Are you teaching regression analysis in a course? Or are you interested in developing a better understanding of regression concepts yourself? The new **Demonstrate Regression** teaching module may be just what you need.

Statistician and JMP training specialist Mark Bailey and I, along with professors Amy Froelich from Iowa State University and Bill Duckworth from Creighton University, developed this interactive module. It is the ninth in a series of 10 Concept Discovery Modules, available via the JMP File Exchange (download requires a free SAS account) and the JMP Academic website.

These modules were born out of the results of a survey administered to high school and college statistics instructors. We asked questions such as:

• What statistical concept do students in introductory statistics courses have the greatest difficulty learning?

• Which of these concepts lend themselves to visual and interactive demonstrations or applets?

Regression, it turns out, was near the top of the list on both fronts -- an ideal candidate for a JMP teaching script!

The resulting regression module allows you to control population characteristics, such as the intercept, slope, and correlation, and simulate data at a given sample size. Then, you can generate several samples to study sampling variation. Or explore your own data set or create data from scratch. Then, explore concepts such as residuals, sum of squared errors, least squares fit, and correlation. Add new points, and drag the points to study influence and leverage.

Mark used JSL (JMP Scripting Language) to develop the module. He has written many useful teaching scripts over the years (many of his most popular scripts can be found in the JMP Sample Scripts directory - Help > Sample Data > Teaching Demonstrations). But, with this module, Mark really had his work cut out for him! According to Mark, “This script is the most complex that I have written yet. The nature of making an interactive, responsive, and useful tool for demonstrating the statistical principles in simple linear regression demanded many advanced JSL techniques and features.”

Feedback on the final script has been extremely positive. Here’s what a few folks have had to say:

• "WOW!!! I like this!!!"

• "I can easily see using it in classes and workshops. What a nice teaching tool!"

• "I’ll be able to get more across in 10 minutes with this tool than in 90 minutes (or more) of lecture."

Of course, we couldn’t have done this alone. We’d like to thank the dozens of you who have tested early versions of the script, provided feedback and ideas for enhancements, and patiently awaited this final version.Thank you! Thank you! Thank you!