Efficient Design of Experiments for Quality Agricultural Research : Part III

Abstract

Experiments in which data on several responses are measured from an experimental unit corresponding to the application of a treatment are known as multi-response experiments. Multi-response experiments are of two types viz. complete multi-response experiments (all the response variables are recorded from each experimental unit) and incomplete multi-response experiments (recording of all the responses variables from each experimental unit is not feasible). For complete multi-response experiments, it has been shown that the designs that are efficient for single response experiments are also efficient for complete multi-response experiments provided that the response variables are less than the error degrees of freedom. Obtained a method of construction of designs for incomplete multi-response experiments using combination of randomized complete block (RCB) designs as treatment-wise design and balanced incomplete block (BIB) designs as response-wise design. The designs obtainable from this method are economical from resource point of view. Developed a step wise procedure of analysis of incomplete multi-response designs obtained as a combination of RCB design and BIB design. Developed beta version of software for generation of nested block designs both for independent errors and correlated errors. Developed an algorithm for generation of efficient supersaturated designs for two-level factorial experiments and obtained several efficient super saturated designs. Developed a software for detection and handling of outlier(s) in the experimental data. Efficient designs for asymmetric parallel line assays and slope ratio assays have been obtained by deleting all observations corresponding to a dose (of standard or test preparation) from efficient block designs for symmetric parallel line assays and slope ratio assays. To disseminate the knowledge available on combinatorial aspects of designs and analytical procedures acquired to scientists engaged in research in National Agricultural Research System, advisory services were pursued rigorously. Design Resources Server has been strengthened in collaboration with National Professor by adding links on -designs, designs for bioassays, supersaturated designs, modules for generation of randomized layout of square lattice designs, completely randomized designs, RCB designs, Latin square designs and augmented designs. To provide steps of analysis of data generated through designed experiments using SAS and SPSS, a new page “Analysis of Data” has been launched on Design Resources Server. Discussion Board has been initiated for sharing research with fellow scientists over the globe or for flagging issues for attention of scientific community. A list of experts in design of experiments over the globe is uploaded which will be useful for establishing linkages.