Figure 2.

Average number of cards in each block in each stage. In the standard net score [(C + D)−(A + B)] in the clinical IGT version, the tendency to learn good EV is significant in each stage. The slopes of the learning curves increase rapidly in Stages 2 and 3. However, the standard clinical IGT version involves only 100 trials. The presentation of net scores has raised the question of whether normal decision makers truly avoid all of the bad EV cards and favor all of the good EV cards, and the question of why the difference between Block 1 and Block 5 in Stage 1 is only four cards [2−(−2)].