Choose, rate or squeeze: Comparison of economic value functions elicited by different behavioral tasks
Fig 1
Task design and behavioral results.
A. From top to bottom, successive screen shots of example trials are shown with their duration for the three tasks (left: rating task, middle: force task, right: choice task). Every trial started with a fixation cross. In the force and rating tasks, a single composite proposition, with a gain G for the subject (YOU) and donation D for the charity organization (ORG) was displayed on the screen. Then a scale (for rating) and a thermometer (for force) respectively appeared on the screen, noticing subjects that it was time for providing a response. After response completion (rating or force), feedback on whether the proposition was won or lost was displayed. The probability of winning was fixated to 70% in the rating task and determined by the percentage of maximal force produced in the force task. A loss meant no money for both the subject and the charity. In the choice task, two composite options were displayed and choice was triggered by switching ‘or’ into ‘?’. Feedback was winning the chosen option in 70% of the trials, and nothing in the remaining 30%. B. Average ratings (left), forces (right) and values inferred from choices (right) are shown as functions of the amount of gain and donation. Cold to hot colors indicate low to high values. The value function used to fit the choices was the a priori function that served to optimize the design (linear model with interaction).