Many decision analysts have studied assessment methods of utility function. Decision maker``s utility function is obtained by an interview process. However, assessing preferences is difficult since human decision makers are often uncertain of their preferences or his/her response is not consistent frequently.
Thus, this study introduces a framework of systematically investigating inconsistencies and biases in assessing decision maker``s utility function and develops a new procedure to reduce these inconsistencies and biases by requestioning and to estimate unknown parameters for utility function.