Page 98 - 33-3

Goal Consensus, Subordinates’ Prior Performances, and Supervisors’ Resource Allocation Preferences

in their positions are 8 and 6 years, respectively. The surveys were conducted between
January 1, 2017 and June 30, 2017. The questions on the regional/branch manager surveys
are included in Appendix A. Finally, we collect the case company’s monthly financial
information and data related to the branch offices from January 1, 2017 to June 30, 2017

and match them with the survey data. 4

3.3 Measurement of Independent Variables
3.3.1 Goal Consensus between Regional and Branch Managers (CONSENSUS )
i,t
We operationalize the goal consensus between both regional and branch managers by
measuring how consistently a regional and branch manager pair prioritize the four goals.
This method of measuring goal consensus helps us to identify the level of agreement
between a regional and branch manager regarding the importance of the goals pursued by

the case company.
We use the Analytic Hierarchy Process (AHP) approach developed by Saaty (2000)
to develop a ratio scale from the pairwise comparisons to obtain the priorities of both
regional and branch managers. The AHP approach has several advantages. First, it

addresses the direction of the difference between two profiles and provides an objective
formula by which to process the subjective and personal preferences of an individual,
thereby overcoming the major limitations of Euclidean distance. Additionally, the
normalized relative weight provides more information than the ranking. Second, the

AHP works by developing priorities for alternatives, and the prioritization process solves
the problem of handling each subject’s distinct significance with respect to the values of
different alternatives. Finally, the AHP allows inconsistency in judgments and provides
users with directions for improving their judgment and understanding of the problem (Saaty,

2000).
We ask both regional and branch managers to compare all the criteria in a pairwise
fashion, where each level (six comparisons in all) uses a scale ranging from 1 to 9. The
judgments of the pairwise comparisons help us to establish a matrix. After constructing the

4 The information and financial department in the case company help us collect and confirm corrections
to all the financial data they obtained. We winsorize the 1% extreme observations on each tail for
continuous regression variables to control the effect of outliers on our empirical results.

90