Technical Note: Estimating Statistical Power of Mixed Models Used in Dairy Nutrition Experiments

摘要

Statistical power is defined as the probability of correctly rejecting the null hypothesis. Power calculations may be useful in planning experiments. The objective of this technical note is to outline an applied method that estimates statistical power of a dairy nutrition experiment that employs a Latin square as the experimental design. Because the SAS MIXED procedure (PROC MIXED) is commonly used to analyze data sets, this note outlines basic programming procedures that may be used to estimate statistical power of a mixed model using this procedure.