WTP calculation for MNL model
Posted: 04 May 2022, 03:04
Dear Concern,
I want to calculate the WTP from the MNL model. This MNL model is the base model. Going through the literature, I found out in calculating the WTP, which requires the ratio of the estimated coefficients.
I have looked into the example of MNL model with SP covariates. I have not included any SP covariates, the base model only contains the attributes of the alternatives which shows that they are highly significant.
If I use the delta method for calculating the WTP, I just simply used the following script:
deltaMethod_settings=list(expression=c(WTP="-b_cpeed/b_cost",
WTP_Charging_time_min="-1*(b_cpeed/b_cost)","asc_3"))
apollo_deltaMethod(model_WTP, deltaMethod_settings)
Here, bcpeed= Charging Time or Speed (which is a positive estimate), b_cost is the Charging cost (which is a negative estimate)
I have multiplied with -1 term, in order to have a positive value. The delta method settings require a 3rd argument, which was not fixed in estimation.
I just simply used asc_3.
I would like to know, is this an appropriate approach or I am leading to some fatal error?
I want to calculate the WTP from the MNL model. This MNL model is the base model. Going through the literature, I found out in calculating the WTP, which requires the ratio of the estimated coefficients.
I have looked into the example of MNL model with SP covariates. I have not included any SP covariates, the base model only contains the attributes of the alternatives which shows that they are highly significant.
If I use the delta method for calculating the WTP, I just simply used the following script:
deltaMethod_settings=list(expression=c(WTP="-b_cpeed/b_cost",
WTP_Charging_time_min="-1*(b_cpeed/b_cost)","asc_3"))
apollo_deltaMethod(model_WTP, deltaMethod_settings)
Here, bcpeed= Charging Time or Speed (which is a positive estimate), b_cost is the Charging cost (which is a negative estimate)
I have multiplied with -1 term, in order to have a positive value. The delta method settings require a 3rd argument, which was not fixed in estimation.
I just simply used asc_3.
I would like to know, is this an appropriate approach or I am leading to some fatal error?