Scientific elitism and the information system of science

Scientific elitism must be viewed as a multidimensional phenomenon. Ten variables of elitism are considered and a principal components factor analysis is used to scale this multivariate domain. Two significant dimensions of elitism were found; one in basic and one in applied science. Sociometric techniques were used to identify the elite of a scientific discipline in a large metropolitan area. An abstract analytical continuum, the scientific mission, was generated using a Thurstone‐type scale. The scientific mission is a higher order concept and subsumes both idealized polar types of basic and applied science. A scientist's scale score reflects his professional interests and the breadth of his interest space. There is a distinct tendency for elites to cluster at the basic research end of the continuum. It was found that: (a) the ten variables of elitism provide a scale that successfully discriminates between elites and non‐elites; (b) elites process more information than non‐elites; (c) elites had more narrowly defined, less diffuse interest spaces than non‐elites; and (d) elites prefer literature‐oriented methods of procuring scientific information as opposed to person‐oriented methods.