Comment by WhompingWindows | Hacker News Reader

Comment by WhompingWindows | Hacker News Reader

WhompingWindows

First issue: publications. I think this is not a big problem in the USA relative to China. Is there evidence of American scientists who have greatly benefited in reputation and prestige, despite doing very shoddy work? Many scientists have been rebuffed and rejected from grants/awards due to not having enough publications. American scientists, though, are more attuned to the quality and integrity of journals in which publications appear. Whether for tenure or grants, a Nature or Science or Cell paper will mean a LOT, and many scientists evaluating other scientists wouldn't care much about a journal publication with Impact Factor under 1.0.

Meanwhile in China, there are numerous article-factory-journals that are pay to publish, you can put your shoddy work in those and amp up your publication count easily. Surely, these exist in the USA, but are career scientists at major institutions utilizing these shady Chinese journals? There is evidence that some of these Chinese journals are publishing straight up BS, which is especially easy to do with data analysis where you could "clean" your data easily. Perhaps it is a cultural or political difference, but I don't see nearly as much rigorous self-reflection of Chinese scientists on this front?

Second issue, grants: publications are a small slice of this story. Science departments (not humanities) in universities are MAJOR revenue generators for the University. My university took 1/3 of your grant straight off the bat, to cover overheads like shiny facilities and administration and marketing. Meanwhile, the scientists themselves may make huge salary bonuses or advance their tech/staff substantially when they have substantial grants. So, getting a grant is great for you personally, and improves your chances of further grants.

So, is it really that surprising that there may be pressure to publish at all costs, to p-hack, to reach for those low-impact journals despite their lower reputation/impact, given both universities and scientists are BOTH massively financially benefited from this incrementalism? Does it really pay to reach for pie-in-the-sky, fundamental sea-changes in your field? It seems like a high variance, high risk strategy that only very bold, well-funded, devil-may-care scientists would employ.

I can't comment on China and other fields, but in the US for AI and robotics, in which I do research / publish in, there is definitely a growing trend of an over-emphasis on novelty research that disregards reproducibility in favor of "wow" factors and fancy demo videos. A lot of highly cited research papers/labs tend to be the most heavily promoted ones (on Twitter, in the press, etc), and they're not necessarily impactful/useful for the rest of the field.

Amidst this worrying trend of shallow research, begging for petty cred and grants more than knowledge, you have to wonder what has become of ethics, of integrity in science. Globalized academia needs some soul-searching, IMHO.

" you have to wonder what has become of ethics, of integrity in science"

When was there any more ethics or integrity in science than any other time? The AIDS crisis was a shitshow of choosing prestige and recognition over the lives of a generation. The discovery of DNA was off a woman who was hardly recognized. Henrietta Lacks' cells. The syphilis experiments.

The discovery of DNA was done by Friedrich Meischer, a brilliant Swiss biologist. At the time, nobody knew what the nuclein he purified did (and, the results were considered so surprising that his own thesis advisor redid all the experiments manually before letting the data be published).

What you are referring to is the use of Rosalind Franklin's X-ray fibre diffraction images by Watson and Crick to elucidate the 3D structure of DNA, and, depending on the accounts you read, whether she got due credit is arguable. She did publish in the same Nature journal issue as W&C (https://www.nature.com/articles/171740a0.pdf), she got credit for the photos (see the acknowledgements in the W&C paper, http://www.nature.com/genomics/human/watson-crick/), and she was dead by the time the Nobel Prize decision was made (so she could not have received the prize).

I understand many feel very strongly that she was cheated, and while I do believe she was definitely slighted and not given enough credit, the underlying story is fairly complicated. I recommend reading both Dark Lady and Eighth Day of Creation and then forming your own opinion. personally I thought her personal diaries, which she willed to Aaron Klug and were used in the writing of Eighth Day, were really illuminating.

Sorry, you're right, the story is complicated, but that it is so complicated draws further the question- where are we getting the idea that science was full of ethical humans doing purely ethical things?

> Is there evidence of American scientists who have greatly benefited in reputation and prestige, despite doing very shoddy work?

This is the exact point of the replication crisis. Within social science, the answer is an absolute yes. That's why we have been having this conversation more and more over the past decade.

> Meanwhile in China, there are numerous article-factory-journals that are pay to publish, you can put your shoddy work in those and amp up your publication count easily. Surely, these exist in the USA, but are career scientists at major institutions utilizing these shady Chinese journals?

In China, people know about scientific reputation perfectly well, which is why publication in Nature/Science/Cell ensures great financial reward.

There are shoddy journals everywhere, of course. Yes, top scientists at top institutions and rigorous fields in the US don't use them. The same is true for top scientists at top institutions in China. The average quality is probably lower (due to the inherent difficulties of doing cutting-edge science only two generations removed from mass starvation) but the principle is the same.

Regarding your first point, the West may not be as bad a China, but there are some notably bad actors. Schon [1] fabricated results very impressive results at Bell Labs for years before getting caught. In fact, these results were so impressive that it shifted the focus of research groups in his field for years to try and get his methods to work. It was such a big scandal that my (and I assume other) universities added a research ethics requirement to their engineering PhD programs as reactionary measure.

[1] https://en.wikipedia.org/wiki/Schön_scandal

Reply on news.ycombinator.com