Decision Trees

 

1. What it looks like

…as generated by EvalC3

DTree
2. How read this Decision Tree

The tree should be read from left to right, as though it is a tree that has fallen over

This tree has 7 branches, each with a “leaf” at the end, shown by the color beige or green. Each of these branches is a prediction model , made up of a particular configuration of case attributes, described by the text labels.

The green and beige leaves describe the numbers and types of outcome  found.

Beige= outcome present. Green = outcome present. This is the predicted outcome of this model, given the distribution of cases on the leaf.

First number = Number of cases with outcome present. Second number = Number of cases with outcome absent

The example here uses data from the Krook QCA study of women’s participation in parliament in 26 African countries

Lets read the top branch…Where “quotas” are absent (0) and “Women’s status” is absent (0) this model find there are 0 countries with high levels of women’s participation in parliament, but there are 12 cases where there are low levels of women’s participation in parliament.

In the next branch…Where “quotas” are absent (0) and “Women’s status” is present (1) this model find there are no countries with high levels of women’s participation in parliament, and a “post conflict situation” is absent (0) there are 0 countries with high levels of women’s participation in parliament,but there is 1 cases where there are low levels of women’s participation in parliament.

In this example, each branch represents a configuration that is sufficient for the outcome being either present (green) or absent (beige). But sometimes both outcomes will be present, but one will be more common that the other. In other words the model will have some inconsistency (in QCA terms) and have limited “Positive Predictive Value” or “Precision” to use terms used elsewhere

3. What to do with it

To view a particular model in detail, click on the 0 or 1 cell to the left of the leaf you are interested in.

Then click on Load Model. This will take you to the Design & Evaluate view, where you will see the model attributes in the Design section and its performance measures in the Evaluation section

To save this model, click on Save Model

5. Where to learn more about Decision Tree and how they work

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s