Construct Two-Phase Experimental Designs with Correlated Errors

Tools for constructing and analyzing two-phase experimental designs under correlated error structures. Version 1.1.1 includes improved efficiency factor classification with tolerance control, updated plot visualizations, and improved clarity of the results. The conceptual framework and the term two-phase were introduced by McIntyre (1955) ).

TwoPhaseCorR

Two-Phase Experimental Designs under Correlated Observations

TwoPhaseCorR is an R package for constructing and evaluating two-phase experimental designs, particularly in the presence of correlated observations between phases. It supports the computation of information matrices and Canonical Efficiency Factors (CEF), helping researchers identify efficient designs under various intra-block correlation scenarios.

Features

• Constructs cyclic two-phase designs for a specified number of treatments (v).
• Models intra-block correlation via a correlation parameter (rho).
• Computes information matrices for treatment effects and interactions.
• Estimates Canonical Efficiency Factors (CEF) for assessing design efficiency.
• Visualizes the relationship between CEF and rho to guide optimal design choices.

Installation

From CRAN (after approval)

install.packages("TwoPhaseCorR")