THE STATISTICAL DESIGN OF EXPERIMENTS INVOLVING A NUMBER OF GLASSHOUSES
The major purposes of a glasshouse is to control the aerial environment of a crop. Thus any programme for the improvement of crop production in glasshouses is likely to include the investigation of the effects of changes in the aerial environment. The natural experimental unit to use in such investigations is a whole glasshouse, though one might consider obtaining experimental units by partitioning a house. I am concerned with the use of a group of such experimental units and shall refer to it by the term chosen at Littlehampton, a multifactorial unit.
Cooke, D. (1968). THE STATISTICAL DESIGN OF EXPERIMENTS INVOLVING A NUMBER OF GLASSHOUSES. Acta Hortic. 7, 5-13