Is "Ideal" science very different from how scientist do their work?

Fantasus

Ad Honorem
Jan 2012
2,381
Northern part of European lowland
By reading some popular science sites and some popular books I have got that idea. That if we think scientist only make rock-hard conclusions based upon solid observations and experiments that is far from the whole truth. That what they do very often is as much about modelling, hypothesis and "what ifs". As almost all written about life elsewhere in the universe, about possible "aliens", multiverses etcetera, but even ideas of "planet 9." That most of what we learn scientists don´t allow themselves to do, they do it anyway.
 

Chlodio

Forum Staff
Aug 2016
5,097
Dispargum
Speculation leads to hypotheses. Hypotheses drive experimentation and observation.
 

Todd Feinman

Ad Honorem
Oct 2013
7,064
Planet Nine, Oregon
They do say silly things, like "we will find aliens in 20 years!" .
For people who have seen UFOs close enough to know, it's really hillarious! It's as if they are saying "we might pull our heads out of our posteriors in 20 years or so!" :cool:
 

Fantasus

Ad Honorem
Jan 2012
2,381
Northern part of European lowland
They do say silly things, like "we will find aliens in 20 years!" .
For people who have seen UFOs close enough to know, it's really hillarious! It's as if they are saying "we might pull our heads out of our posteriors in 20 years or so!" :cool:
UFOs? Unidentified Flying Objects? There could be a lot of unidentified objects, flying or not. If they are unidentifed, why then conclude they are not from earth?
 

Todd Feinman

Ad Honorem
Oct 2013
7,064
Planet Nine, Oregon
Ball lightening, sprites, lenticular clouds, swamp gas, airplanes with the sun glinting off their wings, secret test aircraft, optical illusions, hallucinations, corpuscles in the eyes, lies and hoaxes? :)
 

Robert165

Ad Honorem
Jan 2010
4,266
North Georgia
Is Most Published Research Wrong?

Up to 1/3 of published papers can be wrong, at least, for new studies.
A few of the problems:

Bias (unconscious bias of experimenter)
Bias (purposefully dishonest experiments)
Small sample sizes
Failure to distinguish between false negative and an true negative results*
P value not strong enough (how strict the margin of error is)
Failure to publish studies that had negative results (from the initial study)
Failure to negative results from replicated experiments
Trying to test for more than 1 variable at a time
Restricting the scope of the study to very small populations

* kind of hard to describe, see video a point 2:15 to describe false negative vs true negative


I've seen/read elsewhere that another problem (not discussed in the video) is that many research papers in published journals do not get enough scrutiny. In other words, lots of times their peers just give a "light read" to the study because they are so busy with their own work. If the published results seem plausible then they are much more likely to have undetected errors than if there are very obvious errors (that were overlooked in the initial study).
 
Last edited:

M.S. Islam

Ad Honorem
Jul 2012
3,333
Dhaka
Bias (purposefully dishonest experiments)
This is the single most influential factor. Principal causes of bias -

- Catering to whims/interests of those who fund researches.
- Producing politically correct research outcomes.
- Fame-seeking by way of producing controversial research outcomes with the objective of grabbing media attention.
 

aggienation

Ad Honorem
Jul 2016
9,813
USA
Govt grants for university research labs are one of the worst things for science in the US. Even though they are funding entire fields that normally probably wouldn't even exist for lack of funding, probably a good thing, over time the university research system has become utterly dependent on them for their very existence and is now completely geared to applying for and sustaining itself on grants. So they cater all their research to support whatever is the fad science, lest they promote something that isn't popular and somebody else gets "their" grant, and then suddenly the professor loses a significant portion of their own salary, having to cut various grad students loose to go find a home in another lab. Lots of politics, resulting in heavy manipulation of research data, resulting in a very toxic environment.
 

pugsville

Ad Honoris
Oct 2010
10,100
Govt grants for university research labs are one of the worst things for science in the US. Even though they are funding entire fields that normally probably wouldn't even exist for lack of funding, probably a good thing, over time the university research system has become utterly dependent on them for their very existence and is now completely geared to applying for and sustaining itself on grants. So they cater all their research to support whatever is the fad science, lest they promote something that isn't popular and somebody else gets "their" grant, and then suddenly the professor loses a significant portion of their own salary, having to cut various grad students loose to go find a home in another lab. Lots of politics, resulting in heavy manipulation of research data, resulting in a very toxic environment.
Whats the alternative? Private corporation is subject to much the same probelms as well as generally being more relentlessly short tern,
 

aggienation

Ad Honorem
Jul 2016
9,813
USA
Whats the alternative? Private corporation is subject to much the same probelms as well as generally being more relentlessly short tern,
I'd cut the dead weight. Much of academia science produces almost nothing of actual value besides giving six figure jobs to professors and creating grad students trying to become professors. More oversight on what sort of grants are approved. Be utterly ruthless. Save tax payer money and fix that has become a massively toxic industry.