Table 2. Results of model fitting on mouse behavioral data.
AIC and BIC both give model M1 as winner. All models fit the data significantly better than chance: pR2 score > 0. The table shows mean ± SEM of parameter values without outliers (|β|≥1,000).
| Model | α | β | γ | m0 | LL | AIC | BIC | pR2 |
|---|---|---|---|---|---|---|---|---|
| M1 | 0.66 ± 0.04 | 11.7 ± 6.0 | −212.5 | 611.1 | 2259.0 | 0.08 | ||
| M2 | 0.59 ± 0.04 | 9.4 ± 5.8 | 0.34 ± 0.04 | −211.0 | 701.1 | 2957.4 | 0.09 | |
| M3 | 0.59 ± 0.04 | −0.8 ± 15.2 | 0.62 ± 0.04 | 619 ± 411 | −199.5 | 771.1 | 3615.9 | 0.14 |
| M4 | 0.39 ± 0.03 | −15.2 ± 8.2 | 0.90 ± 0.004 | −209.7 | 698.4 | 2952.1 | 0.09 |
Abbreviations: AIC, Akaike Information Criterion; BIC, Bayesian Information Criterion; LL, log-likelihood; M, model; pR2, pseudo R2