Which analysis should I use?

Discuss statistics related things
Post Reply
PhilippB
Posts: 1
Joined: Thu May 11, 2023 9:36 pm

Which analysis should I use?

Post by PhilippB »

Hello,

I'm currently writing my Bachelorthesis and I dont really know which analysis I should use. I'm researching what effect conscientiousness has on food craving during negative/positive emotional states, whether conscientiousness influences the tendency to crave high/low-calorie food and whether the relationship between food craving, emotional state and conscientiousness differs depending on the calorie content of the food.
So my variables are:
dependent variable: Food Craving (continuous)
between-subjects independent variable: conscientiousness (continuous)
within-subjects independent variables: emotional state (positive/neutral/negative), and calorie content (high/low)

Originally I was planning to dichotomize conscientiousness by splitting participants into a high and low conscientiousness group through a median split and do a mixed ANOVA with simple effects analysis. Still not sure whether I should do that or not. I dont know whether it is possible to perform a simple effects analysis through the ANOVA menu in Jamovi.
However I like the GAMLJ menu, since you can use simple effects analysis really easily, and I've read their article (https://gamlj.github.io/mixed_example2.html) on using the mixed models menu for a mixed ANOVA like design. Since I have 3 independent variables, would i just need to add the extra factor into the factors box and continue as described in their article? However, I dont really know anything about the theory of mixed effects models, since we didn't learn about that in our bachelor stats class. So I dont really know how I would write the "statistical analysis" part of my thesis. Would it be possible to add conscientiousness as a continuous variable to the model? maybe as a Covariate?

Do you have any suggestions what I should do? I really need some help..

Thanks :)
User avatar
mcfanda@gmail.com
Posts: 457
Joined: Thu Mar 23, 2017 9:24 pm

Re: Which analysis should I use?

Post by mcfanda@gmail.com »

Hi
I would strongly suggest against median-splitting. You can use a mixed model as you said, adding conscientiousness as a covariate. You can even add its interactions with the other variables, as you wish.
Post Reply