Posts: 6
Threads: 3
Joined: Mar 2021
Reputation:
0
Dear FindingFive,
Hi, FindingFive!
I hope you are doing well.
I am writing because I would like to ask you a question about copied studies.
I mean... I would like to publish three studies to MTurk simultaneously, two of which are copies of the original (basically, the three studies are same).
I am wondering whether there would be any problem in doing so. I am a bit concerned about whether the participants who participated in the previous session are not blocked from any copied version.
I always appreciate all your help and support.
Thank you very much.
Posts: 352
Threads: 22
Joined: Sep 2020
Reputation:
2
Hi Cheonkam,
Is there a particularly reason why you are running 3 identical studies at the same time? Are you trying to run 300 participants at the same time? You may crash the server if that's the case. There's a reason why capacity per session is capped at 100 - that's our estimate of how many participants we can take simultaneously within a 5-minute span.
If you do launch 3 studies, please launch them at least 1 hour apart from each other.
As for eligibility control, as long as you set up the anti-prerequisite correctly, participants should not be able to take the other 2 HITs after they have done one. However, there's a minor 1-minute delay in applying the eligibility control. So the chance of duplicate studies is slim but it's not entirely 0.
Most of these issues won't happen if you launch your studies on FindingFive directly. Eligibility control is particularly tricky with regard to mturk, especially for simultaneous sessions.
Posts: 6
Threads: 3
Joined: Mar 2021
Reputation:
0
Dear Ting,
Thank you for such a detailed reply!
We just aim to experiment with 100 participants, and I wanted to run 24 participants (8 * 3) at the same time and repeat this procedure several times in order to avoid extra-fee.
I always appreciate all your help and support.
(I would definitely try the FF platform for the running experiments as well next time, which will be much more convenient).
Thanks.
Sincerely,
Cheonkam Jeong
Posts: 352
Threads: 22
Joined: Sep 2020
Reputation:
2
In that case please feel free to go ahead! It should work reasonably well. I would still recommend launching those studies about 20 minutes apart from each other.
