Hi I am planning to conduct a within subject study: where we are trying to test the effectiveness of an intervention: the study will test participants on 3 occasions i.e. Time 1 (baseline without the intervention)
Time 2 (with the intervention)
Time 3(with the intervention)
I am testing the (i)Parental distress using the subscale from (PSI short form)
(ii) Perceived stress scale (PSS)
(iii) Brief COPE
I want to see if the intervention affects the stress levels and if there is a relationship between coping and stress. What would be the best analysis to conduct to find this out?
Any help will be much appreciated!!
Repeated measures ANOVA
You have to counter balance subject for order effects to ensure there is not learning, fatigue, habituation or sensitization
I agree with counter-balancing questionnaire order. Then use a separate 6 (order) X 3 (time) mixed model Analysis of Variance (ANOVA) on each of the three measures. If the data do not conform to ANOVAs normality assumption and a transformation is not successful, then ignore questionnaire order and use Friedman's ANOVA (nonparametric, repeated measures) on each of the questionnaires.
ANOVA, although I agree with a previous poster and would consider a control group in order to be able to make more inferences about the efficacy of the intervention.
I would add that in clinical intervention studies it is usual now to have a treatment-as-usual (TAU) as the control condition rather than a wait-list or similar. The ideal situation is to have three conditions: Target Treatment, Comparison Treatment (to control for things like the Hawthorne Effect), and TAU. However, that isn't always possible. As previously suggested mixed model ANOVA or MANOVA (given you have multiple DVs) for analysis would be appropriate for the analysis.
You could do this as a multi-level model (MLM) with time/treatment as a within-subject factor at level 1. You can control for any between-person differences at level 2.
- Badges
- Science topic
- Similar topics
- Psychology
- Developmental Psychology