Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Comparing different processes with t tests

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 17, 2009 10:59 PM
(304 views)

Does anybody know how to compare different processes using t-test in JMP?

I've seen reports such as the one shown below where defect levels from three different processes are shown in columns on the left of a report and the spreads are represented by bubbles on the right of the report.

For any given process the data points are represented by a diamond showing upper quartile, median and lower quartile. The diameter of the bubbles on the right represent the standard deviation.

Bubbles that do not overlap indicate that the difference between the two processes being compared is statistically significant.

Thanks,

Sonu

4 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 18, 2009 5:10 AM
(288 views)

To create such a plot in JMP, you would go to Analyze>Fit Y by X and make X a nominal or ordinal variable.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 18, 2009 7:13 AM
(288 views)

How does one run a t-test on two different populations where the populations can be compared side by side to a given confidence level.

Could you see the example I had posted? It shows quartiles for each population on the left and std deviation "bubbles" on the right. I've seen these kinds of reports before but dont know how to create one from scratch.

Sonu

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 18, 2009 9:47 AM
(288 views)

Please be more specific. I already explained how to do this. In JMP, Go to Analyze>Fit Y by X and make X a nominal or ordinal variable.

If you don't understand, please ask a more detailed question.

And again, your explanation of the plots is not correct. The diamonds do not show quartiles. The circles on the right are comparison circles, not standard deviation circles. The JMP documentation has explanations of these plots.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Mar 19, 2009 8:19 AM
(288 views)

You are correct. I was misinterpretting these plots.

I found a good tutorial called "Many Means" and it walked me through the plots.

Sonu