the model as random factor (Z, the know matrix and u and ε the unknown
vector of the random effect) since there was a high significant
correlation (P b 0.05) with pH30LT but there was no further interest in
the effect of this parameter.
The same statistical procedure was followed to infer one final linear
mixed model for the pH30LT (Y), with all measured variables during the
pre- and post-slaughtering process as fixed factors (X) and the day of
measurement as random factor (Z). Only variables quantified to the
5% level of significance (P b 0.05) were considered in the final linear
mixed model. If strong associations (|r| N 0.6) between covariates
were present, the most adequate and representative one was selected
in the final linear mixed model.