Split plot experiment design and a practice

Split plot experiment design (SPED) is a kind of incomplete block design including special randomization procedures. Applying sped is ideal in an experiment if desired precision level for any factor is higher than for the others or more experiment unit is needed than the other factors. It is also be applied for enlarging experimental citation area to add one more factor. Standard split plot experiment design which is the most common type of sped has a factorial design with two factors. a level of factor A is designed in randomized incomplete block design with r replication. Each of the whole plot experiment unit in which factor A is applied is divided into b split plot experiment unit for factor B levels. Either A or B or both of them could be in a factorial experiment. In this study standard split plot experiment design is explained and a practice is given. To increase the wheat yield for Central Anatolia conditions, different crops and fertilizer applications are tested in a sped with 4 replications and analysis of variance and analysis of means is shown. According to the results of the experiment, decrease of the fertilizer level had a negative effect on the yield, winter lentil summer lentil and chickpea had the same yield level with fallow. On the other hand, growing wheat after wheat had caused significant decrease in the yield. When compared to randomized complete block design, 45% increase in precision for split plots and 28% decrease in whole plot precision is obtained. Furthermore, eight variations of sped including augmented split plot experiment design which is not encountered frequently in literature is explained.