Multidimensional Scaling, the precursor to Principal Components Analysis, Common Factor Analysis, and related techniques Multidimensional scaling is an exploratory technique that uses distances or disimilarities between objects to create a multidimensional representation of those objects in metric space. In other words, multidimensional scaling uses data about the distance (e.g., miles between cities) or disimilarity (e.g., how (dis)similar are apples and tomatoes?) among a set of objects to “search” for some metric space that represents those objects and their relations to each other.