This will not be a popular comment, and I love science and scientists. I have spent my career working with scientists and investing in science. But the research community as a whole has a credibility problem. Does this problem merit financial cutbacks? I can't say, but it sure makes the rhetoric and justifications easier for those doing the cutting.
These stats have been reported by various legitimate media sources--and you can verify these for yourself on the web:
- Outright fraud: about 2% of scientists admit fabrication or falsification, while about 14% report knowledge of similar misconduct by colleagues.
- Image manipulation: roughly 4% of biomedical papers contain problematic image duplications, with some cases suggestive of deliberate manipulation.
- Questionable research practices: up to 33.7% of researchers admit at least one such practice, while up to 72% report seeing them among colleagues.
- Replication failure is serious: results can't be duplicated in more than 50% of major psychology and biomedical research projects.
- p-hacking: researchers can unintentionally or deliberately search across data until a result becomes statistically significant, increasing false-positive risk without necessarily committing outright fraud.
- Biomedical research waste: one landmark estimate argues that up to 85% of biomedical research investment may be avoidably wasted through poor design, non-publication, and incomplete reporting.
- Unpublished research: science is science, but many null or disappointing studies remain unpublished, distorting the visible literature and causing other labs to repeat failed work.
- ...And finally: a minority of funded research (much less than 50%) is reliable, novel, usable, and clearly reported scientific journals.