COMPOSITE DESIGNS IN AGRICULTURAL RESEARCH

Abstract
An experiment was conducted in the greenhouse using tomato plants for testing response to nitrogen, phosphorus and potassium on North Gower clay loam and Castor silt loam.Two designs were used to assess this response, a factorial and a composite design. There was general over-all agreement in results obtained by the two methods. The limitations of the composite design and difficulty in interpretation are discussed.It is suggested that composite designs be used in agricultural research with some caution. With only three factors under test, the standard factorial is preferred.