Table 1 Significance of the Linear Mixed Effect model quality improvement for each of the 17 selected features with the addition of the size and the interaction between the size and the feature.
Compared models | |Z| (f1) | |Z| (f2) | |Z| (f3) | |Z| (f4) | |Z| (f5) | \(\phi\) (f1) | \(\phi\) (f2) | \(\phi\) (f2) | \(\phi\) (f4) |
|---|---|---|---|---|---|---|---|---|---|
Feature only vs. main effect | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) |
Main effect vs. interaction | 0.55 | 0.97 | 0.39 | \(\le \,0.05\) | \(\le \,0.05\) | 0.86 | 0.64 | 0.56 | 0.08 |
Compared models | \(\phi\) (f5) | tan \(\delta\) (f1) | tan \(\delta\) (f2) | tan \(\delta\) (f3) | tan \(\delta\) (f4) | tan \(\delta\) (f5) | P\(_y\) | Min \(\phi\) | – |
Feature only vs. main effect | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | \(\le \,0.05\) | – |
Main effect vs. interaction | 0.84 | \(\le \,0.05\) | 0.39 | 0.09 | 1 | 0.79 | 0.15 | 1 | – |
