In the classical design of experiment, a randomized complete block design (RCBD) is a helpful experimental design because this design uses a small number of experimental units. The randomized complete block design is comprised of two factors, i.e., one nuisance factor and one potential factor. In many real experiments, some data may be missing or unrecorded. The instant formulae were not provided for an analysis of variance in this case. This paper took into account the RCBD with a treatments and b blocks. In this contribution, the RCBD with an unrecorded value was analyzed by means of the exact scheme with the general linear model. Due to no ready-made formula in the past, this research paper provided the mathematical formulae for the fitted parameters and the overall regression sum of squares (ORSS) for the full model of experimental data.
展开▼
