Disturbance Term Regression Test Procedures for Recursive and Nonrecursive Models: Solution From Intercorrelation Matrices.

Abstract

The logic and procedures underlying a disturbance term regression test of logical consistency for structural models are reviewed for recursive and nonrecursive designs. It is shown that in a simple three-variable, complete mediational case the test procedure is mathematically equivalent to a part correlation. Procedures are outlined that allow researchers to conduct disturbance term regression tests using only the matrix of intercorrelations among structural variables. General procedures are outlined (a) to test mediational hypotheses in recursive designs and (b) to test for omitted cross effects in nonrecursive designs. The relative virtues of the disturbance term regression test and other fit indices are discussed.