On optimal trend free sampling designs

This paper considers the problem of estimating a finite population mean in the presence of linear trend. Recently, Mukerjee and Sengupta (1990) have studied the existence of an optimal design-unbiased strategy under this set-up and have suggested optimal sampling designs whenever such optimal strategies exist. In this paper, we consider cases when such optimal strategies do not exist and suggest some sampling designs which are optimal within some classes of sampling designs for the use of sample mean, difference and regression estimators