Balanced residual treatment effects designs of first order for correlated observations

Gill [1992. Balanced change-over designs for autocorrelated observations. Austral. J. Statist. 34(3), 415-420] discussed change-over designs of first-order residual treatment effects and obtained conditions for a design to be variance-balanced, which is uniform over periods and binary over subjects, in the presence of autocorrelated errors. In this paper, some new variance-balanced residual treatment effects designs of first order are constructed in the presence of autocorrelated errors that satisfy the conditions given in Gill [1992. Balanced change-over designs for autocorrelated observations. Austral. J. Statist. 34(3), 415-420]. © 2007 Elsevier B.V. All rights reserved