I am fairly new to mediation analysis and have a question regarding the type of analysis I need to go for in the following design:
Basically, my study has a mixed design. I have a two-level X as between-subject (Stress vs. No Stress) that i measured four times throughout one session for each participant (Tpre, Tpost, T3, T4). I measured Y once (Social Conformity). I measured the supposed mediator M (Self-Threat) also four times.
I'm seeking help for how to conduct the mediation analysis in my case.
Currently, I'm thinking of conducting Sobels-Test with data from X and M simply being taken from one of the measures.
I'm wondering is there any other method that makes more sense?
To my mind your are losing a lot of power of this design by only measuring the outcome (Y) once. I hope I have misunderstood.
Good Point, actually I estimate Y in a paradigm of 133 trials within-subjects and merge the data to one measure. Does that help or do you mean something else?
Can you upload a small CSV file showing what the data look like for 2 or 3 subjects? It should be in a LONG (person-time) format, and should have variable names in the first row. I think that would help readers to understand much more clearly what you are describing. Thanks.
In principle I try not to aggregate ( as you lose a lot of information) but I am not really understanding your design in its detail. Bruce Weaver 's suggestion is a good one. And how do the 133 trials relate to the timing of the interventions and Tpre, Tpost, T3, T4? And what does mixed mean here?
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Sampling Design
- Sample Size