Optimum Sample Size in Group Comparisons in Animal Breeding Researches with Simulation Study

Abstract

The power analysis performed prior to the study is intended to calculate the required sample size directly. On the other hand, the number of subjects using post hoc power analysis shows the power of working according to the effect size and the accepted type I error level. This research was conducted to determine the optimal sample size in Animal breeding surveys. Especially, it was done to evaluate the level of significance varying with the increase in sample size. In the study, the breeding rate data of 37201 kid goats were used. Gender and two Mature goat age groups were compared with independent two sample t-test in terms of an effect on the growth rate of kids. The mean, standard deviation, common standard deviation, mean difference (effect size, d), standardized effect size (SES) and p-value for the groups were calculated. When the Type I error is 0.05; optimum sample size and the power of the test are calculated in the specified scenarios. In general, it has been determined that the power of the test is appropriate for the optimum range of influence sample size and degree of freedom. Required sample sizes are shown. The contribution to the sample size of the changes in effect size is quite large. Researchers can calculate the strength of the test by taking advantage of the sample size suggested in the study before setting up the trial setups.

Keywords: Goat; breeding; effect size; sample size

DOI: 10.7176/JSTR/5-2-33