23

I strongly believe that the reviewing process carried out by Language Specialists (LS) needs a radical overhaul to minimize the subjectivity and bias exhibited by some LS'. Despite the several questionnaires/surveys translators were asked to complete in order to share their ideas about the process, I have not noticed any tangible changes. However, this community thread is dedicated to addressing a specific question.

Why is Translator's ID visible to LS' in Their Reviews? 

The Go-Check support article, here, displays a couple of screens from a LS' view while selecting the severity and providing comments in a review. In the upper left corner, the Translator ID is visible. Don't you think that this adds to the sheer subjectivity of some reviewers and the likelihood of bias?

My fellow translators, please consider this as an appeal to Gengo to hide the Translator's ID from the LS' reviewing workbench. If you agree, please give the thread a thumbs-up.

The first screen

The second screen

 

5件のコメント

  • -2
    Avatar
    gunnarbu

    As a general rule reviews should of course be as unbiased as possible, and hiding the translator ID could be one possible way to secure this, but I doubt very much that bias and subjectivity as described above is common among Gengo LSs. Over ten years I have always been treated objectively an unbiased, also for re-reviews when I have challenged the first review.

  • 2
    Avatar
    Rica Tero

    Hi @Abdek! 

    Thank you for sharing your concern on this thread. I want to thank @Gunnarbu for sharing your experience too. It's highly appreciated.

    I understand that hiding the translator ID will be a 100% guarantee that our LSs are not biased. I've shared this with our community manager, Maria, and it's a great thing to lessen your worries. We will do our best to communicate it, but we can't assure you of your expectations. There are a lot of factors to consider in this change.

    Also, I want to emphasize that the quality team strictly monitors the work and performances of all our LSs. We always encourage everyone to submit a re-review request if they are certain that the mark(s) they got on their GoCheck are wrong.

     

  • 8
    Avatar
    Abdek

    @gunnarbu I have clearly stated in my OP "some LS'". There is no need to misinterpret the intention of the post away from its obvious purpose, which is to basically answer the question; I am not generalizing. If you are fortunate enough to have a LS who is unbiased in your language pair(s), that would definitely be your own experience and wouldn't make it a rule. 


    @Rica Tero Thanks for sharing my request with the Community Manager. Could you please elaborate more on the "factors to be considered" in hiding an unnecessary piece of information?

    As Lord Acton said, "Power tends to corrupt, and absolute power corrupts absolutely." Granting absolute power to LS' over the last say is undoubtedly a source bias, to say it lightly.

    1- Even when I have escalated several inaccurate reviews that I have received, they are forwarded to another LS flagged as "disputed tasks". It is evident, based on several incidents I have experienced firsthand, that LS’ often get biased and defensive. This behavior is expected, knowing the fact that LS’ review each other’s work without an impartial party to verify their mostly subjective opinions.

    A suggested solution, in the case of a task accepted for a second re-review, to avoid the expected bias, is to assign the task as a normal review to a different LS to assess its overall quality from scratch. If the disputed errors are not detected, they should not have been flagged in the first place, and any additional errors flagged by the second LS should be cancelled, since they were not flagged by the first LS. I have requested this in my correspondence with the Quality Team, and it was simply disregarded.

    2- Gengo has taken a great initiative earlier last year, asking Language Specialist to share the most common mistakes observed or the monthly error trends over the forums. Since then, I can see every month several Language Specialists sharing the common mistakes in their respective pair, but never seen one post from the En-Ar pair Language Specialist. I have submitted two support tickets, 278276 and 280216, over a year ago. Until now, there is no single post from any of the LS' in my language pair.

  • 8
    Avatar
    Heyke

    Hi Gunnar, bias is more often unconscious than not, and your LS might even have a positive bias towards you by now. Furthermore, someone trying not to be biased might not produce a neutral review for this very reason. This is why double blind methods are being used in research etc. As an example, when I just googled the term "double blind", I came across the concept of "double blind peer review".

    By the way, I very much like Abdek's idea of sending a disputed review to another reviewer without letting them know that is has been disputed.

    Do you think there is any disadvantage for anyone in hiding the translator's ID, Gunnar? 

  • 1
    Avatar
    gunnarbu

    Hi Heyke,

    No, as I said above, that may be one way to secure better objectivity.

サインインしてコメントを残してください。