Hello everybody,
I have just one question regarding control variables. I am running a two-way fixed effects model, based on time and entity fixed effects (where entities are firms), and I have applied it using data from six different countries. Basically the model uses three different control variables, that were included in the analysis because, at least from a theoretical point of view, it is reasonable to do so. However these variables are not significant (p-value > 10%) in every single case: in some countries they are, in others they're not. Therefore, do I have to eliminate them for those countries in which they are not significant?
I have just one question regarding control variables. I am running a two-way fixed effects model, based on time and entity fixed effects (where entities are firms), and I have applied it using data from six different countries. Basically the model uses three different control variables, that were included in the analysis because, at least from a theoretical point of view, it is reasonable to do so. However these variables are not significant (p-value > 10%) in every single case: in some countries they are, in others they're not. Therefore, do I have to eliminate them for those countries in which they are not significant?