Table 1.
Multiple regression models examining the effect of the mean and variability of each microstructural parameter on EFEM, VMCV and VMExp/σapp. The estimate for each term in the model is given with the associated p-value in parenthesis.
| Response | Term | BV/TV(λ) | Tb.Th (mm)(λ) | Tb.N(mm−1)(λ) | Tb.Sp (mm)(λ) |
|---|---|---|---|---|---|
|
| |||||
| (EFEM(MPa))(λ) | Av | 2675 (<.0001) | 8606 (<.0001) | 326.9 (<.0001) A | −474.0 (<.0001) |
| SD | −3148(<.0001) | −15983 (<.0001) | NS (p>0.23) | NS (p>0.39) | |
| Int | 1374 (<.0001) | 901.9 (<.0001) | 206.0 (<.006) | 242.5 (<.0001) | |
| Model | r2 adj | 0.92 | 0.51 | 0.78 | 0.82 |
| p< | 0.0001 | 0.0001 | 0.0001 | 0.0001 | |
|
| |||||
| (VMCV (%))λ | Av | −1.969 (<.0001) | −5.883 (.0001) | −0.2142 (<.0001) | 0.1526 (<.04) |
| SD | 7.124 (<.0001) | 20.17 (<.0001) | 0.7951 (=0.0192) | 1.222 (=0.0106) | |
| Int | −0.6918 (<.0001) | −0.3309 (<.0001) | NS (p>0.39) | NS (p>0.22) | |
| Model | r2 adj | 0.64 | 0.42 | 0.50 | 0.53 |
| p< | 0.0001 | 0.0001 | 0.0001 | 0 0001 | |
|
| |||||
| (VMExp/σapp)(λ) | Av | −74.95 (<.0001) | −229.8 (<.0001) | −9.672 (<.0001) | 15.79 (<.0001) |
| SD | 31.64 (<.01) | 343.2 (<.006) | NS (p>0.72) | NS (p>0.79) | |
| Int | 2.634 (<.0001) | 17.26 (<.0001) | 36.39 (<.0001) | 36.54 (<.0001) | |
| Model | r2 adj | 0.98 | 0.46 | 0.86 | 0.90 |
| p< | 0.0001 | 0.0001 | 0.0001 | 0.0001 | |
Int: Intercept
An adjustment for multiple tests, taking into account the correlations between all parameters, suggested that a p-value of 0.0170 be considered as significant for an α of 0.05.