Author Information

Paul N. Hinz
Mario R. Pareja

Is the first presenter a student?

No

Type of Submission

Paper/Presentation

Abstract

The advantages of repeating experiments in several locations and years are discussed and standard methods of analysis are reviewed. The methods assume that the same treatments are used in each experiment. This paper discusses a method used for a combined analysis when the treatments represent levels of a quantitative factor but differ among experiments. The method makes use of multiple regression analysis in which a continuous variable represents treatment levels, classification variables represent experiments, and products of the continuous and classification variables represent differences among experiments. The method is illustrated on data from a series of experiments designed to study the relationship of grain yield of soybeans as affected by the density of the weed species velvetleaf. The analysis determined that yield loss was linearly related to weed density but that the slope of the relationship differed among years. The slope differences were correlated with August rainfall, and a model is suggested that accounts for both within-experiment variability due to weed density and between-experiment variability due to August rainfall

Keywords

combined experiments, multiple regression, velvetleaf, soybean

Creative Commons License


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

Share

COinS
 
Apr 30th, 4:15 PM

A COMBINED ANALYSIS OF EXPERIMENTS WHEN TREATMENTS DIFFER AMONG EXPERIMENTS

The advantages of repeating experiments in several locations and years are discussed and standard methods of analysis are reviewed. The methods assume that the same treatments are used in each experiment. This paper discusses a method used for a combined analysis when the treatments represent levels of a quantitative factor but differ among experiments. The method makes use of multiple regression analysis in which a continuous variable represents treatment levels, classification variables represent experiments, and products of the continuous and classification variables represent differences among experiments. The method is illustrated on data from a series of experiments designed to study the relationship of grain yield of soybeans as affected by the density of the weed species velvetleaf. The analysis determined that yield loss was linearly related to weed density but that the slope of the relationship differed among years. The slope differences were correlated with August rainfall, and a model is suggested that accounts for both within-experiment variability due to weed density and between-experiment variability due to August rainfall