Determination of Warner-Bratzler shear force thresholds for classifying meat tenderness of Kivircik lambs: A study in Türkiye

Abstract

The aims of the study were to identify the Warner-Bratzler Shear Force (WBSF) threshold value that corresponds to the transition from perceived tenderness to toughness in lamb meat through sensory panel evaluations, and to compare the classification performances of the threshold values determined using linear regression and Receiver Operating Characteristic (ROC) analysis methods. The study data comprised WBSF analysis results of the longissimus thoracis et lumborum muscle from 405 Kivircik lambs, along with tenderness scores determined by a trained sensory panel. WBSF threshold values were determined separately for the “panellists' sensory tenderness evaluation dataset (raw dataset; n=3049)” and the “mean score dataset created by averaging the tenderness scores given by the panellists for each lamb (individual lamb dataset; n=405)”. Higher classification accuracy was achieved in the mean score dataset compared to the individual score dataset using both linear regression and ROC analysis methods. The WBSF threshold value for the mean score dataset was determined as 40.9 N by ROC analysis and 48.0 N by linear regression analysis. The accuracy of these threshold values was calculated to be 82.2 % and 85.4 %, respectively. The threshold value based on ROC analysis provided higher sensitivity and negative predictive value, whereas the threshold value based on regression analysis was advantageous in terms of specificity and positive predictive value.