The Compensatory Model for decision making, widely applied in consumer decision making and in brand/product marketing (to consumers) decision making, is often defined in conjunction with the Noncompensatory Model for decision making.
Compensatory modeling is based on the premise that (1) alternative good attributes and/or (2) acceptable bad attributes can be traded off--or compensated with or by--each other within a given decision making situation. For example, if you want an apartment that has a backyard, a mountain view, a low rent cost, and a good neighborhood, you might be inclined to compensate a sea view for a mountain view if all other factors were conducive to a good match to your apartment attribute rankings.
Noncompensatory modeling is based on the premise that certain preference attributes are not compensatory by nature, that is, certain attributes cannot be traded off with--or compensated by--other attributes even if they are objectively considered "good" attributes. For example, say a noncompensatory attribute in your grocery shopping is that food must be organic. If your market is out of your preferred organic all-tomato, only-tomato marinara sauce, but they have another high quality "all natural" all-tomato, only-tomato marinara sauce on sale and an organic all-tomato, only tomato concentrated tomato paste in stock, though not on sale, your noncomensatory attribute requirement will prohibit compensating with the on-sale "all natural" sauce and require compensating with the higher priced organic tomato paste.
Methods of the Compensatory Model
The University Of Manchester, England, School of Management explains that the four Methods of the Compensatory Model are:
- Scoring Methods
- Compromising Methods
- Concordance Methods
- Evidential Reasoning Approach
Alternative attributes applicable to decision making are ranked according to utility/scores (utility, as used in microeconomic theory: total satisfaction from or specific preference for given goods and services). Ranking utility turns subjective "attribute values" into a "common preference scale" ranked across a range of differences, e.g., 0, 1, 2, 3, 4, 5. Some models within the Scoring Method are:
- Additive Model / Simple Additive Weighting Method
- Additive Difference Model (attribute-by-attribute utility/scoring)
- Analytical Hierarchy Process (pairwise attribute comparisons)
Pro: Attributes of alternatives can be weighted and ranked against each other; subjective "preference" is included.
Con: Limited in early decision making phases as it doesn't exclude alternatives or attributes.
Alternatives are judged and selected by proximity to, or their closeness to, the ideal solution and by their non-proximity to, or distance from the nadir solution (nadir, nadir solution: the furthest, the lowest; the solution furthest from the ideal or the lowest that is within the realm of acceptability). A model within this Method is the Technique for Order Preference by Similarity to Ideal Solution, which "first normalises the decision matrix" then evaluates alternatives based upon closeness to the ideal and distance from the nadir.
Pro: Attains proximity, amongst all alternatives, to the ideal solution.
Con: Excludes consideration of noncompensatory attributes
[concordance: agreement with, harmony with a standard]
Attributes of decision alternatives are ranked by satisfaction as compared to a concordance standard, i.e., the standard to which the decision must be in concordance. The alternative that is deemed acceptable is the one that ranks highest overall on the concordance standard. The Linear Assignment Model is an example of a concordance method.
Pro: Identifies the alternative approximating an ideal situation while avoiding an unacceptable situation; avoids dominance of single subjective, preferential attributes.
Con: Precludes noncompensatory attributes unless these are weighted to offset the combined totals of all other attributes.
Evidential Reasoning Approach
An expanded decision matrix is used in which attributes of alternatives are measured, or "described," by "a distributed assessment using a belief structure," such as "Excellent" "Good" "Average" "Poor" "Worst," with evaluations on percentages matching belief structure, e.g., "Excellent 60% belief degree."
Pro: Objectifies "belief structure."
Con: Eliminates exclusion based upon noncompensatory attributes.