Cheese Yield Experiments and Proteolysis by Milk-Clotting Enzymes

Accurate estimates of differences in yield of cheese are important in experiments comparing treatments.Small differences (.1 %) are economically important.Estimates of least significant difference are appropriate for evaluating those differences; expression as a percentage of yield is preferred to expression as weight.A low least significant difference is essential to prevent a Type II error in statistical inference, viz., concluding no difference when an economically important difference exists.The results of an experiment showing no effect of a treatment, but with a high LSD, are misleading.A least significant difference of ± .1 % seems to be a practical lower limit, although lower values would be desirable.Important factors in accurate yield estimates are: accuracies of estimating quantities of milk and cheese; accuracy of sampling and analysis; and experimental design, replication of trials, and statistical analysis.Adjustments of yields to constant moisture (whey) and to constant fat losses in whey are useful.The significance of yield experiments should include both statistical and ec0nomical inferences; the least significant difference should be low to conclude no effect of a treatment on yield.Most milk-clotting enzymes are more proteolytic than chymosin, when com-