Skip to main content
. 2018 Jan 17;13(1):e0190680. doi: 10.1371/journal.pone.0190680

Fig 3. Predicted latent hedonic scores of tenderness attribute in plain ground beef samples.

Fig 3

Predicted scores assume that orderA = orderB = orderC = 1/3 and that other explanatory variables are observed at their samples means. Parametric bootstraps are used to generate p-values for the null hypothesis that the predicted score for any one beef sample is equal to the predicted score of another sample.