3.2 Logistic model

Examine the bivariate relationship first among variables to high protein status of milk. Check for quasi or complete separation of categories. Shout out to Lisa Avery for this function!

plotuv(data=Milk, response='High_Protein', covs=c('Cow', 'Yard', 'Diet', 'Time', 'protein'));

Below logistic model does not account for clustering of repeat observations by cow to high protein status of milk.

rm_mvsum(model=glm(High_Protein ~ Diet + Yard + Time, data=Milk, family='binomial'), showN=T, vif=T);

	OR(95%CI)	p-value	N	Event	VIF
Diet			1337	386	1.02
barley	Reference		425	183
barley+lupins	0.45 (0.34, 0.60)	<0.001	459	119
lupins	0.29 (0.21, 0.40)	<0.001	453	84
Yard			1337	386	1.01
1	Reference		251	66
2	1.29 (0.86, 1.92)	0.21	276	76
3	1.71 (1.16, 2.53)	0.007	272	92
4	1.31 (0.89, 1.95)	0.17	268	87
5	1.08 (0.72, 1.63)	0.70	270	65
Time	0.98 (0.95, 1.00)	0.044	1337	386	1.00