary-value likelihood-ratio tests" from stata technical bulletin which discusses the distribution of the test statistic but couldn't understand. What's the rule of thumb for it's interpretation? Can anyone help? LR test vs. fixed parameters: chibar2(01) = 0.02 Prob >= chibar2 = 0.4442
-
Login or Register
- Log in with
