I want to calculate risk aversion coefficients using Constant Partial Risk Aversion utility function (U=(1-a)X1-a). But I am not sure on how to go about it. Anyone with this knowledge?
Did you use hypothetical or real game? The real game may be better as it avoids the hypothetical bias. I have attched relevant articles that estimate CPRA. I think besides understanding the theory (conceptual model) it would be important the way we interpret (categorize) the coefficient of risk aversion from the software output. I hope it may help.