مقارنة لبعض طرائق تقدير القيم المفقودة في تصميم القطاعات العشوائية الكاملة

Four methods for tackling missing values were presented in randomized complete blocks design: Yates, Harry, Rubin, Haseman and Gaylar. To prefer between these methods some statistical measurements were used: error mean square (MSe), treatments mean square (MSt) and the value of significant differences between treatments (cal. F) , with addition to the easiest path of statistical analysis. It was found that the most preferable method was Haseman and Gaylar, followed by Harry method and the last one was Yates method. However, Rubin method has failed to go along with the others.