| This dissertation examined Argyris' conceptual framework (1977) which proposes linkages between the influence of organizational defensiveness on the strategic use of double-loop learning behaviors. The study represents a first attempt to empirically test theories proposed in the literature which connect these two concepts. Through the use of structural equation modeling, the study also examined the predictive power of communication variables and their influence as antecedents of organizational defensiveness. These variables included communication satisfaction with superiors, communication satisfaction with subordinates and situational "embarrassability." Participants included members of three organizations: a county government department of social services, a large teaching hospital and an inner-city library system (N = 485). Also, pilot testing involved an additional 202 participants who represented various organizations.; Six hypotheses were proposed. H1, which posited a significant negative relationship between perceived organizational defensiveness and perceived support of double-loop learning, was supported in most models. H2, which proposed a significant negative relationship between communication satisfaction with superior and perceived organizational defensiveness, was also supported. H3, which stipulated that there would be a significant positive relationship between communication satisfaction with subordinate and perceived organizational defensiveness, was not proven. Additionally, H4, which predicted a significant positive relationship between situational "embarrassability," or tolerance to embarrassing situations, and perceived organizational defensiveness, was not supported.; The two remaining hypotheses were descriptive. These included H5, which suggested that most employees perceive that defensiveness is prevalent within their respective organizational climates, and H6 , which proposed that the majority of employees believe that double-loop learning is generally not encouraged within organizations. Both of these hypothesis were supported by the findings.; Finally, to determine directions for future research two other models were tested: one controlled for position (supervisory and non-supervisory) and the other controlled for commitment (committed or non-committed). In general, hypotheses were not supported for the all supervisory and the non-committed sample. The hypotheses were, however, supported when the model was imposed on the committed group. |
