What Role Does Censorship Play in Modern Scientific Research?

[tom aaron]

This is sort of off topic, but..

To what extent does the requirement for sharing raw data enter into the peer review process? I don't recall ever seeing a paper with raw data included. I would expect that reviewers probably don't spend their time sifting through data unless there is something questionable in the analysis. Is it typical for data to be submitted with the manuscript? Also, to what extent is data available? For example, if you were to perform a meta analysis I imagine that having the raw data from the studies would be important. Would you need to contact the authors of each study?

That's a good question

In my field we do not recognize a paper unless there is current or future access. This isn't in writing but the expectation in my discipline. In the past we have rejected findings from China because of issues with having original curated specimens sent to us in a reasonable time frame. This has changed for the positive in the last 20 years. I used to feel bad for Chinese researchers who did excellent work but were frustrated by bureaucratic incompetence.

 

[Andy Resnick]

This is sort of off topic, but..

To what extent does the requirement for sharing raw data enter into the peer review process? I don't recall ever seeing a paper with raw data included. I would expect that reviewers probably don't spend their time sifting through data unless there is something questionable in the analysis. Is it typical for data to be submitted with the manuscript? Also, to what extent is data available? For example, if you were to perform a meta analysis I imagine that having the raw data from the studies would be important. Would you need to contact the authors of each study?

Yeah, this thread is veering off-topic... To the extent that journals provide clear requirements for acceptance, if authors choose not to meet those requirements, one cannot claim censorship. There are many journals to publish in, and each journal has it's own set of requirements. Some journals allow for "supplemental results", often online-only, that can have a lot of the low-level detail (detailed procedures, raw data sets, etc.) that would obscure presentation of the main findings. To be sure, if a colleague requests experimental sample material, I am generally happy to provide it as freely as possible.

In terms of making raw data publicly available, there seems to be two broad approaches. One view is "This research was paid for by tax dollars and is therefore property of the citizenry- by having to purchase a journal article, I am paying twice to obtain results. And besides, wouldn't it be neato-fun to have all those interested citizen scientists look through the raw data and maybe find something that the authors missed?" (for example, the Sloan digital sky survey data releases)

The other side takes a position of "By letting the raw data roam freely, there are no controls on someone with ulterior motives from re-slicing the data to generate alternative results." Think of what would happen if lawyers got access to raw patient data and could identify those who were harmed either by being given a placebo or being that one-in-a-million with a bad reaction. Climate science is another prime example of what can go wrong, and even wrongerer...

The current middle ground is that raw data (lab notebooks and the like) does not have to be made public; the final product (peer-reviewed journal article) must be. To safeguard the journals themselves (publishing papers costs money), many journals are allowed to 'embargo' for a certain time before the articles become free. Similarly, there are data repositories (GenBank, etc.) which experimenters are obligated to use as needed.

None of this constitutes 'censorship', tho.