Understanding Modeling Types

In JMP, data can be of different types. JMP refers to this as the modeling type of the data. Modeling Types describes the three modeling types in JMP.

Modeling Types
Modeling Type	Description	Examples	Specific Example
Continuous	Numeric data only. Used in operations like sums and means.	Height Temperature Time	The time to complete a test might be 2 hours, or 2.13 hours.
Ordinal	Numeric or character data. Values belong to ordered categories.	Month (1,2,...,12) Letter grade (A, B,...F) Size (small, medium, large)	The month of the year can be 2 (February) or 3 (March), but not 2.13. February comes before March.
Nominal	Numeric or character data. Values belong to categories, but the order is not important.	Gender (M or F) Color Test result (pass or fail)	The gender can be M or F, with no order. Gender categories can also be represented by a number (M=1 and F=2).

Example: Modeling Type Results

Different modeling types produce different results in JMP. To see an example of the differences, follow these steps:

1.	Select Help > Sample Data Library and open Linnerud.jmp.

2.	Select Analyze > Distribution.

3.	Select Age and Weight and click Y, Columns.

Click OK.

Distribution Results for Age and Weight

Although Age and Weight are both numeric variables, they are not treated the same. Results for weight and age compares the differences between the results for weight and age.

Results for weight and age
Variable	Modeling Type	Results
Weight	Continuous	Histogram, Quantiles, and Summary Statistics
Age	Ordinal	Bar chart and Frequencies

Changing the Modeling Type

To treat a variable differently, change the modeling type. For example, in Distribution Results for Age and Weight, the modeling type for Age is ordinal. Remember that for an ordinal variable, JMP calculates frequency counts. Suppose that you wanted to find the average age instead of frequency counts. Change the modeling type to continuous, which shows the mean age.

1.	Double-click the Age column heading. The Column Info window appears.

2.	Change the Modeling Type to Continuous.

Column Info Window

Click OK.

4.	Repeat the steps in the example (see Example: Modeling Type Results) to create the distribution. Different Modeling Types for age shows the distribution results when Age is ordinal and continuous.

Different Modeling Types for age

When age is ordinal, you can see the frequency counts for each age. For example, age 48 appears 2 times. When age is continuous, you can find the mean age, which is nearly 48 (47.677)