I have constructed a variable later to convert it into growth variable (as the variable is equal to percentage change in that variable) I have to take log difference . My question is my data is unbalanced panel data
1.log difference will provide same interpretation here ?
2 What if i take simple log of that variable as after implying difference most of the coefficients are insignificant and resulted signs are against the predicted one .
3. is there any other way to create a growth variable?

Regards