Abstract

The two-period repeated measurements crossover design is not often used in agricultural studies. It is, however, an attractive model, involving the confluence of two powerful statistical ideas, treatment crossover and repeated measurements on the same experimental unit. This paper presents one approach for the statistical analysis of such design based on the work of Wallenstein and Fisher (1977). It is shown how the data may be transformed so that it can be analyzed under the framework of a completely randomized repeated measurements design. We formalize the analysis in the context of a forestry experiment conducted on poplar trees (Populus SP.), to compare the efficacy of two treatments to prevent damage by the coleopteran insect Platypus sulcatus (ambrosia beetle). Two insecticides were applied in a crossover fashion to two groups of 8 poplar trees each. Each tree was treated with one insecticide and evaluated on three occasions during the first year, received no treatment during the following one-year washout phase, and then (in the third year) received the other treatment and was evaluated on three occasions. One of the parameters analyzed to test for treatment differences was the number of tree lesions attributed to the insect. We present the results of our work and discuss the potential usefulness as well as the limitations of this interesting design.

Creative Commons License


This work is licensed under a Creative Commons Attribution-Noncommercial-No Derivative Works 4.0 License.

Share

COinS
 
Apr 27th, 2:30 PM

THE ANALYSIS OF THE TWO-PERIOD REPEATED MEASUREMENTS CROSSOVER DESIGN WITH APPLICATION TO A FORESTRY PROBLEM

The two-period repeated measurements crossover design is not often used in agricultural studies. It is, however, an attractive model, involving the confluence of two powerful statistical ideas, treatment crossover and repeated measurements on the same experimental unit. This paper presents one approach for the statistical analysis of such design based on the work of Wallenstein and Fisher (1977). It is shown how the data may be transformed so that it can be analyzed under the framework of a completely randomized repeated measurements design. We formalize the analysis in the context of a forestry experiment conducted on poplar trees (Populus SP.), to compare the efficacy of two treatments to prevent damage by the coleopteran insect Platypus sulcatus (ambrosia beetle). Two insecticides were applied in a crossover fashion to two groups of 8 poplar trees each. Each tree was treated with one insecticide and evaluated on three occasions during the first year, received no treatment during the following one-year washout phase, and then (in the third year) received the other treatment and was evaluated on three occasions. One of the parameters analyzed to test for treatment differences was the number of tree lesions attributed to the insect. We present the results of our work and discuss the potential usefulness as well as the limitations of this interesting design.