In this study, F test conducted after missing value estimation (Yates) in randomized block design was compared in terms of type I error rate and power of the test with different methods. These methods were Yates, group mean and new modification of Yates we developed in this study. The experiments containing 3 groups were chosen. In the experiments, the samples including equal observations were drawn from populations, Z (0, 1) distribution. Type I error rate and power of the test were obtained from the samples of 1000000 simulation trials. As a result, new modification of Yates and group mean provided less reliable results than the Yates for small sample sizes. New modification and Yates provided similar results for large sample sizes.