Table 1 Summed log-likelihoods and AIC values across participants for the full model (with free parameters \(\beta _d\), \(\beta _s\), and \(\lambda\)) and reduced models, from lowest to highest AIC.

From: Quantifying tracking quality during occlusion with an integrated gaze metric anchored to task performance

Model

Log-likelihood

AIC

Remove \(\lambda\)

\(-377.24\)

790.48

Full model

\(-369.88\)

793.76

Remove \(\beta _s\)

\(-383.29\)

802.57

Remove \(\beta _s\), \(\lambda\)

\(-417.80\)

853.60

Remove \(\beta _d\)

\(-456.57\)

949.15

Remove \(\beta _d\), \(\lambda\)

\(-481.78\)

981.56

Remove \(\beta _d\), \(\beta _s\)

\(-511.19\)

1040.38

Back to article page