I changed the structure of a sentence but that needed a change of a word. 😛

Note to self: nvdi

There is an acknowledged issue in the “Soft Sciences”** which is quite serious, called The Decline Effect, aka The Reproduceability Problem

Studies done and accepted even as far back as the 80s and 90s are, when an attempt is made these days to reproduce them (as, for example, a class assignment in a field of discipline), they are failing to be reproduced.

It’s happening a LOT in the Soft Sciences, particularly the biological, sociological, and psychological fields.

A couple articles about the issue from ca. 2010:

https://wolfhowling.blogspot.com/2011/01/scientific-method-its-limits-decline.html

https://www.newyorker.com/news/news-desk/more-thoughts-on-the-decline-effect

An older, published study (2005) by John Ioannidis:
Why Most Published Research Findings Are False
https://www.ncbi.nlm.nih.gov/pmc/articles/PMC1182327/

You might remember that name — he was the epidemiologist who was pilloried early on in the Covid BS for disputing the official party line.

And, more recently…

https://www.sciencealert.com/replication-results-reproducibility-crisis-science-nature-journals

Note that there is also The Reproduceability Project
Psychology: https://osf.io/ezcuj/
Cancer Biology: https://osf.io/e81xl/
General: https://osf.io/search?resourceType=Project%2CProjectComponent

This also ties in, for fun…
XKCD #1
https://xkcd.com/882/

XKCD #2
https://xkcd.com/397/

The REAL problem is “QC” — Quality Control — Unfortunately, all too often stuff depends on peer review which does NOT involve actually reproducing the study, but merely looking over it for obvious errors. If the “error” was not in the actual direct process of the analysis, but in the “cherry picking” of what data is applicable and what were “outliers” and thus to be ignored, well, the peer review won’t do squat about that. Also, if there is a weakness, the reviewer, who may be inclined to believe the study because it confirms what they have already decided they believe to be the case, may rationalize glossing over the problem rather than calling attention to it.

Doing actual reproduction of a study in any bioscience, psych, or sociology field is a pain, for obvious reasons. You need people to produce the data, and time, often, to reproduce the process. So the Soft Sciences are usually just “review type” peer-review, and not full reproduction for verification and validation.

=====
** The “Soft Sciences” : I use this term in the following manner — there are “Soft Sciences” and “Hard Sciences”. The distinction is in how stringently the universe tests a given study’s validity.

Physics, Chemistry, Engineering are generally — not always, but usually — fairly Hard. The devices work, the bridges connect, the buildings stand. Come up with a conclusion that is wrong, and the universe goes, “Uhhh. Nope. Fuck you, stupid!” — an obvious historical example of this was the Tacoma Narrows Bridge — Engineers, having built a variety of suspension bridges, thought they fully understood them. Except, no, they did not. The result was a bridge that fairly quickly destroyed itself.

The bio sciences, psychology, sociology — these are almost always Soft Sciences — because the universe does not stomp on you immediately when you screw things up. This is why so much of the issues with Reproduceability occur in these fields.

Note that there are some subfields of physics — “Climate Science” is the glaring example — which are a Soft Science, even though they are actually physics. Because it’s clear that, for “Climate Science” there is not a lot of immediate feedback when you do something wrong. A study which decides ‘x is happening’ does not have an immediate destruction of something to say, “no, it’s not happening”. This is even more the case when the models are not openly operated, tested, and publicly accessible so anyone can run studies based on them to verify the accuracy of their predictions…

And in the end, the accuracy of predictions is the truest test of any science.

You are totally correct. I’ve long been making the connection between the modern media and the French word for feces. It feels quite apt.

😛 😉 😀

}}} what won’t be found is and evidence* is more properly to be understood as “any“, though I’m open to correction if found to be wrong!

Yup. Typo. DOH! Been making a lot of “and” typos of late, no idea why (well, “old age” but still…).

I sometimes do “and” when I mean “an”, also. Weird.

Seen myself doing some typos like “theb eginning” for “the beginning”, too.

USUALLY I catch it… I’m a naturally excellent proofer, but I might not re-read something (I suspect that’s the usual cause for it slipping past me) is the only explanation I have.

The exception that proves the rule is Winsome Sears.

But the others are all like Michelle Obama — low on merit coupled with an enormous sense of entitlement.

Given DEI/CRT, black women can get away with it more than other sliced-and-diced groups. So black women do.

Nothing mysterious about it.

it seems every turn against academic rigor is actively encouraged, which devalues the end product but they don’t seem to care, take liz warren entire fraud ridden career,