This is a guest post by Philip Moriarty, Professor of Physics at the University of Nottingham
My pal Rapha and I have been craving for stripy data for too long of a time. I had first approached Francesco Stellacci (FS) to ask amicably and professionally for the data behind a few of his controversial papers. The guy did not reciprocate (he may have thought that I am not an amicable professional; and he is right, I can’t be amicable with a deceptive metascientist; but a little lie doesn’t hurt, right?).
Then I told him that I have the right to get the data. Actually, everyone has the right! For the policy of Nature Materials is that you can ask for data to any author of that deceitful magazine, and they should give it to you ipso facto. Which means that FS had to give it to me. No courtship, no pursuit, no nothing. Give it to me, amigo!
Nope. He didn’t play ball. So then I had no other option that try to extort those guys at Nature Materials. Even Rapha’s boss, Dave Fernig, Retraction Watch aficionado, helped me out. There is no harm in extorting despicable people; first, Rapha Z’s mission to save science is a good cause to fight for; second,the editors deserve it, for publishing lots of fake stuff from FS. And third, I am Professor Moriarty!
Now, finally, FS has uploaded the data. I can’t imagine how he has been able to locate data from almost a decade ago when I struggle to find files I saved last week. Of course, he must have generated fakes! In fact, his STM data does not stand up to scrutiny. I have acquired some plain nanoparticles and in future I will write a comparative analysis to show that I can make it appear that the nanoparticles have stripes ad libitum! But I can’t wait, so here I’ll give you, hungry followers, a little taster.
First, though, I should say that, like pal Rapha, I can’t agree with FS that the only appropriate forum for scientific debate is the scientific literature. What about blogging, tweeting, facebooking, youtubing, flirting, pestering, arse-lickering and retractionwatching? A case in point is pal Rapha’s blog, a fair, friendly and fun forum where everyone is invited (even the foolish foe Pep!) and where all voices and arguments are discussed fairly and squarely (and verbosely and loquaciously si c’est moi). Except for deceitful scientologists such as stripe-believers; they should not be allowed to have a voice in scientific forums!
Now the taster: let me highlight just a few of the massive number of problems with the fake data made public by FS.
Who? FS! What? Ringing. Who? FS! What? Ringing. Who? FS! What? Ringing. Who? FS! What?
Do you get it?……. Feedback-loop artefacts!
(By the way, the joke is of my own making. I cracked it to pal Rapha, but did not work the first time; had to tell him that a feedback loop is not this).
(ii) Low pixel densities.
The images from FS are of low resolution. A neophyte to the stripy-nanoparticle creed would think that FS does not have enough money to buy high-quality STM equipment or that it may be difficult to reduce the scan area and increase the pixel density for the stripy nanoparticles. But sophisticated, well versed and consummated microscopists like myself and no others can immediately tell that this is one of FS’s tricks. In fact, FS used the uncertainty principle to his favour (maybe the guy is less than a blockhead than I once imagined). I am sure that readers of pal Rapha’s blog all know what the uncertainty principle is about. But just in case any of those ignorant stripy scientologists land on this page, I’ll explain the principle very clearly: The more precise something about something is, the less precise another something about that same something will be. Quod erat demonstrandum.
And I love hair metal. What? A balding professor of physics loves hair metal? Yes, there I’ve said it.
Hey, I am the coolest guy in Nottingham. No one else has demonstrated the uncertainty principle with a metal guitar. I love lots of distortion, chunking, chucking, crunching, damping and a REAL groove!
But I digress; I don’t want to talk again just about how well I chuck the guitar strings. I prefer to chuck FS’ grooves!
So, this is to say that damned FS has taken advantage of the uncertainty principle: the more precisely he is able to control the scanning velocity of the tip of the microscope, the fuzzier the resulting signal. But even the least experienced users of STM (such as Pedja Djuranovic; poor pal, he was with FS so he could not learn much microscopy) know that to get around the uncertainty principle you increase resolution by reducing tip speed and scan area. Easy peasy.
(iii) Feature selectivity.
Yes, FS selects what he wants to image and analyze. And then goes and only selects places where there could be stripes. What a biased guy. If he would select regions where there are no features, he would not find any! But no, the cheap cheeky cheater could not crunch the truth.
(iv) Error-bar ignorance.
I am going to turn a deaf ear to this one.
In 2005, poor Pedja showed that stripes appeared on bare surfaces when he set the proper feedback-loop settings (no merits here; I can actually do that while grooving). As I was busy exposing FS’ lies, I have told one of my idle students to reproduce Predrag’s results with a simulation. Here is the outcome:
Now compare these against the 3D-rendered experimental data of FS from Fig. 3 of Jackson et al., Nature Materials 2004:
And also with my experiments:
This is how I create fake stripes: in (A) When I boost the gain (see the arrows? There!), I win the big prize: lots of stripy thingies! In (B) I did a zoom of the brightest section after removing some damn noise, in (C) there is a 3D geeky rendering.
I could go on and on and on and on with my erudite loquacity, but in this case the images speak for themselves. The similarity between the fake images from FS and my cuties is striking. I have created stripes on bare nanoparticles! The
emperor villein has not clothes!
Essentially, it’s like simulating the microgrooves in a flat vinyl record by tinkering with the needle. I can make the needle move like it is following the grooves, and I can simulate that in the computer. Because the fake grooves resemble those from FS, these means that FS does not have vinyls with microgrooves! Do you get it? The stripes do not exist! Cheeky creepy cheater!
FS claims that he can distinguish stripes from noise. Surely, he then must then be able to hear Paganini in my hair-metal improvisations. To show how silly is that, I have picked an appropriately highly noisy, low-resolution image from FS’ archive:
FS sees stripes in here. He can even measure the stripe width! This is beyond imagination. What a tard.
Compare and contrast
I have also used the data released by FS to show that I can make his images much more beautiful (Hey, Nature Materials, will you publish me if I send you beautiful stripes? You guys have to recognize that these are much better than the fake, flaky, ugly images of FS you have put on your slimy pages). See:
While I was making awesome stripy images, a pitiful student came to me and dared to ask why the stripes extend beyond the edges of the particles. I told him off: don’t you see that these are phony?