Multiple Comparison of Means and Misuse of Split Plot Design with Quantitative Data

Split plot design is one of the most popular experiments in crop field study. However, several problematic areas still exist when scientists use split plot design with regards to statistical analysis, particularly in multiple comparison. Often many important results were not discovered, thus causing data loss. Therefore, by examples, we introduce the F test and mean multiple comparison for split design in this paper, then survey some misuses of split plot design in papers published and suggest improvement.