Study of different parametric stability measures when the basic data/variables are non-normal

The presence of genotype-environment interactions (GEI) necessitates the developments of varieties or breeds suited for different agro-environments based on their stability and adaptability characteristics. In many situations, the assumptions about the normality and independence of observations as well as homogeneity of error variances are not fulfilled. Therefore, there is a need to investigate the performance of different parametric methods for stability measures when the basic data is not normally distributed. This important aspect is taken up in the present investigation. Using simulation technique, power of the test has been computed for different sample sizes when the underlying dataset is normal as well as non-normal like gamma, beta, t, weibull, log normal etc. In most of the cases it is found that Eberhart and Russell parametric stability measure gives better performance.