What analysis should I use for my data


I’m running a research in which there are 2 groups - intervention group (will go through a 14 day treatment) & control group (will go through a placebo intervention) - comprised of heterosexual couples in a romantic relationship. Each partner will complete the same series of relationship related questionnaires (relationship satisfaction, sexual functioning, general wellbeing etc.) three time (before the interventions, after them and one month follow up).

My hypotheses:

(1) Compared to the pre-treatment evaluation, couples in the intervention group will show fewer symptoms immediately after completing the intervention and one month following that.

(2) Couples in the intervention group will show fewer symptoms compared to the couples in the control group, immediately after completing the interventions and one month following that.

In addition, I might want to compare the effect of treatment on males and females.

Someone said I should create one score for each couple on the dependent variables (questionnaires) by averaging the responses of both partners, but I’m not sure if it’s the right way to go.

What do you think is the best analysis for this study design?

I would appreciate any help.

As the participants are randomized as couples, the male and the female within each couple cannot be considered independent observations. My instinct would be to make a combined score and compare couples.

If the male and females were randomized independently to each couple, it wouldn’t be a problem to make comparisons within females and males separately, but in your case I guess you would need to at least control for the effect on the partner. Most likely there is some interaction going on where the effect of the intervention is also influenced by how it affects the partner, but I’m not sure how you would approach this.

Thank you for your response.

If I create a combined score (by averaging or summing up the individual scores of each relationship partner), do you think I could use mixed ANOVA to test my hypotheses?

I’m simply not sure how to approach it correctly and none of the dyadic data analysis methods seems to fit.

1 Like

Wouldn’t a simple regression model do the trick, comparing score at completion (and at 1 month) adjusting for baseline score? @f2harrell probably have some input on how to model these data.