Science, Fact Or Fault.

seventiesdemon

Superior Member
Gold
Platinum Gold
Joined
May 25, 2019
Posts
4,927
Media
7
Likes
5,609
Points
383
Location
Australia
Sexuality
69% Straight, 31% Gay
Have a look through these few articles and if you wish post an opinion.

2 are articles, 1 is a forum I happened across where scientists, lecturers etc from varying fields are discussing the problem of why up to 70% of research practice and results are faulty. The thread is 7 or 8 years old but it seems to be an issue which is concerning for them.

The articles I think come from unbiased sources.

I think it's important because a lot of people put belief 100% in science these days with no questions attached.....statistically speaking ...we shouldn't do that :)

Has anyone taken seriously that due to statistical errors that...

https://psmag.com/education/scientists-are-wrong-a-lot

One reason so many scientific studies may be wrong
 

halcyondays

Worshipped Member
Joined
Feb 22, 2014
Posts
6,360
Media
2
Likes
10,359
Points
158
Location
US
Sexuality
80% Straight, 20% Gay
Gender
Male
Anyone and everyone can Google a list of the great blunders in science. The list of small blunders is more important. One can also make a study of experimental and statistical error. Armies of PhDs spend their careers working on nothing else. It doesn't mean all science is wrong.

Any student of science history knows that what science got/gets right is built on an enormous pile of what science got/gets wrong or partly wrong. What doesn't work points the way to what does.

The scientific method is a tedious process. The blood, sweat and tears poured into it is astonishing. To find out what works we need more science not less. You should expect that most of it will be wrong or partly wrong. It comes with the territory.
 

seventiesdemon

Superior Member
Gold
Platinum Gold
Joined
May 25, 2019
Posts
4,927
Media
7
Likes
5,609
Points
383
Location
Australia
Sexuality
69% Straight, 31% Gay
Anyone and everyone can Google a list of the great blunders in science. The list of small blunders is more important. One can also make a study of experimental and statistical error. Armies of PhDs spend their careers working on nothing else. It doesn't mean all science is wrong.

Any student of science history knows that what science got/gets right is built on an enormous pile of what science got/gets wrong or partly wrong. What doesn't work points the way to what does.

The scientific method is a tedious process. The blood, sweat and tears poured into it is astonishing. To find out what works we need more science not less. You should expect that most of it will be wrong or partly wrong. It comes with the territory.

Blunders is not what it's about. Everyone makes blunders, in every job. Science is no different, except some blunders or accidents in science lead to good things. Before engineers and modern steel reinforcement became involved in building, everything was trial and error. Even so, modern buildings and bridges still collapse today, not only, because of poor quality workmanship, but because of poor quality engineering, or substandard materials supplied. Cutting of corners because the owner only wishes to spend so much, or the competition to win the contract....etc


It isn't about fail a thousand times and then Eureka !

If you had read the links, it's about research being fudged, undesired results left out etc to give, attain the result wanted or preferred. In the discussion in the link provided to the forum, the concern was research is being fudged to gain the required result for reasons of prestige, bias, funding etc...research being accepted and not checked by piers.

In other links not posted it has been or was suggested that "research registration" become the norm to prevent this from happening so that things can't be changed or manipulated in a study so results can be compared to the data supplied. This became the norm apparently in medical research in 2016 or a bit before. I don't think it extends to other areas of science research or study though.

This is a post from a one of the researchers in the discussion.... it's apparent this discussion was before the introduction of the "registry". But it sheds light on an obvious issue with research in science as a whole.




1st Jan, 2012
Harald H.H.W. Schmidt
Maastricht University
Her is an important paper for the field of ischemic stroke, where over a thousand suggested targets and therapies have all (!) failed when further tested. They conclude a massive publication bias, i.e. negative data are not published leaving underpowered papers in the literature. This is on top of poor quality and lack of power calculations beforehand. What drives people to this? One cause are certainly the funding institutions and bureaucrats in universities who count peas, papers, and impact factors.

PLoS Biol 8(3): e1000344. doi:10.1371/journal.pbio.1000344

Article Publication Bias in Reports of Animal Stroke Studies Leads t...
 
Last edited:

halcyondays

Worshipped Member
Joined
Feb 22, 2014
Posts
6,360
Media
2
Likes
10,359
Points
158
Location
US
Sexuality
80% Straight, 20% Gay
Gender
Male
If you had read the links, it's about research being fudged, undesired results left out etc to give, attain the result wanted or preferred. In the discussion in the link provided to the forum, the concern was research is being fudged to gain the required result for reasons of prestige, bias, funding etc...research being accepted and not checked by piers.

I did read the links. Fudging of data and results ARE blunders.
 

seventiesdemon

Superior Member
Gold
Platinum Gold
Joined
May 25, 2019
Posts
4,927
Media
7
Likes
5,609
Points
383
Location
Australia
Sexuality
69% Straight, 31% Gay
I did read the links. Fudging of data and results ARE blunders.

Arriving at a conclusion by purposely not including negative observations, not listing all the variables, not listing all the conditions especially in the world of modern science where "complex computer modelling" has become a way of predicting future outcomes. These models depend heavily on the quality of information input supplied, in fact they are useless without it.

I'm not aware the meaning of the word blunder includes the description, deliberate omission of manipulation to attain a positive outcome.

That's like the defense omitting all the truth in a legal case, leaving it up to the prosecution to discover it, the defense is innocent until it is. This is where many scientists concerned about this are critical of pier review these days.

Because it's convenient to leave out the variables, leave out negative results, leave out failed manipulations for the purpose of gaining a required outcome, albeit a false one is hardly a criminal offence. But it's not true science, and it is not good for science.

Increasing numbers of 'false positives' and published results not being able to be replicated based on the data and research supplied.. led to to the recent Registry of Research being set up so that manipulation could not take place outside of original experiment guidelines. If changes were made to the guidelines, negative results discovered or omitted etc are required to be added......which is what I would have thought was always was required, especially if results are published.

As I've said, I am not a scientist, but if I'm expected to believe what science wants me to believe outside of my own understandings and interpretations then I expect what I am being told to be based on "all" of the research and information. Not just information conducive to a popular outcome.

I posted this thread because it was obvious there were serious concerns in the scientific community about the quality of certain percentage of research. How much of this research? No one really knows, but obviously enough fact to introduce a registry to police it....

I blundered into this while looking into science, I was curious so I looked further.
 
Last edited: