This reply to Sypher argues that neither his data nor his analysis provide good evidence for confirming the systematic distortion hypothesis in general or showing that it is operative in the communicator style measure in particular. At best, Sypher's research provides additional construct validity for the communicator style domain. It is maintained that “difficult memory conditions” do not prevail for the self-report of style behaviors. Three studies are briefly reported supporting this conclusion. Finally, necessary conditions are identified in order for self-report data to be increasingly useful.

Although the results of factor analyses (and related methods) of sets of ratings are often of immediate and intrinsic interest, there are few personality psychologists today who would claim that obtained factor structures represent inexorable verities (cf. Lykenn, 1971). It is now well recognized that by varying the mix of variables included in a clustering analysis, one can fundamentally alter both the number and nature of the summarizing dimensions subsequently obtained.