In my study, I will investigate if generosity in a dictator game changes as an individual’s wealth level increases or decreases, as compared to a stable wealth level. Additionally, I will examine whether transition direction affects these distributional preferences i.e. going from rich to poor or vice-versa. Finally, I will study whether these preferences differ for participants who believe they deserve their change in social class as compared to those who believe they were simply lucky.
The study will be conducted as a 3 (Transition Direction: RMP, PMR, MMM) X 2 (Success Type: Merit, Luck) X 3 (Participant Wealth: Rich, Middle, Poor; within subjects) mixed design.
However, my current design confounds order effects with transition effects i.e. I cannot be sure if the effects I find are because a person is rich right now or if it's because she transitioned and became rich.
I thought of making the transitions random to remove the confound but this would not adequately answer the research question as I want to look at whether staying at a stable wealth level and THEN transitioning affects distributional preferences. Randomized transitions would mean that participants do not stay at one wealth level for any period of time.
I also thought of adding a PPP and RRR group to remove the confound but I don't prefer to do this as it requires a lot of power.
Another idea would be doing a pseudo-random design where I pre-program the participant's trajectory. Something like a Latin Square Design. However this does not seem to allow a stable wealth level either (MMM).
I'd appreciate it if you had any insights on this and on how I could solve this design issue.
Thanks!
