Yup.

On the upside, I now understand more fully what I was trying to deal with and why it’s pretty much a waste of time.

It is quite natural in partisan debate to see such accusations asymmetrically – on one side they’re fair comment, on the other an intolerable libel. And it is quite natural for those doing it to not even realise.

I don’t believe such serious charges should all be allowed to stand unanswered. However, I wouldn’t want them banned or deleted, either. That’s counter-productive. I think the most productive method of dealing with these is to talk them through.

I had hoped to help you all see that some of the intemperate things commonly said about sceptics can be legitimately questioned. Maybe it could reduce that asymmetry somewhat. I would hope that it would lead to a more civilised debate, in which we all listen more.

-

Incidentally, some other blogs have a way of moderating in which the problematic comments are put up but the problem parts snipped out – with a visible indication that this has been done (and if necessary, why). This has a number of advantages. It means the commenter isn’t left wondering whether the message has simply got lost, they get feedback on exactly what the problem was so they can modify their behaviour, and others in the debate can see that the person replied but the reply was blocked; they didn’t simply run away. It’s based on the same sort of blog ethics as clearly marking updates and significant edits to a controversial post.

Yes, I’d say so.

If somebody isn’t understanding a point then it might be because it wasn’t explained clearly, so it’s usually worth having a few more goes. (I thought it was here.) If ideology is getting in the way, then it’s a useful exercise (and also for anybody else watching) to use the irresolvable disagreement to refine the presentation of the points. It’s verbal sparring – it doesn’t have to get anywhere to serve a useful purpose.

This subject constitutes a major split in society on an important policy decision, which is never going to be resolved if we don’t talk to one another. You have said it yourself on the subject of theists and atheists. You have to talk, and keep on communicating, even if it doesn’t appear to be getting anywhere. You can’t just give up when the other side doesn’t make an immediate conversion. It takes time.

“There was no intention that these should be interpreted as real numbers or that the actual significance should be calculated. It was an analogy. Perhaps I should have explicitly stated this, although I thought it was pretty obvious.”

This is heavily ironic. We’re having a conversation about how it is unacceptable to paraphrase somebody’s technical argument even slightly, and yet I’m supposed to figure out that when you said A you “obviously” meant B. I have no objection to that, but I would like to point it out.

If I understand the point you was making, you wanted to give an example where an effect was very obviously and intuitively significant and equate it rhetorically to the under-95%-confidence case that wasn’t, but which you would like us to treat as if it were. You appear to be trying to give the impression that statistically non-significant results can still be very strong evidence. If I thought you knew it was wrong, I’d call it deliberately misleading. But I am being generous in offering an alternative.

Here’s a better analogy. You toss a die three times, and it comes up 6 twice. That gives us a 92.6% confidence level that the die is loaded for 6s. The results are consistent with the die being loaded, but this isn’t very many rolls, so it does not quite reach significance at the 95% confidence level. The analogy now gives an accurate impression of exactly what 95% significance really means. And I have to say, I think most people in a casino who saw a die roll a 6, a 2, and a 6 wouldn’t immediately leap up and accuse the casino of cheating; that they are 92.6% certain of it. I don’t think anyone reasonable would violently object to saying the sequence 6, 2, 6 was not ‘biased’ in favour of 6s. Maybe I’m wrong – you could ask and see.

“The problem here is that Jones didn’t say that there was no warming”

Since quoting it with the wider context didn’t work last time, I’ll zoom in a little. “Do you agree that from 1995 to the present there has been no statistically-significant global warming?” – “Yes”

Can you see it yet?

Now try to see it like a tabloid headline writer would. Brevity is at a premium, you’re only supposed to report it if it’s news, and you know the level of your readers.

There was no intention that these should be interpreted as real numbers or that the actual significance should be calculated. It was an analogy. Perhaps I should have explicitly stated this, although I thought it was pretty obvious.

Is the problem here that you don’t actually know what significance levels are?

I understand significance levels perfectly well, thank you, and that is not the problem here. The problem here is that Jones didn’t say that there was no warming, the Mail said that he did, and you won’t admit it. That’s the problem here.