Repeated measures design for general linear mixed models testing the effect of region of shaft for antlers (antler burr or below top branching or palm) and femur (middle or upper part of shaft) on mechanical variables derived from 3-point bending tests
| Bone | Hydration | Property | R2a | AICb | Intercept | Section (burr)c |
| Antler | Wet | E | 18.7% | 84.7 | 6.43±0.49 | 1.44±0.42** |
| BS | 24.8% | 195.9 | 103.2±5.9 | 20.4±4.8** | ||
| W | 16.6% | 157.2 | 28.2±2.4 | 5.2±2.5† | ||
| Dry | E | 25.4% | 107.0 | 16.22±0.67 | 2.26±0.84* | |
| BS | 12.6% | 248 | 336±13 | 30±16† | ||
| W | 2.5% | 147 | 24.3±12.4 | - | ||
| Femur | Wet | E | 0.0% | 219 | 22.34±0.48 | - |
| BS | 0.0% | 474 | 263±9 | - | ||
| W | 0.0% | 107.0 | 9.45±0.45 | - |
| Bone | Hydration | Property | R2a | AICb | Intercept | Section (burr)c |
| Antler | Wet | E | 18.7% | 84.7 | 6.43±0.49 | 1.44±0.42** |
| BS | 24.8% | 195.9 | 103.2±5.9 | 20.4±4.8** | ||
| W | 16.6% | 157.2 | 28.2±2.4 | 5.2±2.5† | ||
| Dry | E | 25.4% | 107.0 | 16.22±0.67 | 2.26±0.84* | |
| BS | 12.6% | 248 | 336±13 | 30±16† | ||
| W | 2.5% | 147 | 24.3±12.4 | - | ||
| Femur | Wet | E | 0.0% | 219 | 22.34±0.48 | - |
| BS | 0.0% | 474 | 263±9 | - | ||
| W | 0.0% | 107.0 | 9.45±0.45 | - |
Data for Intercept and Section are shown as mean±s.e.m.
General linear mixed models are based on maximum likelihood and not least squares. Therefore, they do not allow calculation of the variability explained or R2. However, we have run the same final model shown here in a GLM to obtain a rough idea of R2 for comparison between models.
AIC stands for Akaike information criterion. It serves for comparison between alternative models and is better the lower its value.
The coefficient shows the increase in the mechanical property for the burr compared with the section in the top below the palm of the antler. In the case of femora, the comparison was between the mid-shaft and upper part of the femur.