Covariate effects given by coeff_fe() - on the logit of the transition probability?

[RonEfrat]

Hi,

Sorry if I missed it, but I looked all over and couldn't find an answer (maybe because it is obvious?). When calculating the effects of covariates on transition probabilities, are the estimates given by coeff_fe() or confint() referring to their effects on the logit of the transition probability, or on the transition probability itself?

Thank you,
Ron

[TheoMichelot]

Both coeff_fe() and confint() refer to the coefficients of the linear predictor, e.g., intercepts and slopes associated with covariates. You can get the transition probabilities with par() or predict(). This is discussed in Section 7.4 and 8.1 of the main vignette.

Let me know if anything is unclear.