Sorry About my Experiment
I'd like to say sorry about my little experiment (see previous post). I was testing the mTurk system and got a little carried away with my idea. And unfortunately, I can't get the darn HIT to delete. But it will expire soon anyway.
On the results, first off the experiment was a failure, although mTurk could be used in this way, there aren't nearly enough active users on it for such a thing to work, unless the price was bumped up, costing the requester actual money.
Secondly, Turkers please think before blinding following instructions, email me the screenshot? How on earth could I match a Turker with an email, and it would be pointless with so many. The only point in asking for a screenshot, is to make people think I was going to actually check. While in reality every single person will be approved. In general it was a bad idea, one that I should've thought out when it wasn't late at night. But I got rather excited about whether it would work or not. Such HIT's in general are wrong, and against the mTurk TOS. Mine could be justified by being called an experiment, but still it was a bad way to test to see if my idea would work. I apologize to the digg community who hardly noticed it, and to the mTurk community that may have a lesser view of me now. I promise the future test HIT's of mine will be much more responsible.
In short, I let my curiosity loose and abandoned logic somewhere on the road. I will not let it happen again.
On the results, first off the experiment was a failure, although mTurk could be used in this way, there aren't nearly enough active users on it for such a thing to work, unless the price was bumped up, costing the requester actual money.
Secondly, Turkers please think before blinding following instructions, email me the screenshot? How on earth could I match a Turker with an email, and it would be pointless with so many. The only point in asking for a screenshot, is to make people think I was going to actually check. While in reality every single person will be approved. In general it was a bad idea, one that I should've thought out when it wasn't late at night. But I got rather excited about whether it would work or not. Such HIT's in general are wrong, and against the mTurk TOS. Mine could be justified by being called an experiment, but still it was a bad way to test to see if my idea would work. I apologize to the digg community who hardly noticed it, and to the mTurk community that may have a lesser view of me now. I promise the future test HIT's of mine will be much more responsible.
In short, I let my curiosity loose and abandoned logic somewhere on the road. I will not let it happen again.
posted by Eric Cranston at 12:27 PM
6 Comments:
I think there are more than enough active users that this could have worked. However, like you said, most people (like myself) thought that registering on a site, clicking on a link, then taking a screenshot of that and then e-mailing it to somebody was far too much work for 1 cent.
Not everybody who turks is very technical, so having to deal with screenshots may have turned some people off, and exposing your e-mail to someone who may then spam you definitely turned off some people as well. In any case, it was an interesting (if slightly misguided) experiment.
Oh well, it was an interesting idea anyway.
Keep up the good work on the blog though.
Cheers
Best regards.
-cog
I simply was too much for for not enough money. .10 to .25 would have made it work.
It has, indeed, expired.
Thank you for your humbleness, "oops!" would have been enough really.
I looked at the HIT and didn't like it and didn't do it, and not because of the 1 cent pay, I am always keen to help a Requester test.
I am registered with DIGG and find it a useful service and I would not wish to exploit it, although there are many who would I'm sure, (judging by all the spam I get everyday).
I look forward to such genuis as you have displayed with this idea being applied to HITS that are more acceptable to me and my ilk.
Mike
Post a Comment
<< Home