If governments were serious about climate and nature emergencies, they would give the conferences clear and suitable names such as "Climate Emergency Summit" rather than the confusing "COP".

But clearly it's unrealistic to ask people to assess in huge depth all the previous work they cite.

So we can ask what sensible rules might look like, perhaps adapted for different fields and circumstances.

We can ask when citation is justified, and also what is justifiable to say about the earlier work when citing it.

Should researchers note a general apparent level of reliability/unreliability within the field, or which papers they have checked for comments on post-publication peer review sites?

I'm saying that suspicion of existing work is precisely what a scientific approach should include, especially in more plagued fields.

And to say that "scientists should be suspicious of existing work but the public shouldn't" would seem to me to be neither fair nor sensible.

Is there perhaps a significant difference between

1) which news sources people trust

(which might be unknowing)

and

2) which sources they say they trust, in surveys

(which might be influenced by, say, wanting a source to look bad or to maintain a self-image of loyalty to a political outlook)?

Is there perhaps a significant difference between

1) which news sources people trust

(which might be unknowing)

and

2) which sources they say they trust, in surveys

(which might be influenced by, say, wanting a source to look bad or to maintain a self-image of loyalty to a political outlook)?

So one important task, it seems to me, is consideration of what to advise researchers to do about citing.

Obviously, take reasonable care to avoid citing claims thoughtlessly - a question is what kinds of rules/guidance might help.

The Stockholm Declaration seems very good but omits a key, immediate issue:

In fields where evidence suggests a high proportion of seriously flawed papers, what should scientists and others do and say as regards trusting, citing and/or building on previous research?

bsky.app/profile/matt...

Suppose everything in the Stockholm Declaration is done well and relatively quickly.

Could we end up with a harmful and regret-inducing amount of good work on bad foundations?

(Good research - apart perhaps from the lack of enough suspicion of earler work - built on seriously flawed work.)

Bernhard Sabel writes in the Financial Times:

"Politicians and the media...should also be careful not to cast suspicion on science itself, something that could cause unforeseeable damage even greater than the status quo.

Yes, it could do that. But if 5-15% of papers are fake ...

I'm saying that suspicion of existing work is precisely what a scientific approach should include, especially in more plagued fields.

And to say that "scientists should be suspicious of existing work but the public shouldn't" would seem to me to be neither fair nor sensible.

... as he estimates, and more are severely flawed, then it's hard to see how some degree of suspicion isn't justified and necessary.

Of course he may mean trust in science as a whole, beyond areas which are in particular doubt because of fraud and error.

Bernhard Sabel writes in the Financial Times:

"Politicians and the media...should also be careful not to cast suspicion on science itself, something that could cause unforeseeable damage even greater than the status quo.

Yes, it could do that. But if 5-15% of papers are fake ...

How far are rules for this teachable/transferable?

Is it often largely a matter of judgement which combines, for example, knowing about flaws within subfields, in particular journals or types of journal, from particular institutions, and/or particular methods?

What is the role of intuition here?

So I'd be interested to know the practices of, say, expert methodologists and statisticians in deciding their approach to existing work in such fields.

What factors influence their decisions about what to cite and what to say about it?

It seems to me that many researchers may already have methods, perhaps to some degree unconscious, that they use in deciding what to say about earlier work in plagued fields.

Should they say, for example, what precautions they've taken to check the methods are appropriate in the paper they cite, or that they've checked for a well-known flaw?

We can ask when citation is justified, and also what is justifiable to say about the earlier work when citing it.

Should researchers note a general apparent level of reliability/unreliability within the field, or which papers they have checked for comments on post-publication peer review sites?