This paper focuses on econometric issues, especially the common assumption that monetary payoff is subjects' actual utility, in estimating subjects' learning behaviors using experimental data. I propose a generalized adaptive learning model that nests commonly used learning rules. First, I show...