Data Analysis techniques must be appropriate to the purpose of the analysis. Policy analysis, in particular, requires analytic techniques which are parsimonious and robust. Data available on the prices of 22 essential commodities in May 1971 and in December 1973 was subjected to analysis with a view to extracting information on the geographical pattern of price changes. The pattern reveals (a) that wheat behaved differently from all other commodities in 1973, (b) that a concentration of price pressure developed on Ahmedabad and Bombay and away from Delhi and Calcutta, (c) that unexpectedly small price increases took place in two commodities (rice and wheat) in Kanpur. The tools used were Information Theory-leading to a measure of "uneveness", and Cluster Analysis, leading to a simple Guttman scale.