Hello Everyone, I am currently conducting my research about the determinants of financial distress of a set of companies. I gathered data from 20 companies over the period of three years from 2020 to 2022. My dependent variable - financial distress is binary in nature (appoint 1 if the company is financially distress and appoint 0 vice versa) and my independent variables are return on assets, financial leverage and firm liquidity. After carefully reading articles in this area, I found that most of them are suggesting to use panel data logistic regression if the dependent variable is binary in nature (I consider ). Since i am lack of knowledge regarding logistic model as a beginner, I am doubtful regarding the following areas.
1. What are the regression assumption associated under this model and how to test them in STATA? are the heteroscedasticity, auto correlation and cross section dependence valid assumptios? how to test them?
2. How to test the suitability of the random effect model? i run the xttest0 command but i did not worked
2. What are the solutions upon the violation of these assumption?
If you can provide guidance on this it would be immensely appreciated.
Thank you regards, Suren
1. What are the regression assumption associated under this model and how to test them in STATA? are the heteroscedasticity, auto correlation and cross section dependence valid assumptios? how to test them?
2. How to test the suitability of the random effect model? i run the xttest0 command but i did not worked
2. What are the solutions upon the violation of these assumption?
If you can provide guidance on this it would be immensely appreciated.
Thank you regards, Suren
