The Impact Of Functional Issue Classification On Managerial Decision Processes: A Study In The Telecommunications Industry.(Stat

In addition, many marketing decisions (e.g., related to such issues as product design, packaging, advertising) involve selecting from a number of "creative" and potentially controversial alternatives. Often the final selection reflects personal tastes and intuition rather than a reliance on hard facts. As a result, decision makers will be willing to expend considerable personal resources in order to influence the decision process.

The literature suggests that marketing decisions can exhibit both rational and political behaviors. This relationship is reflected in the following proposition:

Proposition 3: Marketing decisions will be characterized as not having significantly different levels of rational and political behaviors.

This review also suggests additional propositions regarding differences in processes across different functional classifications, as follows:

Proposition 4: Human resource decisions will be characterized as having a lower level of rational decision-making behavior than financial decisions.

Proposition 5: Human resource decisions will be characterized as having a lower level of rational decision-making behavior than marketing decisions.

Proposition 6: Financial decisions will be characterized as not having a significantly different level of rational decision-making behavior than marketing decisions.

Proposition 7: Human resource decisions will be characterized as having a higher level of political decision-making behavior than financial decisions.

Proposition 8: Human resource decisions will be characterized as not having a significantly different level of political decision-making behavior than marketing decisions.

Proposition 9: Financial decisions will be characterized as having a lower level of political decision-making behavior than marketing decisions.

METHODOLOGY

Participants

Participating companies represented 12 major firms in the U.S. telecommunications industry (SIC 4813). The 96 participants from 24 decision-making units (Out of 40 which were approached to participate in the study) averaged 44.3 years of age and 21 years of industry experience. In all cases, the manager of the unit was at least at the upper middle level of management and could be considered as having a general manager position. Thus, 24 participants were decision unit heads and the remaining 72 (out of 89 who were sent questionnaires) were members of the managers' decision-making teams. Participants also represented a wide array of functional backgrounds.

Procedures

In order to capitalize on the strengths offered by both scenario and multi-case-based approaches, a two-step data collection procedure was employed. First, the top managers of the 24 decision-making units were asked to identify and describe three important recent decisions that addressed marketing, financial, and human resource issues. It should be noted that, for example, a human resource issue occurring in a marketing organization would be classified as a human resource issue.

This approach is premised on the assumption that, on the whole, managers' recollection of issues are a meaningful representation of the actual issues that led to subsequent decision-making processes. While post-choice revisions in issue classification are possible, they are probably the exception rather than the rule.

In order to reduce potential bias in the selection of the decisions, the managers were told only that they were participating in a study of decision making in the telecommunications industry. No information was provided relative to the focus of the study (i.e., the potential linkage between functional issue classification and managerial decision-making behavior) or the dependent variables (i.e., rational and political behavior) being utilized.

A total of 71 decisions were obtained from the 24 participating organizations (one manager did not provide a marketing decision). Appendix A contains a representative sample of the issues addressed by the decisions in each functional classification

Second, the manager was asked to identify those members of his/her management team (generally direct reports) who were part of the decision-making process and could provide additional information regarding each of the three decisions. A prerequisite for inclusion in the study was participation in all three decisions. These members of the manager's decision-making team, in turn, provided the dependent variables (i.e., assessments of rational and political behaviors) via written questionnaires.

Measures

The questionnaire included three sections, each containing one of the three functional decision descriptions approved by the manager and a series of questions designed to provide data relative to the dependent (rational and political behavior) and control variables. Seven-point Likert response formats were used for all variables.

Independent Variables. Scales used to measure the degree of rational and political behavior associated with each of the 71 decisions were taken from Dean and Sharfman (1996). One item from Dean and Sharfman's political behavior scale was replaced in order to improve the original reliability coefficients reported in their study.

Control Variables. Three classification schema -- threat/opportunity (Jackson and Dutton, 1988), issue importance (Billings and Scherer, 1988), and complexity (Campbell, 1988) -- have received substantial support in the literature as prevalent category labels that influence managers' decision-making behavior. Importance, complexity, and threat/opportunity were controlled in this study through statistical tests. Importance was further controlled by limiting data collection to important decisions. Importance, complexity and threat/opportunity were each measured with three-item scales designed to reflect measures used in prior studies.

As a check against potential response bias, no information was provided ahead of time as to the functional classifications being used in the study nor were the managers' classifications of the issues provided to the respondents. As a reliability check on the meaningfulness of the managers' issue classification, respondents were asked to classify each issue after they had completed the questions associated with each decision. The fact that over 85% of these classifications agreed with the decision unit managers' classifications supports the reliability and validity of the classifications provided by the unit heads.

Analysis

Reliability coefficients (alphas or inter-item correlations as appropriate) were computed for all scales used in the study. The independence of the dependent variables was assessed by a factor analysis of the rational and political behavior process scales.

ANCOVA analyses utilizing decision-unit manager classifications and respondent (i.e., nondecision-unit managers) reports of the variables were used to test the overall hypothesis of the study that functional issue classification does impact managerial decision-making behavior. T-tests were utilized to make pairwise comparisons as direct tests of the specific propositions. T-tests and ANCOVAs were also used to investigate other control variables which could have an impact on results.

RESULTS

Table 1 presents results of a factor analysis which, similar to Dean and Sharfman (1993), supports the argument that rational and political behaviors are two distinct dimensions of the decision-making process. Thus, the dependent variables demonstrate discriminant validity.

Table 2 shows means, standard deviations, and correlations among the variables. Reliability results for the dependent variable scales (rational behavior = .81; political behavior = .75) compare favorably with those obtained by Dean and Sharfman (1993). In order to improve the reliability of the Importance variable (alpha = .51), the single item which had the highest content validity, "How important was this decision to the organization?," was used. Procedures using both the three-item and single-item Importance measures yielded essentially the same results.

Table 3 depicts the results of ANCOVAs analyzing the impact of functional classification (the independent variable) on rational and political behavior (the dependent variables). Also shown are covariate effects for the three control variables. Results from Table 3 support the primary thesis of this study that the functional classification of issues exhibits a linkage to managerial decision-making behavior. Specifically, it was found that functional decision classification had a significant (p[less than].01] linkage to rational behavior. The linkage relative to political behavior was found to be marginally significant (p=.06). These results support the need to further assess the relationship between specific functional classifications and both rational and political behavior.

ANCOVA results simultaneously using all variables show that Threat/ Opportunity (p[less than].0l) and Complexity (p.03) were significantly related to rational behavior. Complexity (p.=03) was significantly related to political behavior. The impact of Importance was not found to be significant relative to either dependent variable. These results demonstrate the utility of representing the relationships of multiple contextual influences to the decision-making process. By explaining variance in rational and political behavior beyond that attributable to Importance, Complexity and Threat/Opportunity, these results confirm the significance of functional classification as a distinctly influential factor on decision processes.

Table 4 provides an analysis of the means for each of the dependent variables as a function of decision type. Results indicate that, with respect to human resource decisions, the level of political behavior was not significantly greater (p=.09) than the level of rational behavior. As a result, Proposition 1 was not supported. Conversely, for financial decisions, the level of rational behavior was found to be significantly higher (p[less than].0l) than the level of political behavior, thereby supporting Proposition 2. Marketing decisions displayed a significantly higher (p[less than].0l) level of rational behavior relative to political behavior, contrary to the relationship posited in Proposition 3.

Print Get the Mag Weekly Updates