The evaluation of arabica coffee cultivars for resistance to coffee berry disease: Some comments

Summary The preselection test as used in the Kenya arabica coffee breeding program was critically examined prior to use in a similar program in Uganda, since the criticism by Van der Graaff (1982) of the an der Vossen et al., scaling technique for measurement of seedling resistance to Colletotrichum coffeanum may alter the interpretation of results.The published data were re-analysed using non-parametric statistical techniques and it is concluded that, despite the criticism, the method appears to give satisfactory results. It is suggested however, that the mode of the distribution obtained by the selection test be considered as a suitable alternative to the mean in the interpretation of results.