Skip to main content
. 2016 Jan 8;29(3):645–656. doi: 10.1111/jeb.12814

Table 1.

Cross‐sex tests for (A) heterogeneity of phenotypic variance (V P) and (B) G×S interactions. Also presented are comparisons of the full G×S model to restricted scenarios where (C) mf is freely estimated but V A assumed homogeneous and (D) V A(F) are allowed to differ V A(M) but mf is constrained to unity. For each comparison, null (H0) and alternate (H1) hypotheses are shown with statistical inference from likelihood ratio tests

Trait Comparison H0 H1 χ2 d.f. P
AM (A) Homogeneous V P Heterogeneous V P 3.80 1 0.051
(B) No G×S G×S 5.90 2 0.052
(C) G×S r MF = +1, V A assumed homogeneous 0.90 1 0.343
(D) G×S V A heterogeneous, r MF = +1 assumed 1.70 1 0.427
WTM (A) Homogeneous V P Heterogeneous V P 11.9 1 <0.001
(B) No G×S G×S 12.5 2 0.002
(C) G×S r MF = +1, V A assumed homogeneous 0.04 1 0.842
(D) G×S V A heterogeneous, r MF = +1 assumed 7.10 1 0.029
SLM (A) Homogeneous V P Heterogeneous V P 0.222 1 0.638
(B) No G×S G×S 14.9 2 <0.001
(C) G×S r MF = +1, V A assumed homogeneous 0.122 1 0.727
(D) G×S V A heterogeneous, r MF = +1 assumed 9.07 1 0.011