The Last Word

[Bazza]

No worries.

[Pariah]

Listening to Vangelis' Heaven and Hell while grading final exams today.

 

Seemed oddly appropriate. 

[Cancer]

Alice Cooper's School's Out is also a favorite.

[Pariah]

Right now, my favorite song relating to school would be "The Sounds of Silence".

 

[Bazza]

Good call

[Bazza]

"Mysteriously, almost everyone who describes themselves as a "free thinker" have almost exactly the same ideas."

~ Existential Comics

[L. Marcus]

"I'm right and everyone else is an eejit."

[Cancer]

"Mysteriously, almost everyone who describes themselves as a "free thinker" have almost exactly the same ideas."

~ Existential Comics

And the same problem: how to get paid for it.

[Bazza]

Become a peer reviewer. Everyone else has the ideas, you just do peer review to confirm/validate them. 

[Cancer]

... That's an overly simplistic view of peer review.

[Bazza]

I'm an overly simplistic person.

 

(Also I don't know what is involved. Care too elucidate?)

[Cancer]

Well, I can only speak with certainty about my own field and closely related ones. No clue what it might be in rather different disciplines.

 

First, make sure it hasn't appeared before, either from this author or others. Subtly different but allied: what related results have appeared before on this topic, and can/does this work improve on those (confirm, contradict, eliminate alternate possibilities, ...)? Why hasn't this result appeared before, i.e., what's new here? Then make sure that this work is properly placed in context of what has come before. What's the current state of knowledge and informed speculation now? Why is this work relevant?

 

Next, check the methods. Do they make sense? Can they reasonably be expected to produce the claimed result? Are they in accord with what other workers in the field have done, or clear extensions of those? Are other requisite basic data current? Is there some hidden circular logic that more or less ensures the result independent of actual measurements?

 

Examine the uncertainty in the results, both statistical (the easy part) and systematic (much harder, but at least as important)? Are the overall levels of certainty plausible for the data and methods? Is there a solid discussion of systematic effects, and reason to believe that the result is safely not spurious?

 

Are the results relevant to the questions presented in the introduction? Or is some other effect more important than the stated discussion?

 

Then there's questions about the actual results and conclusions. Do those make sense? Does the result actually bear in favor of the conclusion drawn, and how strong is the leverage on the question? Is it actually worth publishing?

 

Much more, but dinner calls...

[Bazza]

Cheers. Sounds similar to quality assurance, to get it to a high standard, and eliminate anything that might hold it back: typos, elementary mathematical errors, errors in logic (ie circular reasoning, post hoc ergo propter hoc), etc.

[Bazza]

A while ago I made a connection between apophatic theology and scientific falsifaction, but the more I look into it, the more the two are dissimilar. I have a good introductory book on apophatic theology, but am mostly out of my depth with scientific falsification.

 

Can you describe the later if it is not to difficult?

[Pariah]

Read up on Stanley Pons and Martin Fleischmann for an example. (We had this drilled into us at University, since it happened at my alma mater and in my department.)

[Bazza]

It that an incident, an event, a co-authored book, or would Wikipedia do?

 

Edit: so Cold Fusion.

[Cancer]

Cheers. Sounds similar to quality assurance, to get it to a high standard, and eliminate anything that might hold it back: typos, elementary mathematical errors, errors in logic (ie circular reasoning, post hoc ergo propter hoc), etc.

Referees aren't necessarily typo catchers (they see papers in typescript form, not final galley form) but they should go over equations etc. with extreme diligence because few people can catch errors there.

 

But referees also serve without public recognition, and without compensation. It is a service to the community.

[Cancer]

I actually view the cold fusion episode as a prominent if painful example of how the peer review process, and science in general, works. Yes, lots of shouting, brouhaha, egos, and so on, no doubt painful for those near the event. (The episode happened early in my second postdoc, and the Physics Department (I was in Astronomy, separate department) followed it closely.) In the end, though, no one believes in cold fusion. It was interesting, and it made for a drastic spike in spot palladium prices, but by a decade later ... shrug. Another sexy idea that didn't work.

[Bazza]

Cheers for this Cancer. You're da man.

[Bazza]

Speaking of cold fusion and what science is:

 

PHYSICISTS CAN’T AGREE ON WHAT SCIENCE EVEN MEANS ANYMORE

https://www.wired.com/2017/05/physicists-cant-agree-science-even-means-anymore/

[Cancer]

It that an incident, an event, a co-authored book, or would Wikipedia do?

 

Edit: so Cold Fusion.

Another famous case from the turn of the 20th Century was N Rays, for which the Wikipedia article will do. In effect, Blondlot deluded himself into believing an effect that did not actually exist. While no one could reproduce Blondlot's results (and so there was wide skepticism), Wood's demonstration in Blondlot's lab -- providing an independent test that was badly needed -- was all but brutal, and put paid to the whole thing almost immediately. Blondlot was not a fabricator, but he didn't have adequate tests of his findings.

 

I also recommend Feynman's Cargo Cult Science commencement address. He casts his net wider than science, but he was speaking in a time when lots of ... questionable popular phenomena ... were reaching for the aura of scientific plausibility, and was striving to smack down the more obvious drivel as being, well, drivel.

[Bazza]

Cheers.

[Bazza]

 

Didn't know this was an option.

[Bazza]

[Cancer]

"I think of myself more as Dr. Evil without the megalomania."