Skip to main content
. 2012 Oct 1;2(10):1257–1268. doi: 10.1534/g3.112.003285

Table 4. Multivariate analysis of variance for shape changes in IRCSs.

Strain Mandible
df Num df Den Wilks F P
5A 14 49 0.292 6.23 <0.0001
6A 14 45 0.174 10.81 <0.0001
6C 14 51 0.208 10.3 <0.0001
49A 14 39 0.057 30.3 <0.0001
66H 14 42 0.052 37.74 <0.0001
103E 14 34 0.14 8.92 0.0006
119H 14 47 0.153 13.39 <0.0001
120C 14 45 0.02 89.49 <0.0001
122C 14 47 0.111 19.35 <0.0001
122D 14 39 0.23 6.19 <0.0001
122F 14 34 0.067 20.71 <0.0001
135B 14 45 0.159 12.05 <0.0001
135E 14 48 0.409 3.67 0.0008
137E 14 49 0.131 17.01 <0.0001
137F 14 42 0.232 6.83 <0.0001
137G 14 46 0.092 21.83 <0.0001
157F 14 49 0.242 8.05 <0.0001
157D 14 34 0.05 28.15 <0.0001

MANOVA tests for differences in shape by testing within group (i.e. an IRCS strain with C57BL/6). Differences in shape are calculated using a multivariate regression in which the independent variable is the strain, and the dependent variables are 14 first principal components of the tangent coordinates in the shape space. P values are corrected with a Bonferroni adjustment.