Document Type

Dissertation

Degree

Doctor of Philosophy

Department

Industrial and Systems Engineering

First Advisor's Name

Chin-Sheng Chen

First Advisor's Title

Committee Chair

Second Advisor's Name

Shih-Ming Lee

Third Advisor's Name

Ronald Giachetti

Fourth Advisor's Name

Irtishad Ahmad

Keywords

Span of control, Organization Theory

Date of Defense

3-23-2011

Abstract

The span of control is the most discussed single concept in classical and modern management theory. In specifying conditions for organizational effectiveness, the span of control has generally been regarded as a critical factor. Existing research work has focused mainly on qualitative methods to analyze this concept, for example heuristic rules based on experiences and/or intuition.

This research takes a quantitative approach to this problem and formulates it as a binary integer model, which is used as a tool to study the organizational design issue. This model considers a range of requirements affecting management and supervision of a given set of jobs in a company. These decision variables include allocation of jobs to workers, considering complexity and compatibility of each job with respect to workers, and the requirement of management for planning, execution, training, and control activities in a hierarchical organization. The objective of the model is minimal operations cost, which is the sum of supervision costs at each level of the hierarchy, and the costs of workers assigned to jobs.

The model is intended for application in the make-to-order industries as a design tool. It could also be applied to make-to-stock companies as an evaluation tool, to assess the optimality of their current organizational structure.

Extensive experiments were conducted to validate the model, to study its behavior, and to evaluate the impact of changing parameters with practical problems. This research proposes a meta-heuristic approach to solving large-size problems, based on the concept of greedy algorithms and the Meta-RaPS algorithm. The proposed heuristic was evaluated with two measures of performance: solution quality and computational speed. The quality is assessed by comparing the obtained objective function value to the one achieved by the optimal solution. The computational efficiency is assessed by comparing the computer time used by the proposed heuristic to the time taken by a commercial software system. Test results show the proposed heuristic procedure generates good solutions in a time-efficient manner.

Share

COinS