using a variance of residuals scatter plot and a Kolmogorov–Smirnov
test with Lilliefors correction.
4. Results and discussion
4.1. The prediction formulas of the regression models
4.1.1. Prediction formula for the OGP equipment
The linear regression analysis was conducted on the 145 sample carcasses,
using back fat thickness (X1) and muscle thickness (X2) determined
with the OGP equipment as predictors and the dissected lean
meat percentage in the carcass as dependent variable (Y). Results
show that both independent observations, back fat thickness and muscle
thickness, were statistically significant predictors of the lean meat
percentage in the pig carcass (for back fat thickness: t = −16.929,
P b 0.001 and for muscle thickness: t = 6.629, P b 0.001). They explain
a significant amount of variance in the value of lean meat percentage
(F(2142) = 274.084, P b 0.05, R2 = 0.794, R2
Adjusted =
0.791). The RMSEP was computed by applying the leave-one-out
full cross-validation technique, on all 145 carcasses. The RMSEP