General Information

What is it and how it works ...

Introduction

ESTIMATION & CHOICE is a Web based software used for multiattribute (multicriteria) decision analysis, where the decision problem is structured hierarchically from criteria to lower level subcriteria: Goal - General Parameters - Parameters. ESTIMATION & CHOICE handles the different decision technologies such as Analytic Hierarchy Process (AHP) by T.Saaty , Utility Functions, Simple weighing model, Pareto domination, Benefit-Cost analysis.

Project contains information about alternatives you want to analyze and Plans of Analysis.

Alternatives - the objects you want to analyze and compare. That might be for example cars, houses, employees, business plans etc. In the other words, alternatives are any objects that you can describe with a set of parameters. Within the Project you have just one set of parameters to describe your alternatives.

Parameters may be of three different types.

Type	Parameter	Value
Numeric	Fuel efficiency	3.4 L/100 km
Logical (Yes/No)	Air conditioning	Yes
Qualitative	Fuel Type	Gasoline

For the qualitative parameters there should be classifiers, attached to every parameter of this type. For example, a classifier "Types of Fuel" may be attached to the Parameter "Fuel type".

Classifier - list of possible values of the qualitative parameter. For example, classifier Fuel types may include Gasoline, Diesel, Natural Gas etc. The same classifier may be attached to different parameters. (Classifier Colors may be attached to parameters Car body color and Car interior color).

Plan of Analysis represents the overall goal and objectives of the analysis, includes hierarchical structure of parameters with their weights (=priorities) and Utility Functions.

Plan of Analysis has a hierarchical structure.
At the top level is the goal of the analysis (For Example: To compare Business Plans to choose the best one, To compare Companies on stock market to buy optimal set of shares, To compare Car models to buy family car etc...)

At the next level there are General Parameters influencing your Goal of the Analysis.

General Parameter - characteristic of some important general quality of the Alternative. (For example: Profitability, Risk for Business Plans; Safety for Cars)

Weight of the parameter is a number indicating how that parameter is important for you.

General Parameter: Safety

Parameter	Weight
Safety Bag	0.5
Side Impact System	0.3
Top Speed	0.2

The bigger weight - the more important this parameter. So in this example for the General Parameter Safety the most influential parameter is "Safety Bag".

To assign weights to the parameters you can use either direct weighing or pairwise comparisons

In direct weighing, the weights of Parameters are directly given - just type the number in the corresponding fields. For example, you can provide weights on the scale [1 to 10] or [1 to 100]. Afterwards, if you want, you can normalize the given weights: just click button "Normalize", than system recalculates the weights in the way that sum of all weights equals 1. To do it system summarizes all the weights and then devides every weight by that sum.

Pairwise comparisons (AHP - The Analytic Hierarchy Process by T.Saaty)
In AHP you are asked to compare evry possible pair of attributes. In each pair you must decide which attribute is more important and in what degree. These comparisons are saved in comparisons matrix and the weights of attributes are derived from the eigenvector of this matrix. So, you just have to answer several questions like "Which of these two parameters is more important?", and then the system will calculate and show you the weights.

You have not just prioritize the parameters, but also define what values of the parameter are more preferable. In the other words you have to assign a Utility Function to every parameter.

Utility Function maps the values of parameter on the scale between 0 and 1. Utility Function allows prioritize different values of the Parameter. The close Utility Function to 1 - the better value of Parameter. As you remember we have three different parameter types, and every type has its own Utility Function format.

1. Utility Function for Numeric parameters
You can assign a Utility function to the Numeric parameter using up to 5 points. For every point you have to insert value of the parameter and then select corresponding value of the Utility Function from drop-down list. To assign Utility function for numeric Parameter you have to describe at least two points.

For example, if you are a doctor and you have to estimate condition of a patient you can use parameter "Temperature" with the next Utility Function:

Point	Value (°C)	Utility Function	Description
Point 1	35.0	0.2	Mediocre
Point 2	36.6	1.0	Ideal
Point 3	37.5	0.3	Satisfactory
Point 4	42.0	0.0	Minimally Acceptable

Based on this informasion system calculates Utility Function for the others values of the parameter. (We presume that for all values less than minimal value in the table - less then 35.0 in our example - the Utility Function will be the same - 0.2. And similarly for all values more than maximum value the Utility Function will be the same.)

2. Utility Function for Logical (Yes/No) parameters
To assign a Utility Function to the logical parameter you just have to answer a question - if this feature desirable or not.
For example when you compare cars for the parameter "Air Conditioning" you can assign Utility Function "This feature is desirable". It means that the value of the Utility Function equals to 1 for air-conditioned cars, and equals to 0 for others.

3. Utility Function for Qualitative parameters
For qualitative parameters we have to assign value of the Utility Function to every value of the attached classifier. You can do it in the same way as you assign weights to the parameters. In other words, you can give the values directly or use pairwise comparisons.

Parameter "Car body color"

Value	Utility Function
Blue	0.4
Black	0.3
Yellow	0.2
Green	0.1

Group Model

ESTIMATION & CHOICE supports group decision making and provides tools to coordinate opinions of up to 10 experts working on the same project. Experts may be involved in finding coordinated weights for General Parameters, weights of parameters influencing General Parameter, finding values of Utility Function for Qualitative Parameters. Experts may also compare alternatives on the final stage of the Benefit - Cost Analysis.
The system shows you average weights for the group of experts and a Coefficient of Concordance. You can see as well dispersion and deviation for every attribute, and "Extravagancy index" for experts. So, you have very powerful tool to improve co-ordination in the group of experts and get the reliable results.

Tolerance to Incomplete Data

ESTIMATION & CHOICE runs analysis even if some values of the parameters are missing for some alternatives. If at least 70% of alternatives have the value of parameter, then for the rest alternatives missed values will be calculated as average.

Analyzing Results

Once you inserted information about alternatives and created at least one Plan of Analysis you can run the analysis! Just click pad "Alternatives and Analysis" of main menu, check the alternatives you want to analyze, select Plan from dropdown list and then click button "Analyze Alternatives". The bar chart shows the total ratings of alternatives. In the table below you can see comprehensive analytical information including final rangs and ratings of alternatives, results of the Pareto domination analysis. In case if some missed values have been restored as average you will see the corresponding warnings. You have this valuable analytical information for every alternative and every General Parameter.

Single Alternative Analysis
ESTIMATION & CHOICE provides you the unique option to analyze every single alternative separately. To run analysis for single alternative just click its name at the page with the results for the group of alternatives. You can as well check only one alternative for analysis on page "Alternatives and Analysis".
The system shows final rating of the alternative, list of General Parameters sorted by weight with their ratings and verbal interpretation of ratings (for better emotional perception).

You can see as well list of Parameters influencing current General Parameter with their values, ratings (derived from Utility Function) and verbal interpretation of ratings.

Making final Decision with Benefit/Cost Comparisons
Sometimes it might be not enough to create hierarchical structure of the Plan of Analysis, assign weights and Utility Functions and calculate ratings of alternatives. If your decision problem may be described in Benefit/Cost terms, it is more reliable to use Benefit/Cost Comparisons at the final stage of the analysis.

For example, you compare cars and you created Plan of Analysis including General Parameters: "Driving", "Comfort", "Safety", "Economy". Every General Parameter is influenced by number of parameters. So "Economy" may be influenced by "Price" and "Expenses".

To run Benefit/Cost Analysis you have to split up "Efficiency Indicators" and "Cost Factors". How to do it? Pretty simple! Just assign zero weight coefficient to the "Cost Factors" General Parameter. In our example it is "Economy". Than you will get ratings for this General Parameter, and at the same time exclude influence of the "Cost Factors" on the total ratings of the alternatives.
And then, after running analysis for group of alternatives, you can apply Benefit/Cost comparisons to the results of group analysis - just click button "Benefit/Cost analysis" at the page with group results.

You will be asked about axes for Benefit/Cost chart. Select "Final Rating of the Alternative" as Axis Y (Benefit) and Rating of General Parameter "Economy" for Axes X (Cost). And then simple Benefit/Cost analysis starts as a number of pairwise comparisons of alternatives at the chart. After completing the comparisons system shows the bar chart with final ratings of alternatives resulting from Benefit/Cost Analysis.