turn on suggestions

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

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- Mixed designs anova

Topic Options

- 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

Jan 27, 2010 9:01 AM
(992 views)

Help! I am trying to do a mixed design where one of my factors (TIME) is between subject factor - so I have (pre data) and (post data) - that comes from different subject populations.

What I am interested in is how TIME effects a particular interaction. So within the experiment there are two factors of interest: PRIME and WORD. In general, I am interested in the interaction of PRIME*WORD. For this particular analysis I am interested in how PRIME*WORD changes as a function of TIME.

What I would like to see in the output is whether the 3-way interaction of PRIME*WORD*LOCATION is significant, given that PRIME*WORD should have within-subject variances and LOCATION should have between-subject variance.

Thank you for any suggestions!

-S

What I am interested in is how TIME effects a particular interaction. So within the experiment there are two factors of interest: PRIME and WORD. In general, I am interested in the interaction of PRIME*WORD. For this particular analysis I am interested in how PRIME*WORD changes as a function of TIME.

What I would like to see in the output is whether the 3-way interaction of PRIME*WORD*LOCATION is significant, given that PRIME*WORD should have within-subject variances and LOCATION should have between-subject variance.

Thank you for any suggestions!

-S

2 REPLIES

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

Jan 29, 2010 6:00 PM
(974 views)

It depends. I would perhaps use time as too Y responses (pre data) and (post data) and analyze the model terms of PRIME and WORD along with the (PRIME*WORD) interaction. I would set up a design of experiments and run in random fashion. You could then visualize your pre data response and post data response with the prediction profiler and visualize the behavior of the response versus your factors of interest. I am confused about the very end of your question where "LOCATION" was added to the mix.

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

Feb 1, 2010 6:21 AM
(974 views)

oh my goodness that was a typo! I am actually doing 2 separate analyses (one with time as a between subject factor - and one with location). So in my previous question, you can just substitute LOCATION for TIME. Will try running your suggestion. thanks.