Features
Decisor has many interesting functionalities, such as:
- Easy pairwise comparisons among criteria and alternatives
- Group Decision Making
- Different scales for judgements
- B.O.C.R. Analysis (Benefits-Opportunities-Costs-Risks)
- Easy export to CSV format
Suppose you have a model and several experts or stakeholders, each contributing with a particular set of judgement. How to take into account the amount of pairwise comparisons and integrate it to a single model?
Decisor will let modellers use Group Decision for such cases. The only restriction is that the model should be the same for all stakeholders. The process to use group decisions is as follows: one should use the File > Open Several menu option, and then select all models for analysis. Once a set of models are chosen, the software will iterate among the models and apply a geometric mean for all pairwise comparisons, yielding resulting values into a new model, readily available to the model on screen. Then, it is sufficient to save as a new model and analyze the output.
The tool support the use of multiple scales, beyond the original Judgement Scales originally conceived by T. L. Saaty in his seminal work.
For example, clicking in Settings > Set Scale a modal window surfaces (next figure) showing uses with some possibilities:
This scales are defined and explained in the following reference: Review of the main developments in the analytic hierarchy process, by Alessio Ishizaka & Ashraf Labib.
https://doi.org/10.1016/j.eswa.2011.04.143
Under Analysis > B.O.C.R. > New the user could create a Benefit-Opportunities-Costs-Risks analysis, for example:
The user enter values in the input fields according to specific analysis (for example, costs), and then, he/she should select a formula that will be used in conjunction with previously computed weight values, where possibilities are: a) Addititive(negative) b) Addititive(reciprocal) c) Multiplicative
The user could select only Risks, or only Opportunities to study at a given time. The formulas for each possibility are:
The process functions as follows: Decisor will take the input provided by the user, normalize it (it allows toggling the input value and its normalized value by clicking the button "Toggle normalized"), and then apply the formula by taking into account the weight vector computed for each alternative.
B.O.C.R. analysis is a powerful tool for complex decisions, since it uses advantagens (benefits, opportunities) and disadvantages (costs, risks) and present the user with a numerical value that could indicate a course of action or a direction to follow according to the computed values.
Next figure shows a simple Cost analysis using the Additive(reciprocal) formula for B.O.C.R., where the user has assigned some values to each alternative ($5000 for Tom, $10000 for Dick, and $7000 for Harry):
It should be mentioned that the list of criteria for the leader example did not consider costs, because this can be done in a B.O.C.R. analysis.
It is possible to see that the tool has used the b) formula and a new set of numerical values is presented, where the alternative "Tom" is better (e.g. yields the highest B.O.C.R. value) despite the fact that its weight vector value was not the highest one (it was the alternative "Dick" - the one with the highest rank if one disregards B.O.C.R. analysis altogether and choose to use only AHP analysis). Now, when costs was assigned, the better alternative has shifted towards another one, and it could potentially change the previous decision that was made.
The tool allow to export the model and results to a MS-Excel csv (Comma Separated Values) file. The user must click on File > Save Report and then the tool will ask for a file name and folder to create the file containing the results. After the user has chosen this options, the tool will create a new file and the contents will be just like (for the leader model, for example):
Objective:; Matrizes: ;Experience;Education;Charisma;Age;;WVector;;CR Experience;1;4;3;7;;0.539645;;0.0442351 Education;0.25;1;0.3333;3;;0.131462;; Charisma;0.3333;3;1;5;;0.271544;; Age;0.1429;0.3333;0.2;1;;0.0573493;; Experience;Tom;Dick;Harry;;WVector;;CR Tom;1;0.25;4;;0.539645;;0.035716 Dick;4;1;9;;0.131462;; Harry;0.25;0.1111;1;;0.271544;; Education;Tom;Dick;Harry;;WVector;;CR Tom;1;3;0.2;;0.539645;;0.0633846 Dick;0.3333;1;0.1429;;0.131462;; Harry;5;7;1;;0.271544;; Charisma;Tom;Dick;Harry;;WVector;;CR Tom;1;5;9;;0.539645;;0.0696169 Dick;0.2;1;4;;0.131462;; Harry;0.1111;0.25;1;;0.271544;; Age;Tom;Dick;Harry;;WVector;;CR Tom;1;0.3333;5;;0.539645;;0.0280232 Dick;3;1;9;;0.131462;; Harry;0.2;0.1111;1;;0.271544;; Results: Tom;0.359066 Dick;0.488313 Harry;0.152621 Lambda Max;3.02914
One, in close inspection, can see that a structured file was created using semicolons (';'), a file that is easily opened by MS-Excel (the user must use MS-Excel's features to break the semicolons into columns in a spreadsheet - Data to columns).
This ends our features section. Please, feel free to contact us and suggest new features for next releases (rczekster [at] gmail [dot] com).