Hacker News new | ask | show | jobs
by lotsofpulp 1929 days ago
Yes, but I guess I wouldn’t label them as scientific processes, which I would assume be in line with the scientific method.

These listed problems are all people and people’s incentives problems, and have nothing to do with science. They exist just as much in non scientific endeavors, and so attaching the word science to them seems meaningless, and at worst muddies people’s understanding of science.

2 comments

Scientific processes =/= scientific method

The problem is with the business of science.

Processes are very much a business concept that most adults can understand. We develop new processes every day

Furthermore it was used plural so it can't possibly be the scientific method because not only is that a proper noun but there is only one scientific method while there are many scientific processes.

And yes they can be specific to science. The process of getting published in a peer reviewed journal is unique to academia.

Or what about the process of getting grant funding for research?

I know, you will say but that isn't science.

But nobody is arguing against the validity of the scientific method. At least nobody worth listening to. When people say there are flaws with science that must be addressed they are saying they are flaws with the business of science. And the scientific community is not doing a great job of fixing them.

> Yes, but I guess I wouldn’t label them as scientific processes, which I would assume be in line with the scientific method.

It's that exact assumption which is common, incorrect, and which I'd like to see addressed in K-12 education.

The scientific process in my discipline works as:

1) Graduate students come in to do research

2) Ones which align with hot topics have venues to be published, people to cite them, sources of funding, etc.

3) Some people do diligent work, and write a publication a year. Those don't go anywhere. Some people write 12 publications per year, often based on poor scientific methodology and buggy analysis algorithms.

4) The highest-impact results often come from methodological errors, and never replicate. Most errors are never found, but even if they are, by that point, the people who wrote those papers have found tenure.

5) No one has time for good peer reviews, so reviewers glance at the abstract, sometimes to see if they were cited, and skim the paper.

6) Press picks up the sexiest results, which almost always are nonsense. People also cite the sexiest results.

Increasingly, this has moved from sloppiness to intentional gaming of the system. About half of academics I saw hired in the past decade had some level of intentional sketchiness going on.

This doesn't align well with scientific methods.

In the same way, I think high school civics courses should talk about issues like buying influence, polarization, and corruption as part of civics.