Weird anonymous feedback

Question

I had a look at the low rep and anonymous feedback here (might require some privileges):

https://french.stackexchange.com/tools/post-feedback/least-helpful?filter=all

It seems that questions with a high number of views (probably by users not familiar with StackExchange's ways) get high amounts of negative feedback. And this tendency is new, I never noticed it before.

A few examples:

• Quelle formule de politesse doit-on utiliser à la fin des documents ? — Which valedictions can be used to end a letter?
  (21k views, feedback: +7/-25)

• How to correctly say "my friend" in French to express a friendship relation? — Comment limiter la portée de « mon ami(e) » à une relation d'amitié ?
  (12k views, feedback: +7/-22)

• Jusqu'alors ou jusqu'à lors ?
  (8k views, feedback +6/-25)

I cannot find a good reason for this, except a problem with understanding the user interface by users who just pass by. I guess some might not understand the purpose of arrows (which are now used to obtain feedback from anonymous users). Another guess: maybe they fail to understand that the first post is actually the question, not an answer to the title?

Any other explanation? Are those questions really bad questions?

---

Follow-up questions: How come these questions don't show up first in the overrated list (specifically the third one)? Why do the figures found here (respectively +14, +9 and +11), which are described as “unique anonymous and low rep positive feedback”, not match with the previous figures?

Answer 1

This just looks like confirmation bias - in order to accumulate a large score, a post has to have a large number of views, so things on the "least helpful" page (which you linked to, thanks for pointing out this shiny tool, by the way) will generally have large numbers of views. If you look at the "most helpful" page, you'll also see lots of posts with high view counts.

The main reason for anonymous feedback to disagree with voting from the site's users is that anonymous feedback generally comes from search engine traffic, and the users who find a question (or answer) may have something very different in mind. The Google or Bing or Yahoo or whatever engine can turn up our site as a result that doesn't actually match the person's query, so it gets bad feedback. Anonymous users might also just be less tolerant of basic questions, ones they might see as stupid, or they don't really know what they're voting on (voting on question vs answer). Signed-in users, on the other hand, generally vote on questions according to whether they're useful and well-written, and on answers according to how well the answer suits the question. Also questions are probably voted on more than answers simply because they're at the top.

If you look at the feedback summary, 70% of the total feedback is positive compared to "our" ~97% positive feedback. I think this just means that anonymous users are more readily downvoting, and though they may be voting on their personal query experience more than on the actual question/answer, it doesn't mean they're wrong. But they're not overwhelming. And if I'm reading the Post Score to Anonymous Feedback part, most of the anonymous voting falls in line with how we're voting.