Table 2. Effects of Drug/Placebo Manipulation on the Different Behavioral Measures.
| Behavioral measure | Drug: main effect | Weight: main effect | Drug–weight interaction |
|---|---|---|---|
| Average gambling percentage | F(1,30)=1.66, p=0.208 | F(1,30)=0.42, p=0.520 | F(1,30)=4.96, p=0.034* |
| Gambling slope | F(1,30)=0.00, p=0.960 | F(1,30)=0.00, p=0.943 | F(1,30)=1.34, p=0.256 |
| Corrected context effect | F(1,30)=1.76, p=0.195 | F(1,30)=0.27, p=0.604 | F(1,30)=0.03, p=0.857 |
| Absolute gambling slope | F(1,30)=1.41, p=0.240 | F(1,30)=0.88, p=0.357 | F(1,30)=0.09, p=0.767 |
| Baseline choice precision | F(1,30)=0.54, p=0.390 | F(1,30)=0.00, p=0.948 | F(1,30)=2.30, p=0.140 |
| Gambling bias parameter μ | F(1,30)=1.00, p=0.325 | F(1,30)=0.47, p=0.830 | F(1,30)=1.40, p=0.246 |
| Value function parameter α | F(1,30)=0.15, p=0.698 | F(1,30)=0.241, p=0.627 | F(1,30)=0.756, p=0.391 |
To account for body weight, participants were assigned to low/high-weight groups (based on a median split) and mixed-effect ANOVAs on the behavioral measures were run with drug as within-subjects factor and weight grouping as between-subjects factor. Results of these analyses are reported here. The interaction effect on average gambling percentage alone is significant and is marked with an asterisk.