Hahahaha! I’m a scientist in the drug discovery industry. Want to talk about regulation, checks & balances, testing & market competition? Want to talk about competition for grants, research opportunities and positions within the academic science community?
As long as you are working in an industry…especially something to do with medicine… I am sure you have more than enough people on your back… that’s one field that makes engineers look like they are running around unsupervised…
I am sure there is a lot of competition in academic sciences and the work is tough. However there is a difference between competition and regulation.
Guess what again?
Umm… I’ve worked with computer engineers on software projects and I’m not too sure about the level of supervision they were held to. Perhaps they are more in the ‘wild west’ stage?
Usually it’s for products involving human safety… for example automobiles…
Computer engineers have comparatively more leeway. But end of the day, their customers sometimes are far more demanding … and so they do have fairly robust quality control in the bigger firms.
Science certainly is for the most part self regulated especially for the actual content of the science, and for good reason. It doesn’t make sense for government officials or religious bodies or laypersons who do not have the relevant expertise to regulate science. (Edit: to be clear, here I am referring to the CONTENT of science, not ethical standards or funding.)
That being said, science is indirectly regulated as government officials and donors often have a large stay in determining which kinds of projects get funded and thus what kinds of science end up being done.
Don’t forget all the ethical disasters that arose (and will arise again) from that attitude. We absolutely are regulated by government officials and members of the public, including religious people. Do you know what an IRB board is?
Sorry if I was being unclear. In that sentence I am referring to the content of the science - the propositions that have truth values deduced from empirical observation. As I explained in the next paragraph, other aspects of science such as funding and certainly ethics can be regulated by outside entities.
This is the case with engineering or any other technical subject. When it comes to content and actual quality checks, the testing is done by experts.
However, these experts answer to a third party (often government bodies, sometimes an independent body certified by the government/industry).
This is definitely possible for science. There can be mandatory requirements before publishing results like checking repeatability/reproducibility through an independent internal or external team. It’s also interesting that quality checks require an expertise in testing in addition to domain knowledge.
Peer review in many ways is a limited excercise.
Already done where it counts, in drug development with the FDA.
I am sure it’s covered in pretty much any product sold to the public… especially where safety of the user is involved such as drug development. This is something all governments will do to protect the general public.
Extending to other areas is totally upto the initiative of scientists/universities etc. Doing so will be great for improving the quality of peer reviewed papers.
As @swamidass mentioned, there is a ton of regulation in research, even for studies that aren’t developing drugs for human use. At least in the US, all research has to go through an Institutional Review Board which includes scientists, administrators, and people from the general public. Below the IRB are several subcommittees, including Biosafety, Chemical safety, use of animals and ethical treatment of animals, and human subjects, just to name a few. The paperwork needed to get a funded research program up and running is nearly as large as the paperwork for getting the grant funded to begin with.
When it comes to the ideas that scientists will pursue, that is somewhat regulated. Most granting institutions are looking to fund grants that cover certain topics, so it may be difficult to find funding for “out of the box” ideas that don’t easily fit into those areas of interest. Also, there are guidelines for the treatment of animals and human subjects, for obvious reasons. Currently, there are also guidelines for stem cell research and genetic modification, such as the ban on using human embryos and using CRISPR/Cas9 to alter the genome of living human beings.
However, once a study is funded and approved the scientist has a lot more leeway and can follow the evidence to its conclusion.
And sometimes, perhaps even often, these conclusions cannot be reproduced by others when they try to.
There is a difference between the regulations you cited and what I am talking about.
If you look at peer reviewed publications as the products of science. What is the quality control in ensuring they meet a particular standard? Currently that quality control is peer review. It seems to have a fairly high failure rate in some cases.
There are occassions where the data/results can be difficult to replicate.
Peer review is imperfect, but it is the best system we have found. The quality of the review depends on the quality of the peers and the quality of the journal. Do really bad papers make it through peer review? Yes, they do. However, the only solution I have seen for this problem is better peer review.
The thing to keep in mind is that the reviewers don’t repeat the experiments to see if the results are legitimate. For the most part, results are taken as true unless they wildly deviate from what the peer reviewers would expect to see. Most of the scrutiny is placed on the methodology and conclusions, at least in my experience.
Peer review is part of publishing in engineering research journals…
A seperate team confirming the work by reproducing it is definitely possible. It could be made a minimum standard requirement before peer review. It’s not impossible. Time consuming and costly… yes…
On the long run, it will help in showing which conclusions are reliable.
At least in the field I am familiar with, repeating experiments just isn’t practical. If we are talking about animal experiments, it could also be unethical. For example, if a study infected mice with a virus and measured survival it would be extremely unethical to submit upwards of hundreds of mice to the same suffering just to see if the data repeats.
Most peer reviewers are researchers themselves, and all of their time, money, and effort is focused on testing their own hypotheses. It just doesn’t make sense to commit their own lab to months of work just to confirm the findings in a paper.
Ultimately, the results in papers are tested indirectly or tangentially. If the findings are of interest then people will base new research based on their methods. Those experiments will include controls that try to repeat the findings in other papers, as well as experimental groups that add to our knowledge. If a paper contains fudged data then it will become apparent when others try to build on that research.