Hi everyone

First of all, I like the new forum! :)

Now to my question / problem:
I conducted a study on Mturk using oTree two weeks ago. I uploaded the three different treatments (Delegation, One-Step-Delegation, No-Delegation) as one project via oTree Hub and created one session for each treatment, each session was published simultaneously on MTurk.

The responses were great, 1600 MTurk-workers participated in the study.

19 assignments were auto-rejected, since they were not found in the data-base.

The problem I encountered was, that 15 workers whose assignments got auto-rejected actually participated in the study, but with a different Assignment-ID in the data. They worked on a different treatment than the one, their assignment got rejected.

The workers should not be able to work on two different treatments, since I used the same qualification-id for all three treatments.

Has anyone an idea what happened here and how to prevent it from happening again? I plan to conudct more studies on MTurk with the same (or simliar) settings.

Thanks in advance for your help!
Daniel