Table 2.
Estimated regression coefficients from partial least squares models for prediction of flavour liking using concentrations of selected odour-active volatiles in aged beef. Models based on one latent variable or factor.
| Variables | Regression Coefficient |
|---|---|
| Constant | 69.9888 |
| 2-Ethyl-6-methylpyrazine | 0.3508 |
| 2,5-Dimethylpyrazine | 0.3466 |
| 2-Ethyl-3,5-dimethylpyrazine | 0.344 |
| 2-Acetyl-1-pyrroline | 0.3343 |
| 3,5-Diethyl-2-methylpyrazine | 0.3189 |
| 3-Ethyl-2,5-Dimethylpyrazine | 0.317 |
| 2-Acetyl-2-thiazoline | 0.2976 |
| 2-Methylbutanal | 0.2880 |
| 2,3-Diethyl-5-methylpyrazine | 0.2820 |
| Butylformate | 0.2718 |
| Ethanol | −0.2631 |
| Acetic acid | −0.2569 |
| 3-Hydroxy-2-butanone | 0.2556 |
| 1-Hexanol | 0.2547 |
| Methylbutanoate | 0.2314 |
| 2-Methylpropanal | 0.1713 |
| Acetone | −0.1299 |
| Dimethyl trisulfide | 0.1238 |
| Heptanal | −0.0908 |
| Nonanal | 0.0764 |
| Octanal | −0.045 |
| Hexanal | 0.0442 |
| s.d. | 4.61 |
| Osten’s F-test | <0.001 |
| RMSECV | 2.31 |
| Correlation coefficient | 0.94 |
s.d.: standard deviation. RMSECV: root mean square error of cross validation.