. 2011 Aug 11;27(20):2851–2858. doi: 10.1093/bioinformatics/btr472

Table 1.

Three hypothesis tests for homogeneity, heterogeneity and total strength of an interaction across two experimental conditions

Test on an interaction across two conditions	Null hypothesis	Alternative hypothesis	Test statistic	Significance
Heterogeneity test	β⁽¹⁾_d=β⁽²⁾_d and β₀⁽¹⁾=β₀⁽²⁾	β⁽¹⁾_d≠β⁽²⁾_d or β₀⁽¹⁾≠β₀⁽²⁾	F_d	p_d
Homogeneity test	β_c=0	β_c≠0	F_c	p_c
Total strength test	β⁽¹⁾_d=β⁽²⁾_d=0 and β₀⁽¹⁾=β₀⁽²⁾	β⁽¹⁾_d≠β⁽²⁾_d or β₀⁽¹⁾≠β₀⁽²⁾	F_t	p_t