1. API Information

    The TurkerView API page has instructions as well as contact information for support in getting set up with our new API infrastructure.

    You can also email us directly at hello@turkerview.com if you run into any problems!

    Dismiss Notice

01/25 - Fungi Friday!

Discussion in 'Daily mTurk HITs Threads' started by Melting Glacier, Jan 25, 2019.

Thread Status:
Not open for further replies.
  1. Ornac

    Ornac Survey Slinger

    Messages:
    5,677
    Gender:
    Male
    Ratings:
    +5,258
    Well now that I'm here I think I'l run off and watch the new half season(what's up with that) of Kimmy Schmidt.
     
    • Like Like x 1
  2. moingeor

    moingeor New Turker

    Messages:
    12
    Gender:
    Female
    Ratings:
    +4
    Do you wait to add bonuses until they actually pay?
     
  3. ChrisTurk

    ChrisTurk Administrator

    Messages:
    37,262
    Ratings:
    +93,252
    If you know what the bonus is its way preferable to go ahead & add it in. You can confirm them (or mark them as unpaid!) later on which will improve the data, but the faster they're added the more accurate the expected wage data will be for others :)

    Requesters who don't pay out bonuses will end up w/ flags on their profiles for it, so its good to throw that data into TV ASAP
     
    • Useful / Informative Useful / Informative x 2
  4. Ornac

    Ornac Survey Slinger

    Messages:
    5,677
    Gender:
    Male
    Ratings:
    +5,258
    Nice. I would love to see something that drags the average wage into your range for the API. I know how to adjust for my speed, but it would be great as a long term thing.
     
  5. Ornac

    Ornac Survey Slinger

    Messages:
    5,677
    Gender:
    Male
    Ratings:
    +5,258
    Would also save me hitting the actual reviews to see who reviewed something when there's only one or two reviews.
     
  6. RobertDevious

    RobertDevious Survey Slinger

    Messages:
    3,890
    Gender:
    Male
    Ratings:
    +5,774
    Man, every decent survey I'm grabbing tonight has the most convoluted set of rules ever, followed by a comprehension check that I don't even want to attempt. Batches plz
     
  7. A6_Foul_Out

    A6_Foul_Out Organic Meme Panda

    Messages:
    13,792
    Gender:
    Male
    Ratings:
    +17,899
    [​IMG]

    [​IMG]

    [​IMG]
    (me thinks people w/similar experience didn't get 16$ surveycomets)

    [​IMG]

    [​IMG]

    working as expected @ChrisTurk

    :leftyhyperflail::leftyhyperflail::leftyhyperflail::leftyhyperflail::leftyhyperflail::leftyhyperflail:
     
  8. nick2412

    nick2412 Active Turker

    Messages:
    996
    Ratings:
    +1,331
    That urn one was fire tho.
     
    • Like Like x 1
  9. jessers

    jessers Survey Slinger

    Messages:
    11,409
    Gender:
    Female
    Ratings:
    +31,947
    the nhl skills comp opened with the freaking shark song and i blame all of you
     
    • LOL LOL x 3
  10. dan

    dan Survey Slinger

    Messages:
    19,507
    Gender:
    Male
    Ratings:
    +41,062
    What goes into similar experience? Amount of reviews submitted?
     
    • Like Like x 1
  11. A6_Foul_Out

    A6_Foul_Out Organic Meme Panda

    Messages:
    13,792
    Gender:
    Male
    Ratings:
    +17,899

    a write up/pop up/question mark hover thing would be nice here tho
     
    • Like Like x 1
  12. awdrf

    awdrf Well-Known Turker

    Messages:
    3,441
    Ratings:
    +4,537
    M Research [A1S1F2242EUN4M] Academic Survey: Your Opinions on Trade Disputes (about 10 minutes) - $1.00 | PANDA


    Fair

    Unrated

    Unrated
    $12.12 / hour
    00:04:57 / completion time
    Pros:
    Cons:
    -A fair bit of writing.

    Qualifications: HIT approval rate (%) > 95; Location EqualTo US;
    To read M Research's full profile check out TurkerView!
     
    • Good Review! Good Review! x 1
  13. Randomacts

    Randomacts Survey Slinger

    Messages:
    86,237
    Gender:
    Male
    Ratings:
    +114,043
    upload_2019-1-25_18-33-54.png

    Slowpokes
     
    • WOW WOW x 2
    • LOL LOL x 1
  14. jessica93

    jessica93 Well-Known Turker

    Messages:
    5,314
    Gender:
    Female
    Ratings:
    +12,784
    EPL_Beckman [AU08GAAGXATA8] English Language Study - CHOOSE ONE HIT ONLY - $3.00 | PANDA


    Generous

    Unrated

    Unrated
    $16.69 / hour
    00:10:47 / completion time
    Pros:
    Typical sentence-reading experiment
    Cons:
    No progress bar
    Press F/J to answer a question after every sentence

    Qualifications: HIT approval rate (%) > 95; Location EqualTo US;
    To read EPL_Beckman's full profile check out TurkerView!
     
    • Good Review! Good Review! x 1
  15. dan

    dan Survey Slinger

    Messages:
    19,507
    Gender:
    Male
    Ratings:
    +41,062
    • Your Wages: $7.35
    • Similar Speed: $18.56
    • Similar Experience: $0.00


    Utterly useless on requesters with wide range of HITs like Research Tasks
     
    • WOW WOW x 1
    • Useful / Informative Useful / Informative x 1
  16. ChrisTurk

    ChrisTurk Administrator

    Messages:
    37,262
    Ratings:
    +93,252
    Changing the average itself --- I think --- is actually going to produce really ugly results.

    Imagine changing it for something like C-SATs? Woooo buddy that'd be ugly, and I think you'd end up seeing a lot of cases where it'd randomly throw super crazy data at you that turns out being really wrong :s
    No, its your # of approved HITs on MTurk (which is why the data is a little lacking atm, since that change only went through w/ TVJS10).

    Both similar speed & experience filter your times/reviews out of the aggregate though, so your own reviews don't change the data, its legitimately just other folks.

    Honestly, for most of the core TV reviewers you really wont see much use out of this, because we're all kind of the same person to begin with so if you have 8420,234003 reviews already well... "your" average is going to be == everyone elses average and that is going to be === to the requester's average lol. But it does make it more resilient to 25,000 "slowpokes" joining the system since their data would get filtered out of these kinds of results.

    It should also be a flag if you see a requester profile and then no data in that column, it means the only people who have reviewed it are highly dissimilar to you and you should give the profile a closer look. If you're a speed demon & no one else like you has reviewed, expect to do better than the average hourly displayed. Vice-versa for careful/relaxed pace.
     
    • Useful / Informative Useful / Informative x 4
  17. ChrisTurk

    ChrisTurk Administrator

    Messages:
    37,262
    Ratings:
    +93,252
    or I mean right here too :emoji_joy:
     
    • LOL LOL x 1
  18. RobertDevious

    RobertDevious Survey Slinger

    Messages:
    3,890
    Gender:
    Male
    Ratings:
    +5,774
    Seungcheol Jung [A1YLWWJ5SIGR8Q] Smart Watch Survey (3 minutes) - $0.20 | PANDA


    Fair

    Unrated

    Unrated
    $8.09 / hour
    00:01:29 / completion time
    Pros:
    Cons:

    Qualifications:
    To read Seungcheol Jung's full profile check out TurkerView!
     
  19. Neven123

    Neven123 Well-Known Turker

    Messages:
    6,442
    Gender:
    Male
    Ratings:
    +7,337
    Leopold [ABOMZF5VDXSEO] your imagination and perception - $1.50 | PANDA


    Good

    Unrated

    Unrated
    $14.71 / hour
    00:06:07 / completion time
    Pros:
    Cons:

    Qualifications: HIT approval rate (%) > 95; Location EqualTo US;
    To read Leopold's full profile check out TurkerView!
     
  20. Ornac

    Ornac Survey Slinger

    Messages:
    5,677
    Gender:
    Male
    Ratings:
    +5,258
    Fair enough. I don't have a solution then, but as a suggestion of something to think about.... If I see a hit, not the requester, with 1-2 reviews, I end up going to the site to see who reviewed it to find out if the hit is worthwhile. Someway to make that information accessible from the mturk page would be 10 sorts of useful, at least for me. Not neccesarily who reviewed it, although for me that would be fine, but a way to compare to the average speed or something like that:dunno:
     
    • Like Like x 1
    • WOW WOW x 1
Thread Status:
Not open for further replies.