1

We see a lot of substandard questions here. Some are more troublesome than others. Some good questions get deplatformed for being mistaken as "push" or whatever.

This meta post is about clearly substandard questions being pushed onto the Hot Network Question list apparently solely for the opportunity to push a virtue signaling answer, even if that's somewhat oblique to the vague claims presented.

'Pushed' here means that enough regular users upvote quickly a substandard post without making sure that it is indeed a 'good question'. That HNQ-post is indeed chosen by an automated algorithm but also a result of collective action, or rather: inaction and neglect.

Notice that it was the first revision of this question that ballooned onto the HNQ:

Is the erradication of the Polio Virus directly caused by the production of pesticides such as DDT

Not enough close votes, not enough editing into shape, and that despite a bag of problems that is filled to the brim:

Typo in title, mismatch of unclear title and body, misspelled author of a book that's presented as being the 'real' source for claimS. No quote given for the content of the book. Apparently not even access to the book contents by QP. Just a link to a video of unknown provenance (at least it's a short video).

First claim presented is "wasn't as virulent as thought", then follows a list of five separate claim_s_.

One claim from that bullet list commented on by QP with a link (the FDR (mis-)diagnosis) — that promptly says the exact opposite of what the QP comment asserts. This was only found out to be that contradictory in nature — after 10 regulars voted this into HNQ — in a comment, where that info rested for a while before not QP but I myself tried to edit out this inconsistency. Those regulars upvoting the question in that state obviously didn't check the links.

Then follows a graph that's not attributed transparently. The question made it look as if from the book. It is really taken from a link that QP asserts would be like the book (in question presented quite understated just as "others"). But we have no way to tell if that's true or a misrepresentation, as the content on that link never mentions said book, except for in comments on that page. But that comment there seems to indicate that webpage and book are really quite different. The graph from that website then is taken seemingly at random, as it is from a series of similar graphs building up to the argument on that webpage that it would be the combination of effects that might explain something, not just the observation from that one graph in isolation. The webpage has a compound effect graph, supposedly showing what the webpage author concludes, the question leaves that out — thereby misrepresenting the content of the argument on that webpage. And which claims do we address now? Those from the unquoted book or that website?

Then we get an explanation from QP on the supposed purpose of that graph:

The purpose of this presentation from the author, is to show that the Polio vaccine wasn't as effective (if not effective at all) as once claimed. Many anti-vaxers are using this as a reason to prove that vaccines aren't actually as effective as claimed.

Which author? Without digging it would seem to me 'the book author'? But it is taken from the other website. Do we know whether the book author uses the same or at least a closely matching graph? Are book author and webpage author really saying the same thing? I don't know from the question and can't ascertain either way without any quotes or access to that book. And as we see, no-one, neither QP, nor regulars liking the question, has accessed that book. All reference to that book on tis site are based on some level of hearsay. That's bad.

We notice that regulars are content with it, do in their mass not click on links and do not follow up what's claimed. That's reason for concern.

Because that first graph again is a also a misrepresentation of what's actually said. That website goes on to claim that the graphic would show that the commonly attributed virulence and disease effects would show a correlation and explanation between chemical poisoning and paralysis. Thus a large and previously neglected confounder. Not that eitehr 'there is no virus' nor that the vaccines would be completely ineffective, rather that they would present a sort of solution to a problem that wasn't caused by any virus alone.
Although that website then goes on to indeed construct an argument as QP presented. But crucially — also in light of the warmly received answer — at this point QP jumps ahead and thus distorts the source with his foreknowledge and interpretations, leaving out too many steps.

The question then goes on to mention pesticides, giving two examples (arsenic & DDT; remember that the first and lonely graph only showed two DDT correlates and polio).

The explicit subquestions we read are:

  1. "Is there any research into these claimS? __[plural emphasis added, LLC]
  2. Specifically, if Arsenic, DDT, or other pesticides used during that time caused paralysis similar to that of the Poliovirus? [Question mark added, LLC]
  3. Has there been any research into the lowering of Poliovirus cases directly related to the reduction in the use of these harmful substances?
  4. Finally, is there definitive proof that the vaccine is directly tied to the "eradication" of the Poliovirus?

Apart from obviously asking multiple questions, and not 'challenging one notable claim', these questions also frame the problem in peculiar ways:

  1. Will have to be answered with a yes: QP itself presents such research from apparently two sources. The website linked to contains 24 references. Are they relevant, professional level, good quality etc? That's another subquestion for this one…

  2. This asks if there are other possible causes, confounders of diagnosis, for poliomyelitis paralysis. In short: is it possible/true that (related to bullet point 5 at the start: "parents pushed for a Polio diagnosis due to the funding") the case numbers also rose due to wrong incentives for posing a diagnosis?

  3. Confounds 'cases' with paralysis, ie different disease outcomes—for allegedly different problems (according to claims). As one comment expressed it:

    misrepresentation of the actual source. It claims the effects of polio were caused by other factors as were their decline not that polio was eliminated by other factors. [And it is indeed an established fact that polio must have been around for far longer but only started to be a pandemic-like problem with often much dreaded outcomes from first description in 1860 over to 1880, when severe cases took off, LLC]

  4. What I read for the 'book as claim source' this again looks like misrepresentation, as it seems that the book author seems not to claim that the polio vaccines were totally ineffective, but that they would have played a much "smaller role" in freeing humanity of debilitating effects attributed to the virus and the resulting disease alone. That those other factors contributing to virulence were overlooked despite being really quite large.

    Epidemics of paralysis began to rage in the early 1900s, seemingly out of nowhere. Doctors, parents, and health officials were at a loss to explain why this formerly unheard of disease began paralyzing so many children (src)


So which of the many claims do we address? If we allow multiple questions in one post we must address all of them in an answer. Are they from the website, the book, or mainly 'made up' by the asker — meaning: misrepresented, misunderstood, or maybe even manipulated (I do not say that here it is the latter, honest mistakes happen all the time). I have no way of making sure from the question alone, not even when trying to dig around the net & literature myself, that what I read in question is what the book says. Do we chase an imaginary rabbit?

Such a bunch of quality problems, quite a few of them noted rightfully in comments, should not be allowed to stay in that way on HNQ. Any answers on such questions run a high risk of making things worse.

As the much loved answer shows: it focused on DDT alone (not pesticide usage as the wider claim), ignores the confounder issues (as the claim seems to assert those to be crucial), and finally presents a testimonial for 'vaccines for polio do work' (while the claim seems to not deny that, but describes just different weights for contributing to observed effects?). In short: the answer ignores the complexities of the claims and in effect largely talks past the claim, but all the while saying 'vaccines are great'.


The problem:

We get a bunch of claims, have no way to ascertain who really claims what and whether the question accurately present those claims. Thus we have no focus in the question and answers are free to pick one or a few and then just run with those, as the answers also didn't bother to assure those quality issues to be addressed — whether in question or the answer itself.

Thus we see highly upvoted posts on HNQ, that would achieve probably not much for actual readers of that book who are impressed by its contents into convincing them otherwise (if that would be warranted, and I can't really say…)

Note that the quality problems of that post are really secondary for the meta-post. Although everyone reading along here is really invited to take the criticism and indeed improve the specific example post on main…
The primary problem is about the collective efforts or lack thereof and the resulting detrimental effects in general. This is not the first time to observe these effects.

Why can't we fix questions with so many problems in time but instead send those as 'representatives for the site' onto HNQ?

13
  • I've only read the first few sentences, but I want to check you understand how questions end up on hnq. It's formulaic, not intentional. No one pushes it there. Mods can push it off however.
    – fredsbend Mod
    Aug 6 at 16:23
  • @fredsbend Well, I also do not have an exact formula. But I know that it's pushed by (regular) users by upvoting quickly, with UV'd A present. In this case: 9 or 10 upvotes on rev1. Then the formula picks it up… If that's not 'a push' as a result from the collective action (& here quite some ignorance on the issues pointed out), then I'm open to vocab alterations, but it is not a random formula. More like reward for popularity quickness, awarded by users (seems that as few as 2 on most sites I frequent within the hour (~up to 6, rarely after 24h) suffice, here we had even much more.) Aug 6 at 17:05
  • And that's one half of the criticism. The other is that we users and mods might prevent the HNQ mayhem by closing it, fixing it, then RO, so it may re-enter the spotlight (mod HNQ removal is permanent, so a less desirable option). I tried all tools for closing & fixing I saw, but everyone else seems quite fine with this heap on HNQ, despite all the distortion this regularly ensues. Aug 6 at 17:08
  • @fredsbend Looking at the timeline now reveals mysterious entries: close votes started 12:08, then were invalidated 22:05— by a mod-close 22:05 and mod-reopening 22:09 (no action on content of post), then RO votes got invalidated 22:11? What's that going on there? Even the CV-review queue 'sabotaged' (in effect, perhaps by quick change of mind or mistake?)? Aug 7 at 11:14
  • 1
    Idk, wasn't me. Will check on later
    – fredsbend Mod
    Aug 7 at 21:04
  • Votes are far from the only thing to get a question on the hot questions list. meta.stackexchange.com/questions/11602/…
    – Joe W
    Aug 9 at 12:50
  • @JoeW If you have a more complete description of the currently active algo, please share it. AFAIK, it's mainly UV's for Q, A present, also UV'd, views, all in short time of a couple of hours. ((MIN(AnswerCount, 10) * QScore) / 5 + AnswerScore)/((QAgeInHours + 1) ^ 1.4)). A net-score of 0 or a closed Q don't go hot. So, how is your comment relevant for this? Reception by fast voting regulars is what propels HNQ, is it not? Aug 9 at 12:58
  • A more recent Meta with speculation about the secret formula. Why so secret? Maybe they don't want people gaming the system, maybe everyone likes a puzzle, or maybe the code's so impenetrable that no-one's quite sure anymore after more than a decade of tweaking. Aug 9 at 13:02
  • 1
    @ARogueAnt. The basics of these algos still apply (as said: not privy to exact algo)? Point here is that we need to edit-into-shape before HNQ (really: before UVing!?), or do a fast close/RO-cycle (with sufficient improvements), as no matter which exact sauce is used, closed≠HNQ is no secret? Which means we can predict a bit/enough out of the black box, and use the system to improve what's shown on HNQ (which also causes the most drastic distortions in voting: multiplying the (IMO) spurious to begin with UVs on substandardQ by regulars)? Aug 9 at 13:25
  • Of course I don't have more info on the secret formula I am just adding a reminder that there is more to a hot network question than the number of votes it gets.
    – Joe W
    Aug 9 at 13:46
  • 2
    Honestly if you are so concerned about the quality of a question and the errors it has why don't you do more to fix it?
    – Joe W
    Aug 9 at 13:47
  • @JoeW 1st: it is about the collective failure 2nd: I did, voting, flagging, (CV was even 'sabotaged' as shown in timeline & here in comments…), commenting, highlighting issues in chat, editing (as much as I am comfortable keeping QP's intent, need more cooperation for more drastic edits?) and finally this meta post. What's your idea for "do more" then? Aug 9 at 13:54
  • What is my idea? I am not the one complaining about a bad post while doing almost nothing to get the post fixed up. You made a couple of comments on the post but don't seem to post much compared to the issues you raised in this question.
    – Joe W
    Aug 9 at 14:12
2

Frankly I find you long rant quite a bit off the mark. It would have been better if the Q quoted more of the source instead of just the graph, but here's a quote:

A clear, direct, one-to-one relationship between pesticides and polio over a period of 30 years, with pesticides preceding polio incidence in the context of the CNS-related physiology just described, leaves little room for complicated virus arguments, even as a cofactor, unless there exists a rigorous proof for virus causation. Polio shows no movement independent from pesticide movement, as one would expect if it were caused by a virus.

In the laboratory, however, poliovirus does not easily behave in such a predatory manner. Laboratory attempts to demonstrate causation are performed under conditions which are extremely artificial and aberrant.

There's little doubt the author of that weston-a-price-foundation page ["Jim West"] intends to seed serious doubt that polio is even caused by a virus. And they put extreme effort into showing that DDT causes "polio like" symptoms as the alternative/"real" explanation.

If there's something to quibble about is that the authorship of the article is anything but clear; it seems to have come down the grapevine of denialism, as it also carries these notices:

Copyright: ©HARPUB 1985, 1998, 1999. All Rights Reserved. Website: www.geocities.com/harpub. Email: harpub (at) hotmail.com.

This article appeared in Wise Traditions in Food, Farming and the Healing Arts, the quarterly magazine of the Weston A. Price Foundation, Fall 2002.

But the message inside the article isn't at all in doubt.

The Maready book, although I can't find any direct quotes from it on the web, is probably not being misrepresented either:

So intent on the germ or virus theory, Maready contends, researchers might have missed an alternate cause or factor. His prime suspects: The pesticides DDT and lead arsenate. He contends polio vanished not so much because of a massive vaccination campaign but because these toxic substances were finally banned.

Yeah, the question could be edited to remove some of distracting remarks, like the one on Franklin Roosevelt's case, but it's otherwise a fairly solid question.

4
  • Thx for confirming all my complaints about substandard quality: "would have been better", unclear authorship, "message of the article" while no-one has access to that book_— about which you now speculate over a review. What's the _precise claim? In which publication? Why include book at all? Notice your dichotomizing interpretative reading your 2nd hand src: "not so much" means the reviewer reads from book about a multi-causality theory, while answers read that as "there is no virus, it's toxins alone". There is quality & even logic missing from Q&As. (No joke: claim perhaps as well ;) Aug 9 at 8:38
  • @LangLаngС: to repeat myself... mountains... molehills. The 1st quote and the graph from "jim west" are enough to make a good Skeptics Q, IMHO. I have certainly seen much worse political drivel promoted as a top question on HNQ here on Skeptics.
    – Fizz
    Aug 9 at 12:17
  • @LangLаngС: aside: some users here are really bad at writing good, focused Qs as well. I can think of some who write very divagating ones full of [political] side-points on their 5th or 10th question here. This was a Q from a fairly new user, by the way. Instead of this lengthy critique here, of which you didn't even notify the OP [don't worry, I've done that for your now], you could have edited the Q some more...
    – Fizz
    Aug 9 at 12:25
  • Yo, molehills are indeed like icebergs, most of the quality work being invisible and somewhat 'below'? ;) — Which graph again of the two? And was (or even "is") that quote in the question on HNQ? If not why not? You say 'with that quote would be good', but it's still not there—so you agree it's 'not good'? (Plus the other issues…) How is the inaccessible-to-all book to be treated? Is that Q about the "JimWest"-page, known to Q-readers only as "several others"? // My point is less about the actual Q (if you see possible edits, do them), but improving Qs enough before they go hot. Aug 9 at 12:28
2

This question includes a number a premises (which shifted after several edits making it hard to write an answer). Let's start with the agreed facts.

  1. Aug 5 at 3:53Z: A question was asked.
  2. Aug 5 at 6:38Z: An answer was provided.
  3. Some early readers presumably viewed/upvoted and/or downvoted it.
  4. Aug 5 at 10:57-11:48Z: Comments appear criticizing the quality of the question. It receives a close vote.
  5. Aug 5 at 11:54Z: The automated algorithms decided that it was suitably good/controversial by whatever metrics they use to put it in the HNQ list.
  6. Aug 5 at 12:01Z: It receives a flag suggesting it should be closed.
  7. Several edits are made.
  8. I come along, and see the comments saying the question isn't notable. I close the question, and prepare to dig deeper and edit it.
  9. I look at the edit history, and see it has substantially been addressed, and the claim does appear to be notable. I look at the top answer, and I don't wish to invalidate it. I regret closing the question, and reopen it four minutes later.

Now let's look at the various conjectures being raised by this meta-question.

Are we ignoring quality problems blissfully if we see a chance to fight some strawmen from 'the other camp'?

Errr... That is a loaded question with many side-claims.

  • Is there "the other camp"?

If the other camp is "pseudo-science" or "belief without evidence", perhaps, but there are plenty of regular users here who try to practice skepticism and yet have different views on many political issues, so putting everything in one camp or the other camp is a false dilemma fallacy.

  • Are we fighting a strawman?

The original edit of the question failed to appropriately quote from sources. It failed to demonstrate notability.

But it appears that there is a wide-spread belief that polio symptoms were caused by pesticides rather than the virus. So, even if the OP failed to demonstrate the notability, it was still present. Ultimately, this does not appear to be a strawman claim. So, No.

  • Are we ignoring quality problems?

No. There were multiple edits made within 48 hours to improve the quality. Inappropriate answers and comments were weeded out.

  • Do we see a lot of substandard questions here?

Before or after people edit them? I think our standard is fairly good for a crowd-sourced site. I compare what we have here to what I see on Quora, Reddit and (in its day) Yahoo Answers, and I am proud of what the community produces.

  • Do good questions get 'mistaken as "push" or whatever'?

I don't think so. I don't really think I understand the claim. The question under discussion isn't evidence of such a thing.

[Since I started writing this answer more edits to the question have been made, it seems "pushed" just means "upvoted without meeting your standard". I will come back to that.]

  • Do good questions get 'deplatformed'?

By normal definitions of the term, no. [Deplatforming] came from the concept of universities not wanting to lend their imprimatur to unqualified speakers.

There is no requirement to have qualifications. There are no taboo topics that have been identified. We expect questions to be respectful, to be notable, to be answerable. But they can and do challenge claims from across the political spectra on a wide range of issues.

[I note the irony that you seem upset that the HNQ is putting the Skeptic.SE imprimatur on questions that don't deserve it, and you would like those questions not to make it onto the HNQ.]

I understand from previous discussions, you consider include downvoting to be deplatforming. I dismiss that as too broad a definition to be interesting.

But it raises another question:

  • Do good questions get downvoted?

If we define "good question" according to our individual subjective opinions, of course. None of us are going to agree with the general voting 100%.

If we define "good question" as Stack Overflow does - highly upvoted questions - it is a tautology.

So what definition could we use here that would give us a useful answer?

  • Are questions being pushed on the HNQ to "virtue signal"?

No.

Virtue Signalling is a loosely defined, ad hominem attack questioning the motives of someone who disagrees with you, without evidence.

There is no evidence here that anyone is using an anonymous voting system to show others their good character. There is no evidence here that anyone is voting against their personal opinions on the questions.

The claim is without merit.

Phew, that's the first "paragraph" dealt with

  • Were there "Not enough close votes, not enough editing into shape" of a problematic question?

There was plenty of editing over 48 hours.

Ultimately the claim seems notable and got a reasonable answer in a short time, so closing it (for longer than a quick edit) would have been counter-productive.

Ideally, it would have been quickly closed, and quickly edited, and quickly reopened, before the first answer was written. Once an answer has been written, it becomes harder to justify any changes that might invalidate the answer.

[I only skimmed a lot of the complaints about the question here. I don't have to agree or disagree with the individual flaws; that can be dealt with edits/comments on the question.]

  • Did some users upvote a question that had flaws?

[I think this is the substance of the meta-question here. It is an incredible long question that touches a lot of topics, but this seems to be the essence: that a question you didn't like got upvotes.]

It seems like the answer to this is yes.

Now what?

I am open to hear specific suggestions of how we can persuade regulars to use their upvotes to reward well-constructed questions, to close badly-constructed questions (and not to downvote good questions about false claims). (If we could stop the annoying pseduo-answers and political opinions in comments, that'd help too.)

But I didn't see any suggestions here.

  • Can we be sure to fix questions before they go on the HNQ?

The algorithm rewards recency. So unlikely.

Fixing a question in 48 hours seems pretty good to me. Fixing it within 8 hours seems like it is asking too much.

Without a complete and total overhaul of the HNQ and the voting systems on Stack Exchange, I don't see any opportunity for this to improve, and given the success of the HNQ in general in increasing site usage, I see no motivation for such an overhaul.

Ultimately, the HNQ is a force for good as well as bad.

Sure, it rewards clickbait. Sure, it brings thousands of people here who aren't skeptics (as we define it) and don't know how to answer, assess answers, or to vote. Oh, and the comments! Ugh.

But along with those thousands, it brings hundreds who do, and some of them remain behind and become regulars and grow the number of good answers we get.

6
  • Was told to not include 'suggestions' in metaQs, to keep such Qs clean. MetaQ's long cause the example Q has so many flaws (your "a lot"). No good Q, As tend to miss. Essence is not "didn't like", it's "still bad". Q was edited, but is far from fixed. Eg: Nobody, not even QP, read 'the book', can quote it. It only takes ~3 regulars to UV bad Q&A for HNQ, distortions incl. Last line works also both ways: those attracted by lowQ HNQ-As staying also post a lot of lowQ-As. / Suggestions: eg extend FAQ to guide on voting? Establish policy to try to CV/RO lowQ, more reminders in comments… Aug 9 at 21:36
  • CV was not for 'lacking notability' but "being unclear", and imo still is at rev7. "Other camp": if Q or A seems suited to whack [label] (eg antiX, Y-denialist, Z-sth, politician-ω, etc), then guards are lowered around here. (Cf A that just runs w/ the DDT angle alone…) Part of 'essence' is also: 10 UVs came in —obviously—without check'n eg the link for FDR. Not tautology: a contradiction to 'the model', breaking it, enabled by HNQ: even low quality gets more UV simply from being fast & being featured. Let's feature good ones! If that means more CV/RO, or more editing, then let's try that? Aug 9 at 21:59
  • 1
    "Sure, it brings thousands of people here who aren't skeptics (as we define it) and don't know how to answer" Which is opportunity to teach them about skepticism, which doesn't just make this site and the internet better, but the world. More skeptical minds operating in the wild will lead to better societies.
    – fredsbend Mod
    Aug 10 at 15:07
  • @fredsbend Yo, but not so much or not as easy if we don't try harder to make posts on HNQ better/more precise/a stronger "force for good". Without the imo still necessary increase in qualy (like exampleQ: 'd be fine if better) we fail to deliver 'the best'. 'Without HNQoverhaul' I still suggest to increase qualy-demands/thresholds for HNQs, with eg a tighter CV/RO. [A mere 'increase in user numbers' attracted by too low qualy posts (baity/fastest) is alone a weak metric, as eg increasing mainly weak posts is surely not optimal]. Plus it's known that HNQ itself distorts voting, massively. Aug 10 at 17:48
  • Frankly it's more worrisome when crappy & superficial answers that agree with the "mainstream POV" (of the userbase) get a lot of votes. I'm pretty sure we can think of examples of those...
    – Fizz
    Aug 29 at 1:26
  • Come to think of it; this also came up in another recent question (that was fixed though): is it mandatory to include a quote? Some questions only provide a summary of the claim (in the poster's own words), and that's where there's trouble... including in the case of this Q...
    – Fizz
    Aug 29 at 1:36

You must log in to answer this question.

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