5.50USD HIT 10AM CDT USA/Canada

citizendog

New Member
Requester
Joined
Dec 3, 2019
Messages
1
Reaction score
1
Points
3
Age
32
Gender
Male
I will start a HIT today, 4th of December
  • 6.00 USD with time-limit of 50min
  • for 330 MTurks
  • greater than 95 %
  • only US-Americans and Canadians (for scientific reasons).
  • at 10:00 AM CDT or 08:00 AM for Los Angeles/Vancouver and 11:00 AM for New York/Québec. That is 5PM for me in Munich.
  • this is me: https://turkerview.com/requesters/A3LU6XQIFDYX1J-ulrich-strunz
I will require many participants in a short amount of time, due to the experiment's specifics. Please inform your fellow MTurks, it will help me out a lot.
Bring time. The experiment might easily last more than 30 minutes.
During the entire experiment, I will monitor and help via email support. I usually answer instantly. I can proudly say that this experiment is very special.
In any case, you would do me a huge favor, when after having finished the HIT, you refrain from sharing details online on what is going on during the experiment. Please stick with criticism that does not give away how the experiment works.

edit: changed payment from 5.50 USD 45min to 6.00 USD 50min and increased from "greater than or equal to 95 %" to "greater than 95 %".

The following is from my reddit post:

About the low HIT approval rate, which will make it hard for you to join:
So far I decided for the "lower quality boundary greater than 95". Experiments with greater than or equal to 99 % have shown to me that I will not get enough participants in a shorter period of time. This is very important for this particular experiment, and I have tried 4-5 different design changes, with 2 different software prototypes. When I do not get enough participants to join, all MTurks will suffer by earning less, and paradoxically my data will even be worse.
I listened in to all of your suggestions, and I realized that this is really a big issue for experiments like this, if you want to advertise them. I might actually write a paper about the problem. It is a really hard decision between data quality paradoxes, ethical payment, false incentives etc.
This actually shows me that experiments in my field of interest, which are really scarce, are in crucial need of solutions...the way Amazon Mechincal Turk works, makes a nice solution impossible.
Thanks for all your inputs! And sorry that you might fail to catch the train.
 
Last edited:
  • Like
Reactions: scrmcasey