Table 2.
Linear mixed-effects models for each experiment with patch reward rate at leaving time as the dependent variable
| PE | T statistic | F statistic | dfa | p | |
|---|---|---|---|---|---|
| Study 1 | |||||
| Intercept | 2.4 | 34 | 1128 | 38.9 | <0.00001 |
| Patch type (P) | −0.05 | −4.0 | 16 | 74.6 | 0.0002 |
| Environment (E) | −0.1 | −6.3 | 40 | 37.5 | <0.00001 |
| P × E | −0.008 | −1.1 | 1.1 | 1927 | 0.3 |
| Study 2 | |||||
| Patch type (P) | -0.03 | −2.3 | 5.1 | 57.3 | 0.03 |
| Environment (E) | −0.06 | −4.1 | 16.9 | 28.1 | 0.0003 |
| Dopamine (D) | −0.03 | −1.37 | 1.9 | 29.3 | 0.18 |
| E × D | −0.01 | −2.29 | 5.22 | 200.1 | 0.023 |
| P × E | <−0.001 | −0.1 | 0.01 | 197.1 | 0.92 |
| P × D | −0.006 | −1.14 | 1.3 | 186.6 | 0.26 |
| P × E × D | −0.003 | −0.56 | 0.31 | 186.4 | 0.58 |
adf was calculated using the Satterthwaite correction method.