Table 5.

Behavioral choices did not depend on the outcome of the previous trial. On average there were 67 trials per game, which were equally likely to be a win, loss, or draw. By chance subjects should have switched to a different movement (non-repeat) two thirds of the time and stayed (repeated) one third of the time. Given 22 trials in each outcome category (win, lose, or draw) chance would dictate switching 15 trials and staying 7 trials. The mean number of trials in each group was compared to the expected number of trials. A t-test showed that the actual number of trials did not differ significantly from the expected number of trials in any of the categories.

Trial Type	Mean # of Trials (Standard Error)	T value	P value
Win stay repeats	7.7 (0.5)	0.51	0.61
Win switch non-repeats	14.5 (0.6)	−0.42	0.67
Lose stay repeats	7.1 (0.5)	−0.52	0.6
Lose switch non-repeats	14.9 (0.7)	0.37	0.7
Draw stay repeats	6.7 (0.5)	−1.79	0.08
Draw switch non-repeats	15.9 (0.6)	1.31	0.2