Statistical methods for estimating individual-level relationships from aggregate data. Used extensively in political science where individual-level data (surveys) are unavailable or unreliable, particularly for historical elections or in international contexts. Associated with Gary King's EI method.