Blatantly inappropriate data analysis still counts as lying

BlogBrother Ethan Siegel just caught (yet another) scientist manipulating data to generate the conclusion they want... not the conclusion the data actually supports.

Exposing a Climate Science Fraud

Yeah, thats how we roll on SciBlogs.

But in my example, the scientists actively hid relevant data from other scientists they were speaking to/the journals they were submitting to/the patients they were preaching to.

In Ethans case, all of the data was happily and knowingly laid out for anyone and everyone to mess around with themselves. So either this Judith Curry character is a fantastic idiot, or she was just arrogant enough to think everyone would believe her, and no one would check for themselves. Ugh.

Also-- Sample size of two now... dont trust a scientist named 'Judy'...

More like this

Abbie, you're awesome for helping keep the science community honest.

If everyone can do their part to be vigilant against these kinds of scientific frauds, maybe, just maybe, we can get the actual scientific truth out there. The authority isn't the scientist, it's the science.

For what it's worth, even the pot-calling-the-kettle-black tactic of accusing me of disinformation for accurately representing her statements and point-of-view is getting blown apart in the comment threads of Judith's own blog.

How irresponsible! It's unbelievable the lengths people will go to presumably push a political agenda (though I suppose I really shouldn't be surprised). On the other hand, this why I think that even though there are bad scientists, the scientific process is worthy of our trust and confidence.

Three cheers for folks like ERV and startswithabang who expose the frauds to the scrutiny they deserve!

You typically learn in a 100-level statistics (or econometrics) course that a non-stationary time-series can have subsets that appear stationary. That Judy Curry would try to pass it off as a significant observation speaks to either rank amateurism or, as Siegel pointed out, outright deception.

The article Ethan is responding to is a tabloid article, and as I'm sure you can imagine, tabloid journalists don't always fully capture the nuanced viewpoint of scientists. For Judith's full nuanced view on this topic, I would encourage you to read this article on her blog:

A quick example quote:

In any event, identifying an AGW signal on this short timescale isnât useful. What is of interest on this timescale is whether natural variability (forced and unforced) can dominate the AGW signal on decadal timescales and produce a âpauseâ or a âstopâ. This is the issue addressed by Santer et al., searching for the AGW signal amidst the natural variability noise.

This post represents her full, considered view and if after reading this you still consider her to be a fraud or fantastic idiot, so be it. I would just discourage reading too much between the lines of her views as parsed by a journalist at the Daily Fail.

Hi Spence! That's you, right?

I'm waiting for Curry to go ahead and writes up that article that clearly states that -- when one does their statistics properly -- there's no reason at all to believe that global warming has abated, lessened, or paused. Because that is what the data show, when you do your analysis properly. That's what she needs to be saying if she truly believes that the "Daily Fail" misrepresented her; she should be clearing up the misrepresented part and should replace it with what the data best represents!

Instead, in the article you link to, I see her trying to change the subject, obfuscate, and answer an uninteresting question:

whether natural variability... can dominate the AGW signal on decadal timescales

that has already been answered, emphatically, as I stated in my "cherry-picked" article. (It's yes, and it's yes with a giant "duh" after it.)

that has already been answered, emphatically,

Yes, by Santer et al, the article she quotes. So now you're upset with her stating the obvious? *shrugs* OK.

But note she quite clearly claims that AGW needs to be detected on longer time scales (something I strongly agree with her on, and you do too I believe). As far as I can tell (and I MIGHT be wrong, I accept that) she does not believe the current "pause" is anything other than noise, no matter what the Daily Fail wrote.

I personally don't like claims of lying or fraud unless there is clear cut evidence. Even for Judy Mikovits - and clearer prima facie evidence of fraud you will not come by - I tried to be careful to caveat my language as "looks like fraud" and encouraged people to wait for the full investigation before coming to any conclusions. It's a serious accusation and should not be used lightly.

I would say the same here, but let's face it, there isn't even going to be an investigation here. I have not seen evidence which amounts to fraud, I'm not even convinced by Abbie's claim of "fantastic idiocy". My take is that she accepts global warming and absolutely does not believe the recent "pause" is anything other than an interesting research opportunity to look at natural variability.

I might be wrong. Just giving my viewpoint.

OK, I've got a comment in moderation but missed a couple of points from your comment, Ethan so I'll just add them here:

there's no reason at all to believe that global warming has abated, lessened, or paused

Indeed, and in her article that I linked she makes it clear that it hasn't.

That's what she needs to be saying

That is what she says in the article I linked, which was written after the Daily Fail article, but before your article was written.

"(It's yes, and it's yes with a giant "duh" after it.)"

Exactly. I read Curry's "full, considered view" and was decidedly unimpressed. It's not anything that a competent statistician would find non-trivial.

It smacks of someone trying to walk back an obviously erroneous statement without actually admitting error.

As everyone is aware, I'm am no statistics guru. But, Spence, I've read the article of hers to which you link. I am suitably unimpressed. I will detail my primary reasons, which I'll preface by saying I haven't cherry-picked. I've taken them in the order she presented them, in an article you suggest is a fair representation of her thoughts.

Her definitions leave much to be desired. Consider:

Warming: any change in temperature greater than 0.
Stopped: any change in temperature equal to 0, or less than 0.
Paused: any change in temperature that is greater than 0, but less than .2.

The ambiguity in the data she's chosen (more on that later) is in that precise range such as I understand her complaint. It is therefore an unwise definition because it sets up a lovely cafeteria from which one may order whatever one desires. All data that lies between 0 and .2 is defined as justifying the conclusion that there is warming, and that there isn't warming - simultaneously. If this is the zone of ambiguity, I fail to see how defining that ambiguity as simultaneously arguing two opposite conclusions is useful.

Why not just define warming as something beyond the noise to signal zone causing her frustration? Note she doesn't define it as indeterminate. She defines these values as determinative.

Do you see how this might present a problem? If not, can you explain to me how having the same data determine two contrary conclusions is useful?

Next up is her citation to Santer, et al. She quotes them as saying

Our results show that temperature records of at least 17 years in length are required for identifying human effects on global-mean tropospheric temperature.

The very next sentence she writes begins 'in this context' meaning in the context of the amount of time required for a proper analysis of 17 years or greater, it is 'not unreasonable' to choose a period of time that is below the required interval of time.

Actually, it is by definition unreasonable. By analogy, consider a friendly, well-behaved curve. At least 3 points are required to determine the equation of that curve. In that context, it is necessarily unreasonable for someone to say, well, I only used 2 points, and here's its equation.

Or, say, determining one's taxes owed. If a full year of income is required to determine what one owes for that year, it is unreasonable to decide to only count 9 months of that year.

Are these points some of what informed your suggestion of 'nuance'? If so, I'd be mightily appreciative if you can explain to me how this is 'nuance' as opposed to just sloppy?

Ah, I think I see what happened here.

You see, everyone in the UK knows that the Daily Mail is not a real newspaper, merely a racist middle-class tabloid with delusions of grandeur. For other examples of their science reporting, see the Daily Mail's ongoing project to divide all inanimate objects in the universe into things that either cause or cure cancer (various blogs keeping track of this one, example: Or to get a feel for their general philosophy, try one of the many Daily Mail Headline Generators (Sample: "Do Lesbian Asylum Seekers Cause Cancer?").

Occasional misunderstanding can occur when persons from other countries not familiar with this curiosity of British culture come accross articles from the Daily Mail, often on the internet, not realising that no sane person would ever take such an article seriously.

Subsequently, either this Judith Curry person is actually a quack (can you use that in a non-medical sense? I'm going to anyway) and can be safely ignored; or she has been misrepresented and failed to defend herself. Either way, you shouldn't take it seriously: it's in the Daily Mail.

I hope this helps clarify things.

Not sure where the fine line is drawn between lying and stealing, but it seems one Judy has crossed it. She'll soon have her day in court to prove otherwise. Check out today's Science Insider by John Cohen.

Never trust anyone whose name starts with a 'J'...