Samarendra Nath Roy received MSc (1931) from University of Calcutta. He was Scientist, Indian Statistical Institute; Head, Department of Statistics, University of Calcutta; and Professor of Statistics, North Carolina State University, USA.
Academic and Research Achievements: Roy’s contribution to statistics related to: (i) multi-variate analysis, (ii) testing of hypotheses, and (iii) analysis of categorical data. In some of his earlier work on multi-variate analysis, Roy depended heavily on rectangular coordinates and higher dimensional geometry. Later, he used the more elegant method of matrix algebra, and obtained useful results on matrix calculus. Studying the various aspects of testing composite hypotheses, he proposed a heuristic method of test construction, later known as the union intersection principle. He devised methods of analysis for multidimensional contingency tables of various types. This required a new formulation of the concepts of correlation, regression, dispersion, etc. He also discussed the analysis of mixed types of models in the sense of being partly normal variable type and partly categorical type. He was also interested in the design and analysis of experiments with more than one factor and more than one response. He wrote a book titled Some Aspects of Multivariate Analysis.,/i>
Awards and Honours: Professor Roy was elected Fellow of the International Statistical Institute; and President, Statistics Section, Indian Science Congress (1948).