# W5_S.Algheilani_using non-compensatory in decision making

1. Problem Definition
After graduating from high school the Omani students have many options to start higher education locally or abroad. In this blog the student options are going to be studied and investigated using a simulated scenario of a student graduated from high school.

2. Identify the Feasible Alternative
Ali is making a choice between four universities. There are five criteria that concerns Ali and will affects his decision. The four elements are the Country, living cost, tuition cost, passing grade and the percentage of getting accreditation and he come up with the following inputs.
Criteria Options

Table.1 Summary Information for Choice of a University

Obviously, Ali best option economically wise is to study in Oman. Omani College average rate are much better than the other three options since there is no tuitions fee for the Omani students. Although, there is no tuitions fee to study in Omani College, but that is not the main concern of Ali. Ali wants to find a good university with top quality education level with reasonable tuition cost and acceptable living cost.
3. Development of the Outcome for Alternative
In non-compensatory models, there are four alternative ways for making a choice when multiple attributes are present. We will try to use these models to select one best alternative:

a. Dominanceh0nd
Dominance is a useful screening method for eliminating inferior alternatives from the analysis. By comparing each possible pair of alternatives to determine whether the attribute values for one are better off another, it may be possible to eliminate the inferior ones.
b. Satisficing (Method of Feasible Ranges)
This method requires the establishment of minimum or maximum acceptable values (the standard) for each attribute. Alternatives having one or more attribute values that fall outside the acceptable limits are excluded from further consideration.
c. Disjunctive Resolution
Disjunctive resolution relies on comparing the attributes of each alternative to the standard. The difference with satisficing is that the disjunctive method evaluates each alternative on the best value achieved for any attribute. By applying disjunctive method, all universities will be acceptable because each has at least one attribute value that meets or exceeds the minimum expectation. This model does not discriminate well among the four alternatives.
d. Lexicography
This model is particularly suitable for decision situations in which a single attribute is judged to be more important than all other attributes. The alternative having the highest value for the most important attribute is then chosen. When two or more alternatives have identical values for the most important attribute, then the analyst examines the next most important attribute until a single alternative is chosen.

4. Selection Criteria
Below, results using non-compensatory models?

• Dominance:

Table.2 Check for Dominance among Alternatives

• Satisfying

Table.3 Feasible Ranges for Satisficing

• Lexicography

Table 4. Illustrates the application of lexicography to the ordinal ranking.

5. Analysis and Comparison of the Alternative
From Table 2, it is clear that option 1 dominates option 2. So option 2 should be dropped for further consideration. For dominance, it is not possible for Ali to select the best university. Based on Table 3, option 3 joins option 2 on Ali list of rejects. Satisficing itself did not produce the best alternative. Using lexicography model, option 1 is selected because accreditation is the top-ranked attribute and Option 1 accreditation is the best of all.

6. Selection of the Preferred Alternative
Option 1 is the best alternative among the four other alternatives. The best alternative for Ali is going to University in option 1. It is selected using lexicography model and accreditation is the top-ranked attribute. If other options have same accreditation with option 1, then the choice would be made on the basis of passing grade.

7. Performance Monitoring and the Post Evaluation of Result
Besides non compensatory model, another model called compensatory can also be implemented to choose best alternative. To compare those two models, next blog posting will use application of compensatory model to choose best alternative from same case study.

References :
1. Sullivan, W. G., Wicks, E. M., & Koelling, C. P. (2012). Engineering Economy (15th ed.), Chapter 14.6; Prentica Hall.