2
$\begingroup$

How can I improve this question that has been downvoted? and now flagged as a duplicate of my previously closed question?

The down vote help says "This question does not show any research effort; it is unclear or not useful."

I have spent hours trying to formulate this question ( as evidenced by the linked closed question.) Thus I discount lack of research as the reason.

Thus how is my question unclear?

$\endgroup$
6
  • 3
    $\begingroup$ As a meta-comment, attitudes vary on whether it is good practice to declare oneself as a downvoter and/or the precise reason for downvoting. Naturally the advice is to downvote for the reason given by a mouse-over, but otherwise people are human. Some downvotes I get for answers can only be decoded as someone not liking me or how I wrote, but as Tolkien wrote in another context, I probably would have similar opinions about the downvoter. Other way round, I will willingly downvote spam or very rude questions or answers, regardless of whether anyone has explained a post to be unacceptable. $\endgroup$
    – Nick Cox
    Commented Oct 25, 2022 at 12:46
  • 1
    $\begingroup$ Downvoting without an explanation is a lot like gaslighting, with a similar likelihood to lead to a positive improvement. $\endgroup$
    – dipetkov
    Commented Oct 26, 2022 at 10:46
  • 3
    $\begingroup$ @dipetkov As said, attitudes vary. It's a good ideal to explain to people the errors of their ways, but that is not always appreciated. I don't expect positive improvement from spammers. I just want them to disappear. You've perhaps not had the personal experience of being hounded on- and even off-site by people bearing a grudge against evident and explained downvoting. (Sure, I make myself fairly identifiable, but that is my version of openness too.) $\endgroup$
    – Nick Cox
    Commented Oct 26, 2022 at 16:24
  • 2
    $\begingroup$ @dipetkov Gaslighting is deceit committed with intention to fuck with someone's trust in their own mental faculties. Downvoting simply reflects a discredit of a Q or A by the community. Describing anonymous downvoting as gaslighting is absurd, and demonstrates poor understanding of downvoting, gaslighting, or both on your part. $\endgroup$
    – Alexis
    Commented Oct 31, 2022 at 19:02
  • 2
    $\begingroup$ @Alexis Thank you for your feedback; I admit I should have written "feels like", not "is like", it's a big difference. Your comment made me think whether there have been studies on this/similar topic and I came across Emotional consequences and attention rewards: the social effects of ratings on Reddit. Okay, it's about Reddit but the conclusion feels right to me: "... defection from normative community values (as indicated by downvotes) harbors emotional consequences..." $\endgroup$
    – dipetkov
    Commented Oct 31, 2022 at 21:08
  • 3
    $\begingroup$ Research shows that downvotes are discouraging and upvotes are encouraging. Who would have guessed? CV is one of the kinder and gentler parts of the internet: a decent question stands a good chance of getting volunteer time and effort from people with expertise who are likely strangers to you. It's in practice quite hard to remove very poor questions here. I am mindful of the likelihood that some people asking may be struggling with poor mental health and I really don't want to make anyone feel really bad about themselves. But a poor question (as decided by the community) is a poor question. $\endgroup$
    – Nick Cox
    Commented Nov 1, 2022 at 9:11

1 Answer 1

9
$\begingroup$

Your question wasn't closed as unclear, it was closed as a duplicate. The duplicate question has the answer to the question you ask, right in the first comment:

Substitute "H" for "E" in the tower formula for conditional expectation: there's nothing new to memorize.

This is a complete answer to the new question. If you read further in the comments, we find the term of art "Law of Total Expectation" and a link to a Wikipedia article.

While I did not down-vote this question, it seems to fit the description "This question does not show any research effort."

$\endgroup$
8
  • $\begingroup$ I guess in isolation this explanation makes sense. But I really have difficulty understanding why the question by @Kirsten shows lacks of research effort while this question: var() is not equal to sd()^2 in R got 3k views, 13 upvotes and 4 answers. The categorization of questions is very inconsistent, noisy & sometimes unfair, in my opinion. It basically depends on the user that first spots a question. $\endgroup$
    – dipetkov
    Commented Oct 27, 2022 at 10:35
  • $\begingroup$ @dipetkov I agree that the question you link is poorly researched, and it's disappointing when poor questions get so much more attention that high-quality ones. Part of the reason that it got so many views and votes is that it became a Hot Network Question. Since you correctly identified the issue as floating point precision in the first comment, one thing that you could have done to to forestall it becoming a HNQ is to identify a suitable duplicate about floating point precision and vote to close as a duplicate. This is the neat thing about Stack Exchange: most users can moderate somewhat. $\endgroup$
    – Sycorax Mod
    Commented Oct 27, 2022 at 12:55
  • $\begingroup$ @dipetkov As a purely general observation, the questions that get the most views and upvotes aren't always the best questions. In some cases, they are the questions that have a "clickbait" title ("Is xgboost Taylor Swift's favorite machine learning algorithm?") and an easy answer that is obvious to someone with a journeyman's knowledge. This is a problem with how people interact with Stack Exchange sites overall, but especially HNQs. $\endgroup$
    – Sycorax Mod
    Commented Oct 27, 2022 at 12:58
  • $\begingroup$ Thanks for the advice, I'll keep it in mind and try to use the moderation tools more. I admit I struggled with that question (mostly because the questions I find most rewarding never get that much involvement). Then decided it's best to be chill about it. Since I'm writing these comments, obviously haven't achieved zen yet. $\endgroup$
    – dipetkov
    Commented Oct 27, 2022 at 13:50
  • $\begingroup$ @dipetkov Yeah, I definitely know that feeling well. I've seen a number of poor questions get undue attention. We're all here to learn and have fun. Whenever I feel like I'm not learning or not having fun, I step away to do something more rewarding. $\endgroup$
    – Sycorax Mod
    Commented Oct 27, 2022 at 14:01
  • $\begingroup$ By the way, I've noticed that a question tends to become a HNQ once it gets two answers. And then it becomes a self-reinforcing loop: more views, more answers ... until the weekend is over. Something's not quite right with the SE ranking algorithm. $\endgroup$
    – dipetkov
    Commented Oct 27, 2022 at 14:01
  • $\begingroup$ @dipetkov Yeah, it's definitely a source of some consternation. There have been a number of threads about this over the years: meta.stackexchange.com/search?q=hot+network $\endgroup$
    – Sycorax Mod
    Commented Oct 27, 2022 at 14:11
  • 2
    $\begingroup$ PS: I did an experiment this weekend & did lots of voting to close bad or duplicate questions. I wanted to see if I can be helpful (to others but also to myself, to help me learn & improve). I discovered that many OPs didn't appreciate my feedback and, all in all, I didn't enjoy doing it. So I'll leave it to those that are good at writing useful comments. And my respect for you, the mods, has really grown; it's a tough job. $\endgroup$
    – dipetkov
    Commented Oct 31, 2022 at 10:46

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .