Dear STATA-Users,
it is my first post and I hope I am complying with all the forum-rules.
Background: I am analysing the farmer's adoption decision of sustainable practices using a discrete choice model.
I have farm census data available for the years 2010, 2016 and 2020.
In the first step, I filter out all farms that already adopt sustainable practices in 2010 and run the logit models to analyze the adoption decision between the 1st time period (2010 - 2016) and 2nd time period (2016 - 2020).
EXAMPLE: the y variable will be 1, if the farm has answered yes wether it adopted a sustainable pracitce in 2016 and 0 otherwise. The covariates will be the strucutral characteristics and external factors of the farm in 2010 (demographic characteristics of the farmer in 2010, farm size in 2010, etc.)
Now, I am wondering, if the choice to adopt sustainable practices in 2016 influence also the decision in 2016, wether to adopt it or not during the period between 2016 - 2020. To make it clearer, I attach a screenshot with the nested strucutre I have in mind.
Now, I am trying to run a nested logit model, considering the first decision between 2010-16 as the "first nest" (taking the observed variables in 2010 as the individual specific variables) and the second decision between 2016-2020 as the "second nest" (taking variables in 2016 as the individual specific variables).
There is no alternative specific variable here.
I am new to nested logit model and would like to ask if such an approach could be feasible for this research question and the available data or if I should rather try with sequential logit models or any other approach.
The big concerns are: (1) I have no alternative specific variable; (2) The decision is not taken simultaneously; (3) The variables that I include are the same for the two nests, only in different time points (ex. farm size in 2010 for the first nest and farm size in 2016 for the second nest).
I hope you can help me claryfing my doubts.
Thank you in advance for any comment.
Best
Davide
it is my first post and I hope I am complying with all the forum-rules.
Background: I am analysing the farmer's adoption decision of sustainable practices using a discrete choice model.
I have farm census data available for the years 2010, 2016 and 2020.
In the first step, I filter out all farms that already adopt sustainable practices in 2010 and run the logit models to analyze the adoption decision between the 1st time period (2010 - 2016) and 2nd time period (2016 - 2020).
EXAMPLE: the y variable will be 1, if the farm has answered yes wether it adopted a sustainable pracitce in 2016 and 0 otherwise. The covariates will be the strucutral characteristics and external factors of the farm in 2010 (demographic characteristics of the farmer in 2010, farm size in 2010, etc.)
Now, I am wondering, if the choice to adopt sustainable practices in 2016 influence also the decision in 2016, wether to adopt it or not during the period between 2016 - 2020. To make it clearer, I attach a screenshot with the nested strucutre I have in mind.
Now, I am trying to run a nested logit model, considering the first decision between 2010-16 as the "first nest" (taking the observed variables in 2010 as the individual specific variables) and the second decision between 2016-2020 as the "second nest" (taking variables in 2016 as the individual specific variables).
There is no alternative specific variable here.
I am new to nested logit model and would like to ask if such an approach could be feasible for this research question and the available data or if I should rather try with sequential logit models or any other approach.
The big concerns are: (1) I have no alternative specific variable; (2) The decision is not taken simultaneously; (3) The variables that I include are the same for the two nests, only in different time points (ex. farm size in 2010 for the first nest and farm size in 2016 for the second nest).
I hope you can help me claryfing my doubts.
Thank you in advance for any comment.
Best
Davide
