Hello, i'm new here and still have a difficulties to do regression
i'm doing panel fem estimation and one of my independent data type is index with range -2.5 to 2.5 (but all of my data is negative value). the result of this variable giving a positive coefficient which is different from the expected (should be negative based on the literature).
then i try to change the negative data to become a positive (with removing the negative sign, just want to look what if the result if my data is positive) and the coefficient result shown its match with the expected (negative).
my question is
1. is it possible to something like this is happened? or maybe is there any way to interpret this result?
2. what is the best solution to get result match as the expected? (i try to log transformation but i can't be use since all of the data is negative value)
Thank you in advance
i'm doing panel fem estimation and one of my independent data type is index with range -2.5 to 2.5 (but all of my data is negative value). the result of this variable giving a positive coefficient which is different from the expected (should be negative based on the literature).
then i try to change the negative data to become a positive (with removing the negative sign, just want to look what if the result if my data is positive) and the coefficient result shown its match with the expected (negative).
my question is
1. is it possible to something like this is happened? or maybe is there any way to interpret this result?
2. what is the best solution to get result match as the expected? (i try to log transformation but i can't be use since all of the data is negative value)
Thank you in advance
