# Calculating Kendall’s Tau, a nonparametric ranked correlation statistic - ID 1029

Revised: 5/17/2010

Applies to

• Minitab 16

#### Description

How can I calculate Kendall's Tau, a nonparametric ranked correlation statistic?

#### Solution

You can use the Kendall's Tau macro available from our Macros Library. Kendall’s tau is a coefficient of ranked correlation which measures the level of disarray for a pair of samples such as (x,y). The value of Kendall’s tau lies between +1 (complete concordance) and -1 (complete disagreement). If the x and y values are independent, Kendall’s tau will be close to 0.

1. Click the link, Kendall's Tau Macro, below.

2. To the right of the Kendall's Tau description, click Code.

3. Save the page as a macro, called REGTAU.MAC, in your Minitab macros directory (for example C:\Program Files\Minitab\Minitab 16\English\Macros). Instructions are available by clicking "How to save this page as a macro."

4. To run the macro, type the macro command in the Command Line Editor ([Ctrl]+[L]) or at the command prompt (MTB>) in the Session window. To display the command prompt in the Session window, activate the Session window and choose Editor > Enable Commands.

For example, suppose a subject is given eight different flavors of ice cream and is asked to rank the flavors on the concentration of chocolate. The x values stored in c1 represent the actual concentration ranking. The y values stored in c2 represent the subject’s ranking. The data is set up as follows:

`c1	c22	35	57	68	83	21	16	74	4`

In the Command Line Editor, or in the Session window at the command prompt, type:

%REGTAU C1 C2

The value of Kendall's Tau for this example is 0.85714.

#### Related Documents

Still looking for answers? You can...

• Click on another category at top left
• Contact Customer Support
• Search for text or a specific answer:

Help with Search

+1-814-231-2682