[Solved] How can a requester find my worker id in qualtrics?

Discussion in 'Help & Guides' started by L Lemon, Jul 27, 2020.

  1. L Lemon

    L Lemon Survey Slinger TurkerView Masters

    Messages:
    4,240
    Gender:
    Female
    Ratings:
    +12,417
    I received a rejection for wrong code. I'm positive I did the hit, so I emailed the requester (H.M. Schwartz). He got back to me right away and said he had no way to verify my worker id, but I'm pretty sure there's some way he can. He's very responsive and wants to help me, but he's not quite sure how to verify this.

    Do we (TV) have some sort of documentation or somewhere I can lead him to access the worker id info?

    Here's his actual email if it helps. Thanks everyone!

    I so appreciate your using turkerview (especially because you're not the only person I'm reviewing; the steps to overturning rejections are a huge help and this is the first I've found it)! Thank you!

    Looking on my end, the issue is that the five-digit code you submitted through MTurk (48972) does not match any of the codes generated automatically by the Qualtrics survey. I see that you spent a couple of hours according to MTurk, so I tried matching up based on duration spent on the survey but everyone that spent as long or longer than you did according to MTurk sent a code that matched on the Qualtrics side. I did sign up for turkerview... but can't seem to view the evidence you posted there. Is there any way to email it to me?

    That said, it definitely does sound like you took the survey to me! Unfortunately, the only information I collect on the Qualtrics side is the questions asked of you, the randomized ID (within specifications) generated to match on the MTurk side, and the start and end time on the Qualtrics side; all of this is to protect participant confidentiality as much as I can since this is an empirical/experimental study. Without a matching code or more information, I can't find your survey and reconfirm and without your explicit permission to search and make those confirmations more in-depth, I cannot do so.

    If you'd like me to try to do this, please respond with an email saying so, and then let me know if you still have the resumes that you downloaded? If you can send your versions of the resumes (or simply tell me the names listed at the top and the first job listed for each resume; they might still be in your downloads folder if that's where you downloaded them to) I might be able to match up for you and get this overturned in the next day or so. I can't have you re-take the HIT without also confirming what your previous responses were anyway (because it's academic research), but if I can find your response from those on the qualtrics side that didn't match up with an MTurk number and narrow which experimental group you were in, I might be able to fix this for you tomorrow. :) I haven't yet gone through the qualtrics data to see which items don't match up, on that side, to MTurk items, but I know there's about three or four where folks' codes were incorrect who likely participated in the experiment; I'm doing this tomorrow.
     
    • Like Like x 1
  2. L Lemon

    L Lemon Survey Slinger TurkerView Masters

    Messages:
    4,240
    Gender:
    Female
    Ratings:
    +12,417
    Update: I sent him this:

    https://www.cloudresearch.com/resources/blog/workerid-and-all-mturk-fields-sent-to-qualtrics/

    and here's his reply:

    You're correct; my IRB would have considered that method inappropriate (it makes participants too easily identifiable) so I could not use it. Instead, the only links I have to MTurk are those final randomized codes from Qualtrics since they don't directly identify the individual outside of the process of verification for MTurk. Unfortunately, it's designed to make exactly what we're discussing here (me connecting individual participants to their MTurk ID) impossible outside of that matched code from the survey.
    If anyone has anything that might be helpful to me, it'd be appreciated. I think I may have to just eat the rejection.
     
    • Like Like x 1
  3. L Lemon

    L Lemon Survey Slinger TurkerView Masters

    Messages:
    4,240
    Gender:
    Female
    Ratings:
    +12,417
    If anyone's still reading this, the requester is really going the extra mile to fix his issue - apparently there were of glitches in hundreds(?) of his submitted hits and is trying to make it right. He's confirmed mine will be overturned, but he sent a bulk email asking the following. I'd really love to help this guy out: he's a good egg. I don't have enough experience to offer anything useful to his question, maybe all y'all's do:

    One piece of feedback I will ask for, if you're interested enough to give it, is if you've participated in similar academic research like this through MTurk, what have researchers done to validate responses (from your perspective)? This was the method most preferred by my IRB, but I know there are others folks use, and you all seem to have some experience working on academic research.
     
    • Like Like x 2
    • Love Love x 2
    Last edited: Jul 30, 2020
  4. Social Cognition Lab

    Social Cognition Lab A3FCJX0FTFAB96

    Messages:
    35
    Ratings:
    +94
    Hi L-Lemon! Not sure if this thread is still active, but I work for an academic research lab and we regularly use MTurk for our research. To be honest, this thread is a little out of my depth right this second, but I'm happy to try and help out.

    To clarify: was the issue here that the requester was using some sort of automatic approval system, and your code apparently didn't match what was expected, so it automatically rejected you?

    [​IMG]

     
  5. L Lemon

    L Lemon Survey Slinger TurkerView Masters

    Messages:
    4,240
    Gender:
    Female
    Ratings:
    +12,417
    Hi there, thanks for responding. This was quite a while ago and has since been resolved. The requester was able to find a way to get the information he needed and overturned all the rejections. I can't remember exactly what the process was, but it's no longer an issue. Thanks for checking in though!