ABSTRACT
Organization Development (OD) as a miniparadigm in public administration is especially attractive in that it forces increasingly precise attention to what are usually called intervening variables. Specifically, the analysis that follows implies the usefulness of employing differences in social desirability (SD) as an intervening variable in interpreting effects of OD interventions measured by self-reports. The SD scores of respondents clearly affect initial Likert self-reports, especially in the extreme case. In sum, the present research design seeks to test the usefulness of the SD scale as an intervening variable, as well as to raise issues about self-reports as measures of OD effects. The data are consistent with a number of alternative explanations, especially in the crucial absence of independent behavioral measures of OD effects. The evidence clearly suggests that SD differences among participants explain a safe minimum of 5 to 7.5 percent of the variance in initial Likert scores.
