Document Type

Closed Project

Publication Date

Spring 2002


Dundar Kocaoglu

Course Title

Decision Making in Engineering and Technology Management

Course Number

EMGT 530/630


Decision support systems, House buying -- Decision making -- Methodology, Hierarchical decision model


The home-buying decision process was examined for a couple buying a home. A fourstep decision process was developed by the team and tested on a couple considering buying their first home. The steps were: 1) specify primary criteria; 2) give weights to the criteria via pairwise comparison [1]; 3) select a slate of candidate homes; and 4) rank the homes via pairwise comparison for each criteria.

A team member and his spouse whom had purchased homes in the past were asked to specify the criteria most important to them (e.g., price, location, floor plan, etc.). Criteria that were not preferentially independent were grouped to result in five main criteria. Pairwise comparison was used to determine the relative weights of the criteria. Several homes were then selected that met the minimum criteria – price range, number of bedrooms, general location, etc. The couple then selected five homes from the list that they would consider if they were purchasing a home now.

Pairwise comparison was then used to develop weights for each home for each criterion. By multiplying the 1x5 criteria weight vector by the 5x5 house/ criteria matrix, a 1x5 vector of “Values” for each alternative (home) was obtained. Based on these values, the homes were ranked from highest value (most desirable) to lowest value.

The decision process developed above was then tested with a couple currently seeking to purchase a home. The same steps were followed with the “sample couple.” The couple felt the structured decision process was very helpful in helping them to select a home.

A primary assumption made was that buyers considering using this decision process have already considered their finances; that is, they have spoken with a lender to assess their credit and the maximum loan amount they can expect to have approved. It was also assumed that the grouping of criteria within major subheadings (e.g., “Location”) was reasonable and that the team had correctly identified those criteria that are, and are not, preferentially independent. Lastly, for purposes of this project, “virtual visits” to the homes were made via the multiple listing service (MLS) website data. For purposes of ranking the homes, this was considered equivalent to an actual visit to the home.

Note: The presentation associated with this report is included here as a supplemental file.


This project is only available to students, staff, and faculty of Portland State University

Persistent Identifier