Getting Started with Insight Tree

 

1. After Installation

Start menu item Desktop link

  


Insight Tree

When you double-click the start menu entry or the link to InsightTree.exe you will see a window opening like the one on the screenshot below. Take a minute to get your first impression, look into some menus and then let's start working.


2. The First Decision

You see already 2 major elements of the decision tree:

Decision Tree - Start

There are 3 node types:

There are 3 secrets of using the decision tree software:

The dotted frame around the node "New Decision Tree" signals that it is activated. Simply type what your decision tree is about and finish typing by hitting the Return key: 

Decision Tree - Start

Decision Tree - first step

Your first decision:

Right mouse click the end node (the triangle) and chose "Add branches" from the context menu.

Just go on typing into the new nodes the options which will be represented by the branches.

Decision Tree Options

The square represents the first decision you have to make. You may click on the small symbol at the square bottom to edit details of this decision

Show decision attributes

Now add details to the decision and its options. When editing payoffs you should use negative numbers for costs and positive numbers for income.

Attributes of a decision

You can add as many options to your decision as you need (the theoretical maximum is 4,000 billions, just enough for day-by-day decision making) by right clicking any part of the decision node and selecting "Add branch".

Decision Tree - add further branches

The new branch can be edited immediately.

Use the right mouse-click on a branch to move it up or down, to define it as the taken option, to add a text comment to remove it again.

Use right mouseclick to further build the decision tree

The decision tree in the images above has also numbers assigned for payoff (cash flows).

Edit decision tree branches

Please note for payoffs:

Step by step you create your decision and add the information which will later be used to optimize it.

Decision tree with one outline level of decisiions to be made.


3. More Nodes, Chances

If you right mouse click an end node you can add more branches. By default this will give you another decision node with two branches. Right mouse click the node to convert it to a chance node.

Create a chance node in your decision tree by converting a decision node

Chance nodes are traditionally marked by a circle. There are two types of chance nodes: %-type and numeric, i.e. you may assign any combination of numbers to the chance options.

The percent values of the single possibilities are most commonly desired to add up to 100%. Decision Tree will change the color of the node if they don't but this will not be a problem to you. The software can use any given combination of percentages and calculate the correct average. Numeric chance nodes will not change color.

Numeric values can be any integer numbers. The following example shows a set of PERT estimation from three-point estimates each weighted with "1" for the optimistic value, "4" for the most likely value and "1" again for the pessimistic value:

A section of a decision tree with a chance node

The following tree shows different ways to estimate payoffs and their likeliness for each decision:

Decision tree with one decision / four options

The example uses 2 types of chances: Percentage in option 1 and 3 and numeric in 2 and 4.

A click on the symbol in the toolbar optimizes the decision:

Decision tree with opmtimization applied

Improve the existing product is obviously the best solution in monetary terms: It has a positive payoff of $85.833, more than any of the other options, which have a payoff between -$12,500 and $67,250.

This option is marked with stronger lines. You can apply own styles and influence the design of the entire tree.


4. Applying Styles and Changing Formats

After

the complete tree looks now as follows:

Decision tree with the Style "Metal Guru" applied

Note: You can develop your own style in the "File - Tree properties" menu.


© 2006 - 2008, www.visionarytools.com Daniel & Oliver Lehmann, Munich, Germany