MATRIX ORGANIZATION: A SOCIAL EXPERIMENT.

A social experiment was used to assess the effects of matrix structure on organizational processes, role perceptions, and work attitudes. Tests of hypotheses employed a nonequivalent control group design with statistical procedures simulating a complementary, quasi-experimental design termed treatment-effect correlations. Implementing matrix structure caused predicted increases in the quantity of communications, but decreased the quality of these communications. Negative effects on relevant role perceptions, work attitudes, and coordination also occurred. Implications of these findings are discussed in relation to the literature on matrix organization, and suggestions for further research are advanced. [ABSTRACT FROM AUTHOR]