An Alternative Approach for Interpretation of Data Collected Ffrom Groups of Experiments

Abstract

The usual approach of interpretation of data collected from groups of experiments aims at finding out. If any treatment contrasts remains the same or not within permissible error from environment to environment. It does no throw any light as to if any particular treatment has a tendency to behave uniformly or otherwise with changing environment. Keeping the later objective in view, an alternative approach of interpretation of such data has been presented in the thesis. This method was initiated by Finlay and Wilkinson (1963) and later by Eberhard and Russell (1966) for studying the Stability of performance of different varieties of barley. The same technique has been extended here for assessing manorial treatments from data obtained from groups of experiments conducted by adopting regular designs and also from simple fertilizer trials conducted on cultivators’ fields. The methodology adopted consists essentially of obtaining a treatment index as an average of the treatment yield from an experiment conducted in an environment (i.e. a place in an year) and also and environment index as the average of the experiment. Thus for every treatment we have a pair of observations consisting of treatment index and an environment index for every treatment in the experiment. There will thus be as many such pairs as there are environments. An inverse of the repression coefficient of the treatment index on the environment index has been taken as a measure of stability of the treatment with changing environment a treatment with higher stability and sufficiently high productivity can considered as suitable for adoption under uncertain environmental conditions. Again a treatment with high regression coefficient and also a high yield can be considered suitable for adoption under ensured better environment. This methodology has been applied for interpretation of data collected from three controlled experiments conducted in research stations. Of these experiments one was on wheat and two on sugarcane such running over four years. Data from simple fertilizer trials considered for interpretation were collected from trials on sugarcane. Paddy and wheat conducted generally for four to five years in several districts ranging from three to eight in number. From an interpretation of the results obtained from the analysis of these data. Some treatments could be singled out as promising in specific regions.