I recently conducted a study with 4 within-subject conditions, divided over two sessions. My independent variable is degree of emotional eating (as measured by food intake after a negative mood manipulation), my dependent variables are food intake after some other conditions (positive mood, exposure to food, and a control condition). I want to investigate whether high emotional eaters overeat after a variety of cues (like positive mood and exposure). I hypothesize that emotional eaters (i.e., those who eat most after the negative mood manipulation) will consume more food than low-emotional eaters after each experimental condition (positive mood and exposure), but not control. In addition, I expect the high emotional eaters to eat more after positive mood and exposure compared to control, while I don't expect any differences between conditions in the low-emotional eaters.

I know how to analyze my data using a median-split to categorize high and low emotional eaters and follow this by a mixed ANOVA. However, I would like to avoid median-split and treat my independent variable as a continuous variable. But, I'm stuck on how to proceed. In previous studies in which I wanted to treat emotional eating as a continuous variable, I performed a moderated regression analysis (+1 and -1 SD). However, those studies had between subject designs, and I'm confused on how to do these analyses with my current design, if that's even possible. Alternatively, I received some suggestions for doing linear mixed model analyses.

Could someone point me in the right direction? Thanks so much for your help!


If you use R and the nlme package, the syntax would be:

m <- lme(dv ~ iv*emo_eating, random=~1|id, data=d)

where emo_eating is the continuous predictor, id is the subject-ID variable, and d is your data stored as a data frame. I am assuming that iv is a fixed effect. The summary(m) will give you regression output.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.