Efficient statistical design in two-phase experiments on vase life in carnations (Dianthus caryophyllus L.)

Good experimental design is necessary for the success of experiments. Many experiments on ornamentals comprise at least two phases, for example vase life trials in carnations. The first phase entails the cultivation in the greenhouse and the second the vase life assessment in the laboratory. In field or greenhouse experiments, it is common to use efficient experimental designs such as α-designs or row-column designs, while in the laboratory quite often no particular experimental designs are applied. We conducted vase life trials in carnations, using a randomized layout in both phases. We utilized the data from these trials to compare with competing designs based on the response to selection. This comparison revealed that the choice of design is crucial.