Fumbuka, S S2016-04-042020-01-072016-04-042020-01-071976Fumbuka, S. S. (1976) Some agricultural field experiments in Tanzania: a study in their statistical analysis and further prospects, Masters dissertation, University of Dar es Salaam. Available at http://41.86.178.3/internetserver3.1.2/detail.aspxhttp://localhost:8080/xmlui/handle/123456789/53Available in print form, East Africana Collection, Dr. Wilbert Chagula Library, Class mark (THS WRE TD365.J67)This study consists of three chapters. In the first chapter the study analyses an experimental which was conducted at the Faculty of agriculture, Morogoro to investigate the effect of fertilizes N, P and K on the dry matter yield of a grass legume pasture and also any change in the effect of fertilizers were effective with N being the most important factor in this respect and especially in its linear effect. There is also evidence of a strong interdependence among the factor N, P and K as indicated by the high significance of two factor interactions notably NXP. Indeed even three factor interaction shows significance. In the second chapter the study considered two experiments from the 1975/76 set of trials in selection of cotton varieties conducted at the Ukiriguru Agricultural Research Institutes Mwanza. Variety differences are significant but owing to the incomplete nature of the experiments analysed, it is not possible to appraise the effect of fertilizer, spraying or their interaction on the performance of varieties with respect to the different quality variable studied. No variety or group of varieties can be considered to be best with respect to all quality variable and to their selection of best varieties can only be a compromise process. In the third and final chapter it is pointed out that the traditional method at Ukiriguru of conducting the experiments under “unfertilized and unsprayed” and “fertilized and sprayed” condition, does not permit the estimation of the effect of fertilizer spraying or their interaction separately and hence it is impossible to determine the effect of their respective interactions with treatment (variety) effects. As a remedy it is suggested that the experiments should instead be conducted under all the four possible combinations of fertilizer and spraying and a method of analysing such experiments is briefly considered.enAgricultureStatisticsTanzaniaExperimentationThesis