probit model

[hussein bataineh]

Hi

please im following a paper where they use the (simple probit model) with a control for year and industry there sample is the American firms from 2002 -2018

I'm not sure what code should i need to use probit or xtprobit .they wrote in there paper we use (simple probit model) with a control for year and industry

.the paper menthion ( We estimate a simple probit model; .Industry and year fixed effects are included clustered at the firm level in model ) . as i tried xtprobit give us either (random effect or fixed effect) which is not the target. Can use use probit code with control .Industry and year fixed effects.

probit hard_final_Exact_new csopresence1 Firm_Size_w ROA_w Leverage_w Market_book_four_w Non_pension_CFO_w STD_CFO_w Board_Independence_w BoardSize_w Gender_Diversity_w Fund_Status_w FUNDING_RATIO_w Platn_Size_w CSR_Committee SustainabilityScore_w i.ff_12 if jon==0 , robust cluster (id)

Could you advice me what is the right approach

[Maxence Morlet]

I would by default resort to a pooled probit, but in your case the fixed effect would cause the incidental parameters problem. The logistic model has a sufficient statistic for these, so you could go xtlogit, fe to solve that problem. Why not a linear probability model? What is the mean of the dependent variable?