Application of nonparametric covariance analysis in field trial with reference to YLD management trial in arecanut

摘要

Linear covariance analysis (ANCOVA) is generally used to control the experimental error due to the variations in pretreatment values in disease or pest management trials. In linear covariance analysis, it is assumed that the covariate (pre treatment value) is linearly related to the response variable. But in many situations this assumption may not satisfy. In this paper we propose nonparametric covariance analysis (NPANCOVA) which does not require much assumption about the functional relationship between the response variable and the covariate. The only assumption about the relationship is that it is a smooth function. The method is also extended to analyze the data under the presence of treatment x covariate interaction effect. The performance of the proposed method is verified through simulation studies. The method is applied to the data on Yellow Leaf Disease (YLD) management trial in arecanut. The comparison of mean square errors (MSB) indicated that the performance of the proposed method is better than the traditional ANCOVA technique.