A statistical technique used to reduce the differences between variables in order to classify them into a set number of broad groups. In finance, this technique is used to compress the variance between securities while also allowing the person to screen for several variables. It is related to discriminant analysis, which, in simplified terms, tries to classify a data set by setting a rule (or selecting a value) that will provide the most meaningful separation. Although this technique requires a fair bit of mathematics, it is relatively simple. MDA allows an analyst to take a pool of stocks and focus on the data points that are most important to a specific type of analysis, shrinking down the other differences between the stocks without totally factoring them out. For example, MDA can be used for selecting securities according to the statistically-based portfolio theory set forth by Harry Markowitz. Properly applied, it will factor out variables like price in favor of values that measure volatility (beta) and historical consistency. Edward Altman is famous for using multiple discriminant analysis in creating the Altman-Z score.