Assessment of bivariate normality

There are three methods, which are most commonly used to assess the bivariate normality of paired data, two of which are also used to assess the multivariate normality. Nevertheless, none of the methods is very efficient or conclusive in their assessment of bivariate normality. In this thesis we are proposing a new method to test bivariate normality. This new method makes use of a set of if and only if conditions inherent in the theory of bivariate normal distribution. The proposed new method is highly efficient, accurate, and very easy to apply using any available standard statistical software