Power user Edit(orializing) controversial questions to make them weaker?

Question

I had a strange experience with having my question on a controversial topic edited by a power user. In some ways the question was improved by adding more details, but also subtly changed the meaning, expanded the scope of the question and made it easier to debunk. It felt like there was possibly an ideological motive to the edit, rather than just trying to add clarity.

I complained in the comments and edited my question back to include some of the new details but not make the broader claims. The poweruser closed my question due to an editing war

After having this experience, I clicked on the first related question and saw that it had been edited by the same power-user, out of interest I clicked on the history and saw similar "editorializing", adding discrediting context to links and broadening the claims being made.

My Original Question

I asked a question about if a graph is accurate (on the controversial subject of COVID vaccines), it claims to show a correlation between more covid vaccine shots and more hospitalization in NSW, Australia.

My original question was only asking if the graph's data and chart was accurate, not what it implies or proves WRT vaccine effectiveness.

But the question was edited including the following lines:

The graph has been referenced by Joel Smalley's blog and on anti-vaxxer Del Bigtree's show The Highwire.

I don't think that is relevant, who is Del Bigtree, does it matter if they're an "anti-vaxxer"? Labeling them an "anti-vaxxer" sounds like a way to discredit them off hand, and weaken the question by association.

Another added line:

Is there a correlation between COVID-19 vaccine doses and hospitalisations+deaths? Is there a causative link?

This changed the meaning of the question. I specifically didn't ask about causation because I hadn't seen that claimed. If this "anti-vaxxer Del Bigtree" or someone else made such claims, I think that's irrelevant, it certainly wasn't a part of my original question. This edit also makes the question trivial to debunk, by saying "Correlation doesn't equal causation".

Another question on the same controversial topic

In the related links to my question I saw this one which had also been edited by the same poweruser. Out of interest I clicked on the revisions tab and saw a similar pattern of edits which weaken the question: https://skeptics.stackexchange.com/posts/48559/revisions

Before

There are some claims that the flu vaccine injection has a positive correlation with COVID-19 death rate. For example, this site says:

After

There are some claims that the flu vaccine injection has a positive correlation with COVID-19 death rate. For example, the anti-vax site Home Vaccine Education Network says:

In this example, the user added context, but again they're editorializing, why is it relevant if the site is an "anti-vaxx site" that sounds like a cheap way to discredit whatever they're saying.

Again the poweruser subtly changed the meaning of the question so that it is basically asking for causation to be proved by a single population study, making the question easier to debunk.

Before:

What's the consensus on this? Is flu vaccine related to a higher vulnerability against COVID-19 in any way?

After:

Is flu vaccine correlated to a higher vulnerability to COVID-19? Is there a causal link?

Is this normal?

So my question is: What's going on here? Is this kind of edit(orializing) of original questions something that's commonplace and accepted here? Is it just that one user doing it? I was surprised to see it done to my first question, and also on the first related question that I happened to click.

Also I'm interested in what safeguards the site has if a power user "goes rogue" in this way, and uses their editing power to make questions weaker.

I do see that the edits have clarified and improved questions in some ways, and I'm sure that taken on a whole, that power user's contributions here have been overwhelmingly positive and aligned with the pursuit of truth. Also, I'm not wholly free of bias in this either, I think anyone who comes to this site will have opinions and biases.

Answer 1

So, there are two ways of addressing this question.

One is to accept your premise: That there is a rogue on this site, who has let their "poweruser" status go to their head.

Then the answer is flag their contributions for moderator attention. If it is a moderator, one of the other moderators will handle the flag.

If you are unhappy with the behaviour of a moderator or group of moderators, there is a "Contact" link at the bottom of the page where you can report them to the Stack Exchange Community Managers.

---

The other is to point at the poor evidence to your premise, and gently suggest that the meta-question itself is where the editorialising came in.

(In case anyone else is unclear, I am the user the OP is complaining about.)

The OP's question started out poorly. It didn't have any notability links (it was "floating around"). It included a chart with lots of numbers and lines, but didn't say what the claim was. It claimed to show a "unexpected negative correlation" (which was quickly corrected to "positive") but didn't explain what the correlation was between, and image itself didn't contain a claim of correlation. The tags were lacking too.

Overall, it was a fairly low-effort mess, and I could have taken the low-effort route and marked it as Closed, perhaps posted a welcome comment, and moved on.

Instead, I spent some time on it turning it into a real, answerable question. To do that, I needed to do a bit of research.

First, I found the original source of the image, LCHF Matt on Twitter.

Now, we don't investigate the ramblings of random Twitter users - they churn out more nonsense per day than any fact-checking site could hope to handle. We focus on widely-believed claims. So I tried to find out who he was: Is he an epidemiologist? Is he a statistician? Does he work in NSW public health? Is he a journalist? Is he a major Instagram influencer? Is there a reason the public would believe what he said?

After a few minutes, I didn't have answers to any of those questions. I imagine more time would have revealed more info, but as far as I could see, his entire claim to fame is that he has around 1,500 followers because he frequently posts... well, I didn't want to say "anti-vaxx" here, because I didn't see him explicitly state that vaccines are bad, and calling him "pro-COVID" seemed to be going too far, so I tried to stick with a more neutral position and introduced him as "an Australian who commonly tweets against COVID-19 vaccines and mask mandates."

I reject that this is poisoning the well. I think it sums up accurately why he has followers.

But I still had a problem with the question. Even 1,500 followers doesn't mean much. Why is this notable? Is this widely believed?

Luckily, I found the casinoeconomy tweet which showed this graph had been used on Del Bigtree's show. This was good evidence that the claim had been widely seen. But for the casual reader, I wanted to add context: Who is Del Bigtree? Why is his show significant?

While I didn't want to label LHCF Matt as anti-vaxx without some clear statement from him, I have no qualms labelling Del Bigtree as such. His Wikipedia page says in the opening box:

Known for: Anti-vaccination activism

The opening paragraph says:

Del Matthew Bigtree is an American television and film producer as well as CEO of the anti-vaccination group Informed Consent Action Network. He produced the film Vaxxed: From Cover-Up to Catastrophe, based on the discredited[1][2][3] opinions of Andrew Wakefield and alleges an unsubstantiated connection between vaccines and autism.

So (a) describing him as an anti-vaxxer is an accurate summation, and (b) it is very relevant to this question. The fact that this graph is being used to widely claim that COVID-19 vaccinations are dangerous - that there is causation, rather than just an inconsequential and easily explained correlation - is what makes this an important question; worth looking at and answering carefully.

I disagree that "anti-vaxxer" is poisoning the well. It is a statement of his position. If Del Bigtree is right, being an anti-vaxxer is a heroic act for challenging the orthodoxy, and he certainly has his followers who consider that he is right.

In my "rogue" and "biased" way, I then proceeded to delete a lot of comments where users made pot shots at the quality of the data. I added some of the fine print, to try to preclude more cheap shots against it. I downvoted and commented on an answer that didn't answer the question. If we are going to dismiss the claim, it should be done in a rigorous and referenced manner.

When an edit battle started, I did not let it turn into a war. I closed the question, pointed out the flaws, got clarification in comments, fixed it to match so we had consensus, and then re-opened it promptly.

I didn't weaken this claim; I made it as robust as it could be, so if it gets ripped apart, it gets ripped apart for good reasons.

Now there are several ways of addressing the claim that additional vaccinations are linked with additional hospitalisations:

• Checking that the data displayed is actually from the NSW figures, and has been transformed/represented accurately. The OP says that was his key question, but tried to remove the doubt that it was in the question by stating categorically that it came from NSW Health data. I soften that, because it was begging the question.

• Checking that, if this is the data, that there is a significant correlation. All the analysis I have seen so far is just people pointing at the pretty picture, rather than a statistical analysis. What are the error bars? Has this data been cherry-picked or does the relationship still appear in different time periods, with different filtering, in different jurisdictions? Has there been the equivalent of p-hacking in the selection?

• If this data is robust, then the fun begins. What confounding factors are there that might explain this correlation? Is it just spurious? What do the serious analyses seem to say. The claimant seems to simultaneously say it might be in the fine print, and yet seems to rely on it in his claims; is there a mott-and-bailey fallacy happening here?

• Another approach would be to treat this data as suspect, and to go ahead and look for other peer-reviewed analysis of the same issue. Do we see such linkage in other countries with the same vaccines, other states of Australia, other COVID-19 waves, other vaccination efforts? Ultimately, if there is a large body of good science showing whether or not specific COVID-19 vaccinations lead to increased hospitalisation (after other confounding factors are accounted for), then this evidence would be swamped.

It would be a legitimate answer to this question about correlation to address the causality question, or we are just playing a game of spurious correlations. Again, the fact that it is being used to make a causation claim is what makes this question interesting, and not just JAQing off.

I want to conclude with a warning: The word "Skeptics" in the web-site title sometimes attracts people who are most familiar with its misuse by denialists who call themselves "climate skeptics" or "vaccine skeptics", or "holocaust skeptics". They turn up expecting to be surrounded by like-minded people, and get horrified when their claims are academically dissected and dismissed.

This chart is clearly a spurious correlation caused by the focus on the early vaccination of the most vulnerable people in God's Own NSW (Oh, drat! It turns out I do have some biases!). The exploitation of the chart by pseudo-scientists to support their political beliefs is deplorable. Of course, I can't just say that in an answer without appropriate references to support it, because it would be too easy to (correctly) dismiss as merely my opinion, which is off-topic here.

If your goal here is to use this site to promulgate such nonsense, rather than contribute to skewering it, you are going to find that virtual everyone here will appear to be biased against you. That's not true. What they are really biased towards is scientific evidence to support claims. My edits were all in-line with the goal of making the context clear and the question interesting, important and addressable with evidence, rather than closing it as a mess.