Looking for:
Expert choice manual pdf free download |
Expert choice manual pdf free download - Document Information
Resource Aligner ability to optimize resources and allocate funds budgets with varying constraints, groups, funding pools and risks. Network capabilities for group decision-making including Expert Choices Decision Portal using a browser. Resource Allocation a way to select a combination of alternatives that maximize optimize the attainment of objectives while not exceeding a specified budget available as an Expert Choice add on.
Group Expert Choice Expert Choice helps participants of a group define objectives, sub-objectives and alternatives and then organize them into a structured hierarchical tree. Once participants compare and 5. Then the team can examine how changing the priorities of any objective affects the outcome by manipulating one of four dynamic sensitivity graphs. By keeping the participants focused on the problem, Expert Choice contributes to more efficient use of meeting time.
About the Internet and Intranet Features In addition to the features described above Expert Choice is a decision-support software tool designed to help groups enhance the quality of their decisions by bringing structure to the decision making process. It enables group members any place in the world to solve decisions together through the Internet using their own network.
Expert Choices Decision Portal provides web capability and works as a virtual meeting place allowing for each participant's input to be officially heard and included in the final decision. This not only reinforces buy-in from all members but also saves on travel time and costs. For more information about this feature contact Expert Choice. About the Keypads If Expert Choice is keypad enabled you can create a group decision room where members of your team can brainstorm and answer questionnaires.
Then the results can be structured into a decision hierarchy. Using the electronic keypads, up to people can make judgments about the objectives or sub-objectives and alternatives of a decision problem.
Like the Internet version, the keypad version keeps the participants focused on the problem, while contributing to more efficient use of meeting time. What Expert Choice is used for: Expert Choice is designed to help individuals and groups in the analysis, synthesis and justification of complex decisions and evaluations. Thousands of businesses and government agencies worldwide use Expert Choice for a variety of applications including: y. Insert the CD into CD drive. If the CD does not automatically start then from Windows:.
Follow the instructions on the screen. When prompted, enter your name, company name and Expert Choice serial number.
Note: If you do not have a serial number or do not enter a valid serial number, then only the trial version features will be enabled. Dependent upon your computer policies you may need administrative privileges to install Expert Choice. Alternatively, if an Expert Choice icon has been installed on your desk-top you can click the icon. Pane is a portion of a window.
A window can have several panes that change as different portions of Expert Choice are accessed. Mouse Conventions If you have a multiple-button mouse, the left mouse button is the primary mouse button, unless you have configured it differently. Any procedure that requires you to click the secondary button will say, "right-click.
Press the mouse button once on the menu bar item. Move the cursor to the specified item within the sub-menu and press the mouse button again or slide the cursor to the specified item and release the mouse button. For example: Select File, then Open. Then move the mouse to the word File on the menu bar and click once. Move the mouse to the word Open on the drop-down menu and click once.
Press and hold the left-mouse button on the indicated item as you move the mouse to the designated location, dragging the item with the mouse pointer, and then release the mouse button on the desired destination. When you are required to enter data, we use the word "Type", followed by what is to be entered in bold print; for example: Type Performance.
Click and drag the arrows on a scroll bar, located at the right-side or bottom of a window, to scroll within a window. For example, you can press Alt plus F to select the Files command from the menu bar.
This will drop the file menu down and you can then select the item you want by again pressing one of the underlined keys. Context sensitive help is available by pressing function key F1. Hot Keys Most menu items have hot keys that are underlined and can be invoked by pressing the Alt key in combination with the hot key. The hot keys may not be underlined in Windows until the Alt key is pressed. Shortcut keys exist for some, but not all of the menu commands.
You can see the shortcuts when a drop-down menu is selected using the keyboard or the mouse. In this lesson you will learn how to insert objectives, sub-objectives and alternatives into the decision model, make judgments paired comparisons , synthesize to get your results and view sensitivity analyses. Building a hierarchy is as much an art as it is a science. Following are guidelines that you need to keep in mind before beginning to build any model.
By keeping these two simple guidelines in mind, you will be able to model complex decisions correctly and efficiently.
Guideline 1: Try not to include more than nine elements in any cluster because experiments have shown that it is cognitively challenging for human beings to deal with more than nine factors at one time and this can result in less accurate priorities. Guideline 2: Try to cluster elements so that they include elements that are "comparable", or do not differ by orders of magnitude.
In other words, try not to include items of very small significance in the same cluster as items of greater significance. The purpose of a hierarchy is to cluster the more important elements with each other and the less important elements with each other. Differentiating Between Objectives and Criteria In practice, the words objectives and criteria are used interchangeably.
An objective is something that is sought after or aimed at, whereas, a criterion is a principle or a standard that an idea or objective is judged by. When dealing with choice models, we will use the word objectives rather than criteria because it better describes the purpose of the evaluation to select the alternative that best meets an individuals or groups objectives.
Figure 1 shows our ModelView window consisting of three panes that contains our completed model. If you make the same judgments that we suggest during this lesson then your model will look like this. TreeView pane displays the hierarchical listing of the objectives and sub-objectives.
This is where you will enter the objectives and sub-objectives of your decision. Each element in the TreeView is also referred to as a Node.
Alternatives pane displays the active alternatives. Alternatives can be added to your model from this pane or from the Data Grid the latter to be explained in Lesson 3. Information document pane displays information about the highlighted objective or alternative in either the TreeView or Alternatives pane. To create a model y. Then a blank ModelView window consisting of three panes will appear with a Welcome to Expert Choice dialogue box superimposed over the ModelView.
Tip: If you have already created an icon for Expert Choice on your desktop, you may double click the icon to start. Note: To open a recently opened model, select Recent; then select a model from the drop-down list or select Browse existing files. A dialogue box will appear that prompts you to type a file name. Type Laptop; then click Open.
Another dialogue box will appear prompting you to "Enter a description for your goal. Notice your model has no elements below the goal; this is indicated by the gray rectangle next to your goal description.
Now you will enter the objectives and subobjectives of your decision model. Type your first objective, Performance: a laptop that is fast and holds a lot of data, then press Enter. Notice that the gray rectangle next to the Goal changed to yellow after you have entered the first objective; this means that there are items below the goal. Tip: A quick way to insert an objective is to right-click on the Goal and select Insert Child of Current Node from the drop down menu.
Once you have entered an objective, you may right-click on it and select Insert Sibling of Current Node to enter other objectives on the same level. Type your second objective, Reliability: a laptop that is durable, then press Enter. Notice when you add the second objective the yellow rectangle, to the left of the goal node, now includes a red rectangle in its center; this means that the objectives below the goal have not been assessed. We will now continue entering in our objectives. Notice that we briefly define our objectives to help clarify what we are trying to accomplish.
Type your fourth objective, Price: affordable with some bells and whistles; then press Enter. Renaming Nodes and Deleting Nodes If you make mistakes while entering nodes or feel that you have added too many, move to the node that you would like to either be renamed or deleted and do either of the following: y. Once items are in the trash can they appear in a Trash window and can be dragged and dropped back into the hierarchy by positioning the cursor over the node that will receive them.
Adding Sub-Objectives in the TreeView In this tutorial, we will enter sub-objectives for only one of the top-level objectives. In a "real" decision model you might add sub-objectives for most top-level objectives. To illustrate how to do this, we use our objective: Performance: a laptop that is fast and can hold a lot of data. To enter the first sub-objective under an objective: y. Saving Your Model Since some of your model is stored in memory versus the database, it is important to continuously save your model additions or modifications; to do so: y.
Select File; then Save. Alternatively, you can click the save button on the button bar or use the shortcut keys Ctrl S. Tip: If you terminate your Expert Choice session now by selecting File, Exit you can easily retrieve the most recently accessed models by selecting one of the entries in the bottom portion of the File, Open menu dropdown list. To move nodes follow the steps below.
Press and hold the left mouse button while positioned on the node objective or its description that you wish to move elsewhere in the hierarchy. Drag that node over another node in the hierarchy where you would like it to appear, until the receiving node is highlighted; then release the mouse. The node will become the first child under the node on which it is dropped.
During the move a white notepad will be dragged from one portion of the hierarchy to another. Then a list of options will appear; select one. If you want the nodes of a cluster to appear in alphabetical order, select the parent of the cluster and then select Edit, Sort Cluster.
To rearrange nodes within a cluster, select each node within the cluster in reverse order of the way you want them to appear in the cluster and drop each in turn on the parent node of the cluster. Drag the node to the copy location, as you do this, a notepad will be displayed. Then all the nodes below head of the dragged Plex will be copied below the destination node.
Reverting to an Older Version of your Model We shall assume that you practiced moving and copying nodes in the hierarchy. To put them back to their original order: y. From the displayed Revert to box, double-click on the option "after save". This will restore your model to its form right after you executed the File, Save command in this lesson.
Should you revert to the wrong option, select Edit, Revert again and choose another option from the box. Note: Revert is only available during the current session or if there is an abnormal termination due to causes such as power interruptions.
Adding Entering Alternatives in the Alternatives Pane To enter the alternatives into your decision model, do the following: y. Click the Add Alternative button in the right corner of the Alternative pane. Then a dialogue box will appear which prompts you to enter an "Alternative name". Note: For legal reasons, we use the names of mathematicians in lieu of computer manufacturers. Tip: When you see the phrase "Click OK", you have the option of doing just that or pressing the Enter key.
Next you will learn how to make judgments or pairwise comparisons about the objectives and alternatives in your model. Once your model is built, the next step is to evaluate the elements by making pairwise comparisons.
A pairwise comparison is the process of comparing the relative importance, preference, or likelihood of two elements objectives with respect to another element the goal in the level above. You will make pairwise comparisons throughout this model to establish priorities. There are three pairwise comparison assessment modes. Equal means the two items being compared are of equal importance to you. Extreme means an order of magnitude about 9 or 10 to 1.
Judgments between these words, such as Moderate to Strong are also possible. Numerical judgments are made using a nine-point scale, represent how many times one element is more important than another. Importance is most appropriate when comparing objectives or criteria. Preference is appropriate when comparing alternatives with respect to their covering objective. Likelihood is appropriate when comparing the likelihood of uncertain events or scenarios, as in risk analysis.
To change the comparison type: y. Select Assessment, Set Mode and then select one of the three options in the type section: Importance, Preference or Likelihood. Making Verbal Judgments for the Objectives Overview Since judgments about the relative importance of the objectives may depend on the alternatives being considered, we recommend that judgments be made from the "bottom up", that is first for the alternatives with respect to the sub-objectives, then for the sub-objectives with respect to the objectives, and then for the objectives with respect to the goal.
However, for illustrative purposes, we will make judgments "top down" in this tutorial. Then from the menu, select Assessment, Pairwise, or simply click the Verbal tab. The Verbal pairwise window will appear. Figure 2 shows what the Verbal comparison window will look like if you make the judgments described below.
You should use the word that most accurately represents your feelings. The upper portion of the Verbal Pairwise window shows two elements being compared with respect to their parent. On the right side of the Verbal Pairwise screen is the verbal judgment indicator.
By clicking and dragging the indicator, either up or down, you can select the judgment that best describes your feelings. The lower portion of the window shows numerical representations of the verbal judgments; the current judgment is highlighted.
Note for example, a judgment of Strong, while represented by 5, does not necessarily mean that one element is 5 times another since the priorities are derived from the set of judgments, rather than one judgment alone. Now you will make judgments with respect to the objectives. Now make your first judgment comparing Performance to Reliability with respect to your goal. We are saying that Performance is Moderately more important to you than the computer's Reliability. As you make your first comparison, by clicking Moderate, notice that a number representing your verbal judgment is entered in the matrix and you are automatically taken to the next paired comparison to be made.
In this lesson the Autoadvance option is on. It automatically advances you to the next paired comparison. Tip: In our example, the objective on the top is more important than the objective on the bottom.
If the objective on the bottom is more important than the top, select the word from the lower portion of the scale. Now make the remaining five judgments by referring to Table 1. When finished, read the additional information about entering verbal judgments before proceeding. At any time you may alter your judgments by clicking on whichever judgment box in the matrix that you would like to change; then make your change using the verbal scale. There are several ways judgments can be entered with the mouse when using the Verbal comparison mode; for further information refer to context sensitive on-line help by pressing F1 from the Verbal comparison window.
One way to make Verbal judgments is to type the numerical representation of the judgment in a matrix cell. If you want to enter a judgment between two words such as "Very Strong and If you do this, 8 is merely a numerical representation of the wording "Very Strong to Extreme".
Table 2 Numerical Representations of Verbal Judgments When the column element in the matrix objective 2 is preferred to the row element objective 1 first move to the matrix cell and then invert the judgment by pressing Alt I or clicking the Invert button y. After you enter your last judgment, a dialogue box will appear asking if you want to "Record Judgments and Calculate".
Select No, and you will remain at the Verbal window where you may change judgments. See Figure 3. If you had selected Yes, the priorities for the objectives would have been calculated and you will be returned to the ModelView window without displaying the priority graph. Tip: When entering judgments, you may click the calculate button as long as it is colored yellow or green.
The calculate button will be: Red if not enough judgments are made to calculate priorities; Green if all judgments have been made; or Yellow if you have missing judgments but there are a sufficient number to calculate priorities. If the Normalize box is checked the highest priority will be shown as a value of 1. Inconsistency The inconsistency measure is useful for identifying possible errors in judgments as well as actual inconsistencies in the judgments themselves; this is accessed from the "Priorities with respect to" window.
Inconsistency measures the logical inconsistency of your judgments. For example, if you were to say that A is more important than B and B is more important than C and then say that C is more important than A you are not being consistent. A somewhat less inconsistent situation would arise if you would say that A is 3 times more important than B, B is 2 times more important than C, and that C is 8 times more important than A.
In general, the inconsistency ratio should be less than 0. The "Priorities with respect to" window also shows how many missing judgments are in the set of elements being compared.
This can be done from any comparison mode. To see how Expert Choice helps identify inconsistencies, do the following from the Verbal comparison window: y. Notice we entered a judgment of Moderate that is represented by 3. Your Best Fit might be different if you did not make the same judgments. You should only change an inconsistent judgment if you feel that your initial comparison was in error and did not truly represent your feeling. Note: If the Best Fit is displayed in red then the column element of the matrix is preferred to the row element.
To enter an inverted element in the matrix, enter your judgment and if it is not displayed in red then click the Invert button. You can also view the second, third etc. Then, if desired, you can alter the judgment. Inconsistencies can also be suggested when you have missing judgments. Tip: To view the "best fit" of any cell right-click in a selected matrix cell. If you made any changes to your model while you were learning how to use the Inconsistency command you might be prompted to record your judgments.
Since you were experimenting, just select No. The Graphical pairwise window will appear; refer to Figure 5 when you make your comparison. In this example, there is only one judgment to be made. The Graphical mode operates by clicking and dragging either comparison bar the blue or red bar either to the left or to the right. Tip: By dragging the top bar to the right you are saying that the element on top is more important than the element on the bottom, in proportion to the relative lengths of the bars.
By dragging the bottom bar to the right you are saying that the element on the bottom is more important than the element above. The numerical representation of the judgment is entered into the cell matrix; if the number is in red it indicates that the column element is preferred to the row element.
When done dragging a bar you are prompted "Record Judgments and Calculate? Now we will illustrate how to make numerical comparisons.
From the TreeView, click the sub-objective, Speed Click the Numerical tab A window will appear where you will compare your preference of the alternative computers with respect to Speed. When finished, a dialogue box will appear, prompting you to Record and Calculate Judgments.
Figure 6 shows what we entered. If you click on the number above the slider bar you will autoadvance to the next paired comparison. Alternatively, you can drag the indicator bar to make the next judgment. Then click the next empty cell in the matrix. Refer to on-line help to see other ways to enter judgments. After all comparisons have been made, you will be prompted to record your judgments and return to the ModelView.
Select one of the modes of comparison of your choice by clicking on a tab Verbal, Graphical or Numerical. Tip: Remember to save your model often; then if you want to go back to an earlier version you can use the Edit Revert command. You can repeat the above steps until all remaining three comparisons are made, or alternatively, you can: o. Select one of the modes of comparison of your choice - Verbal, Graphical or Numerical. Repeat the above process until all judgments have been made for both the last two objectives.
Once All Judgments are Made and Priorities are Derived Once priorities have been derived, the rectangles in the TreeView are replaced by Priority Fills; they graphically portray the derived priorities Figure 7. These fills indicate the amount of importance of each objective or sub-objective. Figure 7 shows the derived priorities with respect to Speed. To see the overall results, select the Goal. Figure 7 Hierarchy Showing Priority Fills You can also use the View command to switch between assessed indicators yellow and gray rectangles and priority fills.
Each command redraws the TreeView hierarchy around the current node. Other redraw commands are found on the View Menu; they expand and contract nodes and display only the first level of the hierarchy. The redraw commands are useful when viewing large models; you can move to a node in the hierarchy and view what is relevant without losing your position.
When at the Data Grid you can click a node in the hierarchy and you will automatically be positioned to its corresponding covering objective on the grid.
See lesson 3. When you move to a new node in the hierarchy, Autoredraw will automatically expand the current node displaying its descendants while collapsing other branch nodes of the hierarchy. This feature uses Expert Choice's patent 4,, Autoredraw is represented on the button bar.
Redraw Redraw, is only available when Autoredraw is off. It manually redraws the hierarchy about the current node and does not collapse other nodes.
It is represented by. Tip: When at the ModelView's TreeView, depressing the first letter of a node name or a series of characters will take you to that node. Making an Objective or Alternative Dormant in the Pairwise Comparison Matrix Before we tell you how to synthesize to obtain the results, we want to explain how to make an element objective or alternative dormant so it can be excluded from being prioritized.
Alternatively, if an alternative has already been prioritized it can be excluded from the results. From any pairwise comparison window: y.
Press Ctrl and click on an elements name any objective or alternative located in the rows of the matrix. The row for the element that you selected will be blocked.
In addition, any comparison made or to be made relating to that factor will also be blocked. Figure 8 shows this. Now press Ctrl and click on the dormant element to make it non-dormant; any judgments previously made will be shown. Synthesis - Getting the Results A synthesis can be done for either the entire model or a portion of your model. The synthesis window will appear, showing the results see Figure 9. Notice the Ideal and Distributive buttons; in our example, the Ideal mode is selected.
Each mode will be explained. The other alternatives receive weights under each covering objective proportionate to their priority relative to the best alternative under each covering objective. When using the ideal synthesis mode, the addition or removal of alternatives that are not best on any covering objective will NOT impact the relative priorities ratios or ranks of other alternatives. The ideal mode should be used when selecting one alternative from many and when the priorities of the alternatives not selected are not of interest.
Distributive Synthesis The Distributive mode distributes the weight of each covering objective to the alternatives in direct proportion to the alternative priorities under each covering objective. When using the distributive synthesis mode, the addition or removal of an alternative results in a re-adjustment of the priorities of the other alternatives such that their ratios and ranks can change.
The distributive mode should be used when measuring under conditions of scarcity for example when forecasting outcomes whose probabilities must add to 1, or when looking at elections where votes cast for one candidate may alter the rank of another candidate.
Because each synthesis mode combines priorities differently, you will notice that each mode may yield different, although normally very similar, results. For more information about the synthesis modes see the main on-line Synthesis section, Selecting the Synthesis Type. Synthesis Commands There are two synthesis tabs Summary and Details. Summary displays bar graphs of the alternatives' priorities with respect to the node from which you performed the synthesis in our example, the Goal.
See Figure 9 above. Details shows a grid of all the priorities and alternatives with respect to the goal or objective selected. In our example, since we selected the Goal node, we show all objectives, subobjectives and alternatives; see Figure Experiment with the other detail options and refer to the on-line help for more information. Sensitivity Analyses - Graphs Sensitivity analyses from the Goal node will show the sensitivity of the alternatives with respect to all the objectives below the goal.
It can also be performed from the nodes under the goal if the model has more than three levels to show the sensitivity of the alternatives with respect to an objective or sub-objective.
When performing a sensitivity analysis you may vary the priorities of the objectives and observe how the priorities of the alternatives would change.
There are five types of sensitivity analysis. You can open four types of sensitivity analyses at once or each one separately. Each graph has its own unique menu commands and each sensitivity analysis can be compared to a "what-if" analysis because the results are temporary. The remainder of this section explains each graph. By dragging the objectives priorities back and forth in the left column, the priorities of the alternatives will change in the right column. If a decision-maker thinks an objective might be more or less important than originally indicated, the decision-maker can drag that objective's bar to the right or left to increase or decrease the objectives priority and see the impact on alternatives.
Figure 11 shows a Dynamic Sensitivity graph. Experiment changing the priorities of your objectives by clicking and dragging the objective bars on the graph. Drag one objective bar back and forth and you will see the alternatives' priorities change. Also try this for the remaining objectives. To see the original priorities, click the Home button on the button bar. To view a dynamic sensitivity graph, Figure 12, showing the alternative's bars in multiple sections: y.
Sensitivity Command Bar While viewing any sensitivity analysis, you can hold your mouse over each button under the menu bar to show you what it represents. Each graph has different buttons as well as different menus. Note: You can have multiple graphs open at the same time.
However, if desired, you can close all graphs by clicking the X button. Click the appropriate button on the Dynamic graph button bar. The Performance Sensitivity analysis, displayed in Figure 13, shows how the alternatives were prioritized relative to other alternatives with respect to each objective as well as overall.
To see how the best computer compared to the second and third computers, read the overall priority from the intersection of "right y-axis" and the overall priority for each alternative. In our example, DaVinci is approximately. Note the priorities for the alternatives sum to one. To read each objective's priority based on the decision-makers paired comparisons , use the left y-axis.
For example Performance is about. Customer Service is about. To read the alternative priorities with respect to each objective, read from the "right y-axis". In our example, using Reliability, DaVinci has priority of approximately. This graph is also dynamic, so you can temporarily alter the relationship between the alternatives and their objectives.
Gradient Sensitivity Graph Make the Gradient Sensitivity graph the active window by clicking the appropriate button; see Figure This graph shows the alternatives' priorities with respect to one objective at a time. By clicking the menu command " X Axis", you have the ability to select which objective appears on the x-axis. The red vertical line indicates the objective's priority based on the decisionmakers paired comparisons. To indicate where an objective's priority changes drag the red bar to either the left or right; this is shown as a blue dashed vertical line.
In this figure you can see that increasing the priority of Price from. When viewing a Gradient graph you are looking for cross-over points of the alternatives.
Figure 15 shows how two alternatives compared to one another against the objectives in a decision. One alternative is listed on the left side of the graph and the other is listed on the right. The alternative on the left is fixed while the alternative on the right can be varied, by selecting a different tab on the graph.
Down the middle of the graph are listed the objectives in the decision. If the left-hand alternative is preferred to the right-hand alternative with respect to an objective, a horizontal bar is displayed towards the left. If the right-hand alternative is better, the horizontal bar will be on the right.
If the two choices are equal, no bar is displayed. The The overall priority can either be shown based on the objective weights typical or un-weighted. This graph Figure 16 shows the alternatives' priorities with respect to two objectives at a time.
By clicking the menu commands "X Axis" and "Y Axis" you have the ability to change the objectives being displayed. The area of the 2D plot is divided into quadrants. The most favorable alternatives with respect to the objectives on the two axes will be shown in the upper right quadrant the closer to the upper right corner, the better the alternative. The least favorable alternatives will be shown in the lower left quadrant the closer to the lower left corner, the less favorable the alternative.
Alternatives located in the upper left and lower right quadrants indicate key tradeoffs where there is conflict between the two selected objectives. Now save your model by selecting File, Save. Alternatively, click the save button from the button bar or use the shortcut keys Ctrl S. Unlike the way we built our model in Lesson 1, you can think in terms of clustering ideas that are a result of brainstorming, discussion, research and the solicitation of views and ideas from others.
There are two approaches to model building in Structuring. Top-down structuring Affinity Diagramming is better suited for decisions of a strategic nature where the objectives are better understood or known than the alternatives.
Top-level objectives are identified followed by the identification of sub-objectives. Bottom-up structuring is best suited for situations where the alternatives are better understood than the objectives. The pros and cons of the alternatives are used to help identify the objectives that can then be clustered into groups. Theoretical Foundation of Structuring When using Structuring individually or with a group, top-down or bottom-up structuring provides a framework whereby decision-makers use their knowledge of the problem to identify and organize the factors objectives and alternatives of the decision.
With a group the facilitator tracks the discussion, entering objectives and alternatives, or pros and cons of the alternatives. Then the facilitator clusters the concepts as the participants direct. These elements are then assembled into a hierarchical model consisting of objectives, subobjectives and alternatives. Using Structuring, decision-makers will list positive and negative aspects of each alternative in the decision model. The group of pros and cons is a pool of ideas that helps identify the important objectives that need to be met in order to resolve the issue.
Since objectives encompass these ideas, the pros and cons will be used for creating the objectives. The conversion of pros and cons into objectives means organizing the pros and cons of each of the alternatives in relation to your main goal or objective. The pros and cons are evaluated for the alternatives and a list of objectives is developed. Often, you will then continue to cluster the objectives into additional levels of sub-objectives. Structuring provides an intuitive interface to help you create a hierarchical decision model.
This interface includes the following familiar concepts for model building: pros and cons, objectives, and sub-objectives. Pros and cons are translated into objectives and subobjectives. These are actually the same as criteria and sub-criteria. Generally in the Analytic Hierarchy Process objectives are synonymous with criteria, and sub-objectives are synonymous with sub-criteria. You, the decision-maker, are the one who turns pros and cons into objectives. Thus, it is important to remember that converting pros and cons to objectives is a matter of thinking about which objective s the pros or cons point to.
More specifically, using pros and cons to derive objectives means you must re-conceptualize the pros and cons so that they are 1 described in Making this conversion is usually easier than it may appear.
For example, the pros and cons of being inexpensive or high priced in deciding which car to buy would be generalized into the objective of saving money. Another way to think about converting pros and cons into objectives is to consider how the pros and cons of an alternative can serve as indicators of what you actually want, which can then be expressed as an objective. There are two Structuring approaches to model building, bottom-up and top-down.
The terms bottom-up and top-down are not based on AHP but Expert Choice users should not interpret these terms rigidly for two reasons: 1 in bottom-up structuring, regardless of how hard you try to stick with a strict bottom-up approach, you still have to start with some kind of goal in mind and the goal is always represented by the top level of the decision model; and 2 with regard to top-down modeling, Affinity Diagramming provides for a freestyle, interactive "brainstorming" of the various aspects of your decision situation which is a natural, intuitive approach to building a decision model, where identifying parts of the problem is a matter of inspiration, intuition and common sense more often than a formal stepwise process, working from top to bottom.
Bottom-up structuring means that you are starting construction of your decision model with the alternatives at the bottom and working "up" towards your goal. Top-down structuring Affinity Diagramming means, essentially, that you are starting by listing and organizing your objectives. The instructions for top-down structuring emphasize the freestyle interactive approach to top-down structuring because 1 discussions of induction and deduction applied to AHP model building are beyond the scope of this manual; and 2 if you know how to use formal applications of inductive and deductive logic, it is straightforward to implement them in topdown structuring once you are skilled in using Expert Choice if, for some reason, you think you must.
Emphasizing a freestyle, interactive approach to top-down structuring is the best way to lead you into becoming skilled, efficient, and comfortable with the Structuring. Building a Model Using Top-down Structuring - Affinity Diagramming Top-down structuring or Affinity Diagramming provides a simple format for establishing objectives and sub-objectives for evaluating alternatives. From the Welcome dialogue box, select Structuring; alternatively, from the menu select File, New.
In the dialogue box type the new file name, Vehicle; then click Open or press Enter to continue. Type the Goal Description To purchase the most preferred vehicle. The TreeView pane now contains your goal. Additional buttons appear in the right corner, over both the Affinity Diagramming and Alternatives panes. When a dialogue box appears, type an objective; for this lesson, type Style. A highlighted box will appear in the Affinity Diagramming pane containing your objective. The red dot indicates where the next objective will be placed.
Tip: To position an existing objective drag it where desired. To position the placement of a new objective, click anywhere in the Affinity Diagramming pane; drag a box with the mouse. Next enter the objective. Now use the list below to enter the remaining objectives, repeating the same process from above. Select Close. Now you see the entire objectives names; that is if you typed what we suggested. Clustering Ideas Once you are satisfied with the list, it is time to cluster your objectives by identifying those that belong in the same cluster and then repositioning dragging the objectives into that cluster.
An objective that does not fit into any cluster may standalone. If you have several objectives that do not fit into any existing cluster but belong to the same cluster, simply add another objective into which these objectives can be clustered.
Follow the steps below to cluster the objectives entered in the Affinity Diagramming pane. Drag Engine Performance to the right of the list of objectives by pressing and holding the left-mouse button. Release the mouse when done. Drag and position Horsepower over Engine Performance. Once positioned, release the mouse button. Now Horsepower is clustered in the Engine Performance box.
Cluster the remaining objectives listed below, using the same steps just described. Refer to Figure 2 below to see what our clusters look like. Tip: Another way to enter an objective is to click anywhere in the Affinity Diagramming pane and drag the mouse until a rectangular shaped box is drawn, simultaneously, a dialogue box will appear, prompting you to enter your objective. After clustering the objectives you are ready to enter alternatives; note Figure 2 shows the alternatives that you will add.
Tip: You can also cluster objectives by dragging a box around one or more existing objectives. When the mouse button is released, you will be prompted for the name of the new clustering objective. You may need to drag and arrange the objectives to an empty portion of the screen before doing this.
Entering Alternatives Now that the objectives and sub-objectives are clustered it is time to enter alternatives. A dialogue box will appear prompting you to enter an "Alternative name"; type Clydesdale.
Rather than use actual car names we use the names of horses; feel free to use your preferred vehicle names. Repeat the Add Alternative process to enter the remaining two alternatives: Arabian and Palomino.
Tip: To enter an alternative, you can alternatively right-click in the alternatives pane and then type an alternative name in the dialogue box. From this point on you would continue to evaluate your model by making paired comparisons as described in Lesson 1, or you could evaluate your model based on what you will learn in Lesson 3, using the Data Grid.
Now we want to show you how to draw a box around several objectives to make a new toplevel objective. Grouping Nodes in the Hierarchy If you already have your hierarchy structured and wish to create a forgotten top-level objective you can do the following: y. This is a good approach when more is known about your alternatives of choice than your objectives.
To use our patented bottom-up structuring approach, you will first, enter alternatives; second, enter the pros and cons of each alternative; and third, convert the pros or cons to objectives. To learn more about our patent, go to www. Enter your alternatives Clydesdale, Arabian and Palomino , as described in the previous section, Top-down structuring Affinity Diagramming. Figure 4 shows our pros and cons for the highlighted alternative.
Now you will enter pros and cons for the first alternative. With Clydesdale selected, click the Add Pro button. A dialogue box will appear labeled "Add Pro for Clydesdale". If necessary, drag the pro under the Pro button.
Tip: To align the entry of a future pro or con in the pane, click with the mouse where you would like the pro or con to appear before selecting the appropriate button. Now enter a con for the Clydesdale. Select the Add Con button.
When the dialogue box appears, type Very Expensive and click OK Notice that the cons are displayed in magenta while the pros are in blue. Continue adding Pros and Cons for the remaining two alternatives; for brevity we will only enter one pro and one con for each alternative. Next you will convert the pros and cons on the list to objectives using our patented process. Lets begin with the con, Very Expensive, which is associated with the Clydesdale.
Drag and drop it on the Goal in the TreeView. When prompted to re-define the con as an objective, type Low Cost; click OK. You should now be able to see the beginning of the formation of your hierarchy. Notice how the con, Very Expensive, is now grayed out; this indicates that it was converted to an objective.
Gray can be removed from a pro or con by double clicking. Tip: Since alternatives are evaluated based on their preference with respect to objectives, the wording for cons must almost always be changed to state the objective or objectives that the con points to. For example, expensive points to low cost as an objective.
Looking at the list of pros and cons, you can see that some pros and cons could be repeated more than once because they are associated with more than one alternative; if this is the case, you would not need to convert repeats. Those pros and cons not used in the conversion process will remain white.
Tip: On the other hand, a single pro or con may point to several objectives and can be dragged and dropped numerous times. For example, a pro of size for a large car may point to the following objectives: Comfort; Carrying capacity; Safety; Fuel Economy; Ease of Parking, etc. Converting the Remaining Pros and Cons to Objectives Refer to Figure 6 and use it as a reference, practicing converting pros or cons to objectives or sub-objectives.
This is what we did. First we dragged relevant pros and cons to the hierarchy redefining them. When we were done we had a list of objectives.
Second, we clicked the Affinity Diagramming button. We could have added objectives directly in the TreeView by selecting Edit, Insert either Child or Sibling of the current node.
Then we could have dragged existing objectives beneath the new objectives. Both choices produce the same results. Expert Choice is a flexible decision support system designed to work the way you think. Now you are ready to make paired comparisons as outlined in Lesson 1 or use the Data Grid that you will learn in Lesson 3. Before this lesson ends, we want to describe three more features that are available: Information Documents and Notes and Object Linking and Embedding.
Information Documents and Notes Information Documents, referenced briefly in Lesson 1, are displayed in the lower right-hand side of the ModelView. They typically contain text, entered by you, or a facilitator in the group enabled model, which may describe the goal, give additional information as to why particular objectives or sub-objectives were selected, and tell us how paired comparisons were made. Information Documents are rich text objects and can include files such as Microsoft Office Files Word, Powerpoint, Excel, Access , as well as audio, pictures and video.
Information documents can be created for each objective, alternative, paired comparison in the hierarchy or each Data Grid assessment. Information documents can also be created and accessed from other windows. Notes In addition to Information Documents, Notes can be entered for objectives in the hierarchy, alternatives, individual paired comparisons, and Data Grid assessments as well as cells in the Data Grid discussed in Lesson 3.
When working with group models, each participant can enter their own notes to express their views, rationale, concerns and the like about the different parts of the decision process. To create a note from the current highlighted node, click the Note button. Enter your comments and when done, click the Note button to close.
If a note exists for the current node, the bottom portion of the Note button will be colored. Like Information documents, Notes are available from other windows. To learn more about these features, refer to Expert Choices Contents and Index. This very powerful feature enables you to document the rationale of your decision process as well as execute other programs. The process to link or embed files is dependent on your operating system; with certain operating systems you may only be able to embed.
This is what we recommend. With an open Information Document or Note, position the cursor to the insertion point within the Information Document or Note.
For Windows and other versions: Drag a rich text file icon such as. Then a link shortcut to the file will be created. This saves space, but the file won't be available on other computers if the. If you want to "embed the object" the file , select File, Insert from the Information Document or Note menu. Next navigate to the file to be linked or embedded and select it.
Next navigate to the file to be embedded and select it. Windows 98 only supports embedding. Note: You can embed several documents or files in one Information Document or Note, intermixed with text if desired. How to View or Execute the Linked and Embedded Documents There are several ways to open and execute linked and embedded files.
From an opened Information Document or Note, double-click on the embedded file to make the Excel commands available within the Expert Choice window. Then Expert Choices file menu will disappear and it will be replaced by the embedded program's menu, in this case Excel. To close the embedded file and executing program, in this case, Excel, you must click in the upper-right corner of the window or press Alt-Spacebar. When a pop-up list appears, select Worksheet Object, Open. You will be taken to the embedded program.
Now continue as described above. From an opened or closed Information Document or an opened Note, double-click on the shortcut link. Lesson 3 - Data Grid and Ratings Data Grid Overview The Data Grid contains data about the decision alternatives with respect to the decision hierarchy covering objectives. The alternatives appear in the rows of the Data Grid while the covering objectives appear in the columns; see Figure 1.
Instead of deriving priorities for alternatives through pairwise comparisons, as illustrated in Lesson 1, formulas can be defined under any covering objective column. Each covering objective's formula converts data about the alternatives to values priorities.
There are five different formula types that can be used in the Data Grid. Lesson 4 discusses the remaining formula types. Alternatives that are added to a model from the Alternatives pane in the ModelView, as described in Lesson 1 and 2, are known as active alternatives and are automatically added to the Data Grid. Alternatives can also be added directly in the Data Grid. All the alternatives in the Data Grid are referred to as global alternatives.
Active alternatives are checked in the grid; Figure 12 illustrates this. You can mark and unmark active alternatives and then extract them to the Alternatives pane to refine judgments or perform sensitivity analyses.
Extracting alternatives from the Data Grid will be explained later. While alternatives appear in the second column of each row of the Data Grid the covering objectives appear in the remaining column headers, while the ratings appear in the intersecting cells underneath the column headers. The value in the Total column for each alternative see Figure 1 contains the weighted sum of the values priorities for the alternative across all the covering objectives columns.
Each cell in the body of the Data Grid contains 1 a bar representing the value of the respective alternative row element with respect to the covering objective column element and 2 the underlying data or value you can ask to see either data or values. The values in the Data Grid body cells are derived from the data based on one of five formula types: Ratings, Increasing utility curve, Decreasing utility curve, Step function, and Direct data entry.
Why Use the Data Grid? Applications such as personnel evaluation, resource allocation and college admission decisions, may involve a large number of alternatives. Instead of making pairwise comparisons for the alternatives the Data Grid can be used. The use of the Data Grid combines the power of the hierarchy and the pairwise comparison process with the ability to evaluate hundreds or even thousands of alternatives.
Pairwise comparisons are still used to evaluate the elements in the hierarchy itself, but not for evaluating the alternatives. Alternatives priorities are established relative to each covering objective either by 1 using ratio scaled Rating intensities explained below , or 2 using other formulas that can be applied to data that can be optionally imported from corporate databases which are external to Expert Choice explained in Lesson 4.
This is often the case for subjective aspects of an evaluation or when there is no hard data. First we will show you how to do this and derive ratio scale priorities using rating scales that are predefined in a sample model and then we will show you how to define and prioritize your own rating scales. We will also show you how to: y. Figure 2 shows the model. Note: Expert Choice files have the extension of ahp.
If you are upgrading from an earlier of Expert Choice your models will have to be converted. Since we will be explaining how to use Ratings by changing this model, we want you to make a copy of the model first.
Type MyIcecreamSite and then click Save. japanese manual : MX-MN の取扱説明書・マニュアル Item Preview. remove-circle Share or Embed This Item. EMBED for wordpress. com hosted blogs and archive. Want more? Advanced embedding details, examples, and help! plus-circle Add Review. There are no reviews yet. Be the first one to write a review. download 1 file. download 16 Files download 6 Original. The Manual Library.
❿
No comments:
Post a Comment