Re: In multinomial logit regression; NaNs produced, an unexpected sign of one of the coefficients
Posted: 15 Dec 2021, 17:09
Hi
the positive sign simply seems to reflect what is happening in your data. Disregarding the four highest cost levels (where you only have 6 observations), what you clearly see is that the probability of choosing bus increases with the cost.
Stephane
the positive sign simply seems to reflect what is happening in your data. Disregarding the four highest cost levels (where you only have 6 observations), what you clearly see is that the probability of choosing bus increases with the cost.
Stephane