I would like to ask about my research thesis....
I am doing a data panel research on the impact of bank ratios (LDR, NPL, efficiency, ROA, bank size) and macroeconomic (GDP and interest rate) on the proportion of MSME credit to total credit by banks in Indonesia in 2015-2019...
My dependent variable (Y) is proportion of MSME credit to total credit
and my independent variable (X) is LDR (t-1), NPL(t-1), efficiency(t-1), ROA(t-1), bank size, GDP(t-1), and interest rate
So to choose between CEM and FEM I used the Chow test, the result is FEM
then to choose between the FEM and REM I used the Hausman test, the p-value result is 1... with invalid cross-section notes like in the image...
*updates: I found out that the Hausman test of 1 is the result from the GDP and interest rate variables, because I used the same GDP and interest rate for each sample the period... (e.g GDP growth in 2014: 5%, so I input 5% on GDP for each sample banks for the year of 2014)
So I tried to exclude the GDP and interest rates in the model, after that it gave my the p-value hausman of 0.14>0.05 which is a normal value for hausman test (not 1 and without any invalid notes) so choose REM...
but What does this means if the hausman p-value = 1 when I use the GDP and interest rate in the model, should I use REM or FEM or what should I do??
I've done my research and there are 2 contradictions for this problems:
1. use FEM: viewtopic.php?t=1924
2. REM is more efficient: viewtopic.php?t=2294
Thanks in advance, and sorry for my bad English..
For econometric discussions not necessarily related to EViews.
1 post • Page 1 of 1
Who is online
Users browsing this forum: No registered users and 13 guests