Table 8A.
Stepwise discriminant function analysis for left metatarsal bone.
| Functions | Wilk’s lambda | Unstandardized coefficients c | Structure matrix d | Standardized coefficients e | Group centroids f | Sectioning point g | |||
|---|---|---|---|---|---|---|---|---|---|
| Wilk’s lambda a | Chi-square | sig b | Male | Female | |||||
| Function 1 Measurements of the 1st metatarsal bone | |||||||||
| SA: V | 0.283 | 69.969 | 0.000 | -72.779 | -0.686 | -1.098 | 1.536 | -1.589 | -0.027 |
| Bone density | 13.129 | 0.144 | 0.745 | ||||||
| Height | 64.683 | 0.278 | 0.504 | ||||||
| (constant) | -13.556 | ||||||||
| Function 2 Measurements of the 2nd metatarsal bone | |||||||||
| SA: V | 0.380 | 55.225 | 0.000 | 39.926 | 0.946 | 0.974 | -1.257 | 1.257 | 0.000 |
| Height | -35.974 | -0.242 | -0.325 | ||||||
| (constant) | -8.664 | ||||||||
| Function 3 Measurements of the 3rd metatarsal bone | |||||||||
| SA: V | 0.241 | 81.189 | 0.000 | 51.104 | 0.947 | 0.984 | -1.746 | 1.746 | 0.000 |
| Height | -38.990 | -0.210 | -0.324 | ||||||
| (constant) | -13.755 | ||||||||
| Function 4 Measurements of the 4th metatarsal bone | |||||||||
| SA: V | 0.301 | 69.087 | 0.000 | 50.593 | 1.000 | 1.000 | -1.499 | 1.499 | 0.000 |
| (constant) | -20.593 | ||||||||
| Function 5 Measurements of the 5th metatarsal bone | |||||||||
| SA: V | 0.345 | 60.676 | 0.000 | 51.174 | 0.948 | 1.024 | -1.355 | 1.355 | 0.000 |
| Bone density | -4.971 | -0.090 | -0.327 | ||||||
| (constant) | -10.718 | ||||||||
aAt each step, the variable that minimizes the overall Wilks’ lambda is entered. Minimum partial F to enter is 3.84. Maximum partial F to remove is 2.71.
bp value is 0.000, which means the significant level at p < 0.001.
cUnstandardized canonical discriminant functions evaluated at group means. Take Function 1 for example, Y = 13.129 * bone density + 64.683 * height - 72.779 * SA: V - 13.556.
dStructure matrix indicates the pooled within-group correlations between discriminating variables and standardized canonical discriminant functions.
eStandardized coefficients represent the contribution of the variable to sex discrimination.
fUnstandardized canonical discriminant functions evaluated at group means.
gWhen the group mean of male is positive, discriminant score (Y) > sectioning point would be considered as male; while the group mean of male is negative, discriminant score (Y) < sectioning point would be considered as male.