Managerial decision making processes and affective outcomes as a function of individual factors and self-efficacy beliefs

摘要

Making decisions in the business environment is arguably the most challenging aspect of managers' yet also the easiest to fail in. Unlike individual decisions managers as agents for their organizations make decisions amidst high levels of ambiguity, incomplete information and mostly under time pressure. These are the very conditions that make managers vulnerable to the volition-undermining potential of decision-generated affect precisely when they are feeling over-extended to deal with such demands. Effective managerial decision making (MDM) involves more than applying a set of individual abilities. Managers face numerous obstacles, failures, and setbacks that often carry perturbing self-evaluative implications as well as social consequences that undermine their self-evaluations in ways that impair good use of their decision making skills (Bandura, 1997). Given the absence of a coherent theoretical framework in the literature the conceptual model of relations put forward attempts to organize and simplify how managers make decisions as agents of their organizations. Most conceptualizations apply oversimplified models that focus attention on one or a few variables, neglect the joint constellations of individual variable factors and the influence of individual self-generated influences as a contributing factor in MDM. As an ex post facto explanatory-predictive study the present research offers evidence of these links among the theoretically relevant constructs in order to formulate an account of their relations in a parsimonious framework that could guide future insights to explain and predict the intentions and direction of managerial decision behaviour. Conceptual research has outpaced empirical research in decision making of managers in organizations. A number of mini-theories exists that focus on a few variables using linear, antecedent-consequence relations with manipulations in laboratory environments that deal with decisions in contexts that are very different to those faced by managers. There is limited research on managers as research participants and empirical findings based on non-managerial samples and students may not generalize to managers in real life decision making. The present research used a non-probability, purposive sample (N = 196) of experienced managers in the Western Cape region of South Africa, all employed in private and public organizations (mean age 38.9 years, SD of 7.49, ethnic black managers constituted 15.8 percent of the sample). As part of the study it was necessary to construct and validate custom indicator measures in an independent pilot study from the same population. The pilot study determined the factor structures of the dimensionality and internal consistency of the custom-designed measures by way of both convergent, as well as, discriminative validity. The exploratory factor (EFA) and internal reliability analyses succeeded to provide both a comprehensive and empirical grasp on the constructs as was defined. Further, analyses of both standardized and custom-designed also revealed no significant difference between black and other managers across the pilot samples which provided confidence of the substantive relations of interest (i.e., the associations among the variables). Structural equation modeling (SEM) was chosen as the data analysis strategy of choice and a confirmatory factor analysis (CFA) demonstrated that the operational measures by and large succeeded in providing both a comprehensive and empirical grasp on the constructs as defined. The inter-construct relations were also consistent with expectations. Evidence for convergent validity however proved that the indicator measures for the allocation of attentional resources were less than adequate in order to provide an uncontaminated measure as a latent variable. The structural model was subjected to further scrutiny by way of a spectrum of goodness-of-fit statistics. The analyses revealed that the model was not adequate and the null hypothesis that the model fitted the population data was subsequently, rejected. It was also sensible to assess the degree of lack of fit of the model with reference to RMSEA which revealed a value of .08, that suggested a reasonable model fit. The poor structural model fit could however be attributed to the failure of the measured indicators used to provide an acceptable grasp of the allocation of attentional resources as a latent variable. The inherent structural flaws in the model could however not be unequivocally be ruled out as an additional possibility of poor fit. One conclusion is the possibility of an expanded model that requires additional indicator measures and additional paths. Notwithstanding these limitations, the present research provided support for social cognitive theory that underlies the model. In accordance with the literature and empirical findings the present research demonstrated mangers' decision making is much more than reason-based behaviour. The present research demonstrates the interdependencies and cumulative effects among individual factors, self-efficacy beliefs and temporal volitional processes, as psychological iv mechanisms through which social-structural factors are linked to the quality of MDM processes. The present research also presents an argument for the independent contributions of self-efficacy beliefs as causal influences on "hot temporal processes" that promote accuracy in decision making. Although present research demonstrates that the estimates were greater for cognitive ability than for both self-efficacy beliefs and social self-confidence it does not suggest that personality traits and self-efficacy beliefs have no utility. The present research demonstrates that cognitive ability combines with personality traits, self-efficacy beliefs and temporal processes (decision-generated affect and the allocation of attentional resources) in a complex manner through multiple pathways.