Evaluation of the Impact of Errors in the Measurement of Backfat Depth on the Prediction of Fat-Free Lean Percentage

Abstract

The development of regression equations to predict carcass composition assumes that the independent variables, such as backfat depth (BFD), are measured without error. Monte Carlo simulation was used to evaluate the impact of measurement error for BFD on the prediction of carcass fat-free lean percentage (FFL%) in pigs. In the simulation, FFL% was a linear function of carcass weight and actual BFD (ABFD). Measurement errors were generated such that the correlations (rBF) of the measured BFD and ABFD ranged from 0.70 to 0.95. Two types of measurement errors were simulated: 1) errors with constant variance, and 2) errors whose SD were proportional to the ABFD. A total of 1,000 replications of 1,000 pigs were simulated. The absolute values of the regression coefficients and R2 values of the equations decreased as r BF decreased. When measurement errors were proportional to the ABFD, prediction equations included an extraneous BFD2 variable from 71.8 to 99.7% of the time. In addition, the probability of including an extraneous carcass weight x BFD variable increased from 10 to 45.6% as the magnitude of measurement errors increased. Equations developed from BFD with measurement errors resulted in biased prediction of FFL%. The level and type of measure-ment errors should be evaluated.