Date of Defense
Industrial and Manufacturing Engineering
Dr. Steven Butt
Dr. Betsy Aller
A hospital in Kalamazoo, Michigan sought to improve its ability to select the best available products. The hospital did not have a reliable method for product comparison. Intuition, personal opinion, and price were many times the deciding factors in the process. At the beginning of the project, the group worked with the hospital and project advisor Dr. Steven Butt to establish two major objectives. The first objective was to help the hospital select the best available products, according to their needs. The second objective was to implement and compare the use of several analytical techniques for product comparison, including Data Envelopment Analysis, the Analytical Hierarchy Process, and rank-based (non-parametric) statistics. The best available products for the hospital were selected from the results of the three analytical techniques, as well as some additional considerations. Physical product characteristics from the manufacturers and user rating from a non-scientific survey were used as inputs for the analytical techniques. Paper towel model #20305 (manufactured by Georgia-Pacific) was selected as the best paper towel, because of its high rankings from the analytical techniques and its low price. Toilet paper model #2122 (manufactured by Kimberly-Clark) was selected as the best toilet paper, because of its number one ranking in each of the analytical techniques. Through the implementation of the analytical techniques, the group concluded that rank-based (non-parametric) statistics is the simplest technique to understand and the least expensive to apply. Based on these attributes and this project's result of no statistical difference between the techniques, the group recommends that the hospital implement rank-based (non-parametric) statistics for future product selections. A standard operating procedure was developed using an Excel spreadsheet model to allow the purchasing department to easily implement the method. Finally, the group recommends that a future study be completed with more complex products and one or two additional analytical techniques. This type of study would provide a broader base of information for product comparison at the hospital.
Baker, Robert; Manurung, Metz; and Williams, Jesse, "Multi-Attribute Product Comparison Using Analytical Techniques" (2002). Honors Theses. 2052.
Honors Thesis-Campus Only