Most scientific research is false 🤥

In 2005, Stanford professor John Ioannidis proved that most published scientific findings are false. He reached this conclusion based on mathematical modeling.

The problem lies in non-reproducibility. Other scientists simply cannot repeat the experiment described in the article and obtain the same data.

A striking example is the search for schizophrenia genes: about 100,000 genes are tested, of which only 10 are actually associated with the disease. Even without manipulation, statistics will yield thousands of false positive associations. Real discoveries simply drown in a sea of garbage.

John Ioannidis identified 6 factors by which you can identify fake articles:
1. Few studies on the topic.
2. Weak effect (insignificant association).
3. Large selection of hypotheses and flexibility in testing them.
4. Freedom in interpreting data and methods.
5. Financial interest and bias.
6. High competition among research groups.

The system rewards sensational discoveries, not reliable ones. Scientists are forced to "find" significant results under the pressure of "publish or perish." Simply put, you should only trust data that has been replicated by several independent teams. Everything else is just statistical noise.

And yes, in 2015, the editor-in-chief of The Lancet, Richard Norton, stated that perhaps half of the scientific literature is false. And this is in medicine, where people's lives are at stake.

TG | MAX | VK | DZEN | #research #science