Completely random design does not consider confounding factors, while random block design makes confounding factors evenly distributed in the same block by setting blocks. Completely random design variance analysis belongs to one-way variance analysis, and random block design variance analysis belongs to two-way variance analysis. The former mutates and splits: SS total amount = between groups SS+SS error (or within SS group), while the latter mutates and splits: SS total amount =SS blocking +SS treatment+S.
Design key
The key of completely random design is to randomly group the test materials first. There are many methods of random grouping, the most commonly used are random number table method, lottery method and computer programming data processing method (computer method), and the random number table method is better. Because all the numbers in the random number table are compiled according to the principle of random sampling, any number in the table is completely random at any position. Please refer to the relevant instructions for the use of the random number table.