General > General Technical Chat
Dilbert loses newspapers, publishers, distributor, and possibly its website
<< < (144/222) > >>
tooki:

--- Quote from: Nominal Animal on March 12, 2023, 05:19:22 pm ---For an example of 60% of articles failing reproducibility test, look at psychology (Nature, 2015, cited 56 times).  (And that's not all crappy papers, only the ones whose results cannot be reproduced.)

In preclinical cancer research, 89% of articles failed reproducibility testing, ie. their results could not be reproduced.  Also note that this Nature article has been cited 1851 times, so it is definitely main-stream science, not some lone goofball spouting nonsense. 

--- End quote ---
Which is in no way indicative of the scientific process failing. It's literally just the process working exactly as it is supposed to. One of the core purposes of academic publishing is to say "hey everybody, we did this, but we need for others to replicate the experiment to either confirm or refute its findings". Academic publishing isn't intended to mean "we did this novel experiment and it proves this and that". When you have several replicated experiments that all confirm the findings, then you can start to take it to mean that something is proven.

Of course, most people don't know any of this, since they're scientifically illiterate and rely on the often wildly distorted misinterpretations of scientific papers by scientifically illiterate journalists.


The biggest actual issue with academic publishing is that failed experiments are rarely published (that is, they're rarely ever submitted for publishing). So there's a lot of studies we don't know about at all. That and the paywalling of the results of publicly funded research in privately-published journals that add very little value to the entire process.
james_s:
The more times a theory holds up to being tested, the more confident we can be that the theory is truth, but it's never really 100% proven. Science encourages continuous testing and questioning of everything, and occasionally things long believed to be true turn out to be incorrect.
Wallace Gasiewicz:
I'm OK with the theory that everything consists of Earth, Wind, Fire and Water. (although you can get a really good rock group with just three of these elements)
It's the Aether that I am not so sure of.
I dislike these newfangled ideas.
Buriedcode:

--- Quote from: james_s on March 13, 2023, 04:53:06 pm ---The more times a theory holds up to being tested, the more confident we can be that the theory is truth, but it's never really 100% proven. Science encourages continuous testing and questioning of everything, and occasionally things long believed to be true turn out to be incorrect.

--- End quote ---

Just to add, often people will constantly raise the bar for "proof" if they don't believe in a finding, or lower the bar to ridiculousness if they want something to be true.  Because you cannot really prove a negative, cognitive bias skews peoples views on what constitutes evidence so if a study is negative - it just wasn't precise enough, or wasn't large enough.  If its positive - no matter how small or poorly designed - it must prove my theory right!   Case in point the whole silly hydroxychloroquine/ivermectin debacle - where people are still quoting obviously fake studies.

Add that relying solely on placebo controlled double blind studies - without looking at important things like, you know, plausbility, or whether theres even a possible mechanism at work - means that the biases/flaws that can never be removed from said studies end up creating false positives.  For example P<0.05 thresholds - 5% of studies will be "positive" by pure chance, but are seen as showing something meaningful if one only considers the results from the test and nothing else like context.
Zero999:

--- Quote from: Someone on March 12, 2023, 11:31:22 pm ---
--- Quote from: Zero999 on March 12, 2023, 10:14:17 pm ---My point was making a non-sterilising vaccine a condition of work, or to go to public events was the issue. The vaccine is very good at preventing severe disease and death. People get infected, but it's much milder, than it would've been had they not had the disease or vaccine, so it's definitely useful. It's just questionable whether it's beneficial for all.
--- End quote ---
If it wasn't mandatory, the uptake would have been too low to keep the hospitals from collapse (which did happen in the UK and some other countries). Pretty much all the public health measures were just to balance hospital capacity and continues today. Isolation and/or Vaccination, one of those is much easier to sell politically.
--- End quote ---
The fact it wasn't mandatory had nothing to do with the UK's hospitals being over-crowded. Most of the pressure on the NHS occurred at a time, before the vaccine was approved (Spring 2020) and during the point people were queuing up to take it (winter 2020/2021) and it had to be given to those who were deemed to be more vulnerable. Uptake amongst the vulnerable was more than high enough to ensure the NHS could cope. This was one of the reasons why the Delta and Omicron waves didn't cause as much of a problem as the previous ones.

Vaccine mandates would have made the situation far worse, not better. At the end of 2021, just as Omicron hit, if I remember rightly, the government mandated all care home workers take the vaccine, but it resulted in staff shortages. They were going to do the same for NHS staff in early 2022, but dropped the requirement, as the numbers didn't add up. A fraction of the population equal to 1-1/R need to have sterilising immunity in order to stop the spread, yet Omicron has an R0 of 18, the vaccine only reduces transmission by 30% and is short-lived. Avoiding a small number of infections, which are unlikely to be severe in the vaccinated, didn't outweigh the poorer care which would have resulted from understaffing, which is already bad enough as it is.


--- Quote ---You're communicating your position much better than Dave (who dropped some barely coherent "point" and walked away), and there is considerable evidence that natural immunity is as good (or probably stronger) than the vaccines available:
https://doi.org/10.1016/S0140-6736(22)02465-5
--- End quote ---
Exactly. More importantly, protection against mild or asymptomatic disease is unimportant. Getting a cold is part of every day life. The vaccines were approved because they have been proven to be highly effective at preventing hospitalisation and death. The risk vs benefit for the first two doses was definitely worth it for most adults, who lacked immunity, especially given the more pathogenic variants in circulation. The boosters are much more questionable, but even though I find it interesting, I won't get into it again.

I've noticed many people appear to be unable to look at this rationally. There are those who will never trust the vaccine, whatever the evidence says and there are those who appear to be completely blinkered to the fact it does have some risks. Some authorities got it right, others messed up. I do my very best to look at it from both sides, but it's difficult given the lack of data.


--- Quote ---So here in Australia the current recommendation is 6 monthly jab or infection, treated interchangeably that we'll probably agree is a reasonable balance on the known data. Now that there is enough protection within the community almost all mandates for workplaces are lifted/gone (healthcare remains, due to high interaction and with vulnerable groups).

--- End quote ---
Good. They shouldn't have been any in the first place. Plenty of other countries managed fine without them. There are no mandates for flu vaccination and now this not much different, certainly not at this stage.

I didn't want to get into a debate on vaccines. My point was the pandemic response of most governments was not evidence based. I could have talked about keeping 2m distance, facemasks, school closures etc. Some of it can be put down to inexperience but a lot of it was negligent. Governments will try to make excuses for the madness, some of them have some merit, others are completely false. Getting unbiased data on which measures proved beneficial long term and which did more harm than good is difficult. I think most of them resulted in net harm. It's often better to do nothing, than harm, but that's pure speculation.
Navigation
Message Index
Next page
Previous page
There was an error while thanking
Thanking...

Go to full version
Powered by SMFPacks Advanced Attachments Uploader Mod