In the previous discussion of several types of environmental designs for controlling experimental error, we were only faced with one type of Experimental Unit for all treatments and a randomization process for assigning treatments to experimental units. However, in the factorial experiment, sometimes we are faced with another situation where there are several types of experimental units and the levels of the experimental factors are placed sequentially and the randomization procedure is carried out separately. For example, from the two factors that we are trying, we make a plot of the experimental unit which is larger for one of the factors, then for each of these plots we divide it again into several plots with a smaller size which is the experimental unit for the level of the second factor. This procedure is nothing but the principle of the Split-Plot experiment . The experimental unit plot which is larger in size and in which there are subplots is called the Main Plot, while the second experimental unit plot which is smaller in size and randomly placed on the Main Plot is called the Sub Plot.
This article is a continuation of the Split Plot Design
Suppose there is a study that wants to examine the effect of combination of NPK fertilization and rice genotype on rice yield (kg/plot). The combined effect of NPK fertilization (A) consisted of 6 levels placed as the main plot (main plot) and the rice genotype (B) consisted of 2 levels placed as subplots (subplot). The main plots were arranged using the basic RAK design with 3 replications. The experimental data and the steps for calculating the analysis of variance followed by Post Hoc test: Fisher's LSD can be studied in the document below.
You can learn the tutorial analyse data using SmartstatXL and SPSS Software on the following link: