Treatment-By-Time Interaction In Coconut Experiment

Abstract

The appropriateness of split-block analysis and multivariate analysis for testing treatment-by-time interaction in coconnt experiments were discussed and illustrated with experimental data from an agronomic and a breeding trial. It was shown that inference drawn through these procedures not necessarily he the same, emphasizing the need for testing the assumptions and making a judicious choice. The split-black analysis is suggested whenever the assumption of equal variances (at each time) and equal covarianccs (between pairs of time) hold goad but on violation, alternative methods are to he followed; the easiest being the multivariate analysis.

Description

Keywords

Citation

Journal of Plantation Crops 26 ( 1 ) : 20-24, June, 1998

Collections