Subscribe Bookmark RSS Feed

Repeated measures with one subject?

jstrut

Community Trekker

Joined:

Jan 9, 2016

Hello,

 

We are trying to do an analysis, where we simply want to ask whether the pattern of results looks different under different conditions
And I am struggling with how to do it because we have a single value for each condition.
 
So, imagine that we have 4 moods:
1) Calm
2) Excited
3) Positive
4) Creative
 
For each of these moods, we ask people to rate how much 13 different colors represent that mood:  So, imagine 13 different colors.
 
That gives us 4 * 13 values for each subject.
 
The question is whether the pattern of ratings for the different colors DIFFERS for the different moods
 
So, that’s asking if there is an INTERACTION between mood and color.
 
I am somehow stumped on how to test this.  It seems like a simple 2-way ANOVA, but I don’t know how to set that up in JMP.
 
I've attached a graph where the colors are on the bottom and each line represents the different moods and the y axis is the percentage that color was chosen.
chart.png
2 REPLIES
stephen_pearson

Community Trekker

Joined:

Oct 6, 2014

Since the colours fall on a continuous spectrum it might be an idea to order them that way or even rephrase your question using a continuous variable that reflects the colour.
jstrut

Community Trekker

Joined:

Jan 9, 2016

Thank you for the suggestion. However, I'm not sure what you mean. Each subject looked at specific colors, so there wouldn't really be a continuous measure because there was no "in-between" color. But even if I did convert color to a continuous variable, would that allow me to perform the analysis I am looking for? (Whether the pattern of color preference differs by mood within just one subject?) Thanks!