I am having trouble with checking for the assumptions for my models (which i already built, i thought a histogram was sufficient), and i was wondering if you could help me see if i am doing it right.
I need to check to association between independent variables (meandaily...times) and the dependent variables (either calcium or ketosis). In the lme's i am correcting for farm and / or parity (random factors) as well, but i am also doing GLM without.
original dataset layout
and to do the analysis, i widened the dataset
And now i am struggling with how to check for the assumptions. Ive learnt that either plot(modelname) or qqplot(residuals(modelname)) should work, but i am struggling to know which of the models?
Should it be the final model of calcium or ketosis? or should it be a univeriable? and then which one? I really cannot figure this out with all of my willpower. I have made different test models, as shown below. Can someone help me with this? Am i doing something right here? And then, which one?
As statistician nor english is one of my native capacities / languages, i am sometimes really struggeling with explanations with a lot of abbreviations..
If someone could help me point in the right direction, i would VERY much appreciate.
Thanks in advance!