Back on the main prediction page, Watson Analytics shows what is named the prediction scenario selector (shown in the following screenshot). This lets you drill deeper into the insights that may be hidden in the data. With the prediction scenario selector, you can specify how many fields you want to view that act as predictors for your target (other than the Initiator in our example):

If we click on Two Fields, Watson Analytics changes our visualization and we see a new set of visualizations, as shown in the following screenshot. We can now see how these two variables influence the number of illnesses (our target):

There are now three colored dots on our bullseye. If we click on each of the three new dots, we will see that the following pattern was found in our data:

More interesting details. According to Watson Analytics, other than who initiated the dialog between the physician and the patient, the following topics of discussion have an impact on the patient's overall heath:

• The patient's use of alcohol
• How long a patient has had symptoms
• How the patient views their sex life

As a data scientist, you should investigate the prediction scenario selector. For example, if you select Combination, the visualizations provide a deeper and perhaps more predictive analysis of the data by showing how a combination of variables influences your target. Choosing One Field leads to predictions that are easier to understand, but that might be less predictive. Choosing Combination might lead to a prediction that is more accurate, but harder to understand. Choosing Two Fields might give you predictions that are somewhere in the middle.