Open Science Framework: A scholarly commons to connect the entire research cycle | Hacker News Reader

Open Science Framework: A scholarly commons to connect the entire research cycle

osf.io

129 pointsasaibx19 comments

What's interesting about OSF is helping to open up the research process. This is done by connecting tools and providing a centralized platform for communication. They've also been a developer and leader in helping with scienctific reproducibility.

One of their partners is Elsevier. That sets off a lot of alarm bells.

I no longer work for COS, but I was their first dev hire. I assure you that COS is not unduly influenced by Elsevier. Their funding and direction come from elsewhere.

i.e., https://cos.io/about/our-sponsors/

The page has no description of the features. From the guides [1] it seems that the main features are file sharing and versioning.

The recommended way to run an OSF instance is with docker-compose.

https://github.com/CenterForOpenScience/osf.io/blob/develop/...

[1] https://osf.io/support/

Can someone explain to me why the norm for scientific research is to lock papers behind paywalls? That kind of thing has never made sense to me. Isn't a lot of this research publicly funded by the NSF or state universities in the first place?

It's certainly moving away from this.

Anyway, here's the short of it (this is not a justification of the approach, merely an explanation):

1. Researchers are supposed to be judged by their ability to deliver high quality research.

2. A current way of measuring this is a modified form of how many papers they publish. The modification is to look at how many times a paper is cited, but this takes a long time to see. To speed this up, you can use an average measure from other papers in the same journal. This is called the "impact factor".

3. Therefore, people want to publish their research in journals with high impact factors instead of hosting them somewhere for free. Journals act, essentially, as a badge of quality.

Journals do actually spend money on the publishing process [0] (disagreements around the actual cost side, I think we can at least agree that there is a non-zero cost). They need to therefore recoup this from someone.

Making things simply openly available means you can't charge the end user, so the alternative is to charge the researchers.

Unless your grant has money for paying these fees, or you can apply for money from someone, the choice is either pay for it personally or let the publishers recoup the costs.

edit - I forgot to add a key reason this is changing. Grant funders are now largely requiring people to make their research open, and giving them the money in the grant in order to actually do this.

[0] http://www.nature.com/news/open-access-the-true-cost-of-scie...

Essentially you're locked into the system - The people who give you money (grant foundations, universities) judge you almost entirely based on your publications, and in how 'high' of a journal your publications are. As an early career researcher, you only have enough material to publish maybe one to three papers per year, so you have to make the most of it. Let's say I have a paper and I can get it into Nature Genetics (high impact factor, not Open Access) and into PLOS ONE (low impact factor, OA) it would be almost career suicide to put it into PLOS ONE.

The funding committee will look at my CV, see I published 'only' in PLOS ONE, will think 'this guy publishes only in low IF journals, so his research must be mediocre, so he isn't suitable to receive our money' and pass the money on to someone else.

Some high impact journals offer OA publishing if you pay extra - $2000 to $5000. For the CV, funding committee it doesn't matter whether it's OA or not, that's not a bonus (usually!). But again, as an early career or PhD student you don't have that much money, $3000 is better spent on going to a conference to find your future boss.

> Essentially you're locked into the system - The people who give you money (grant foundations, universities) judge you almost entirely based on your publications, and in how 'high' of a journal your publications are. As an early career researcher, you only have enough material to publish maybe one to three papers per year, so you have to make the most of it. Let's say I have a paper and I can get it into Nature Genetics (high impact factor, not Open Access) and into PLOS ONE (low impact factor, OA) it would be almost career suicide to put it into PLOS ONE.

This is also a somewhat circular problem. You need to be publishing where other big papers go, increasing the strength of those journals, so others then also need to publish where you went and so on. The lag in this system can be huge, as we're talking about slow updating numbers (citations happen over many years) and publishing timescales themselves can be months to several years for a single paper.

> Some high impact journals offer OA publishing if you pay extra - $2000 to $5000. For the CV, funding committee it doesn't matter whether it's OA or not, that's not a bonus (usually!). But again, as an early career or PhD student you don't have that much money, $3000 is better spent on going to a conference to find your future boss.

One fix to this is the funding agencies requiring open access which is happening more and more.

A mitigation is policies with journals like Nature Genetics, which allows you to place the paper in your institutions repository (or funders repository) 6 months after publication. [0]

[0] http://www.nature.com/reprints/permission-requests.html

Some people don't even realize that researchers themselves pay money to publish a scientific manuscript, whether closed or open, and not the other way around. So the publishing house (likely) recoups their costs of editing and publication with the publication fee paid by the scientist. Remember peer reviewers are not paid either; editing staff, typesetters etc. are the ones that do get paid.

So my issue with the pay for open access model is researchers have to use their grant money to pay even more to these publishers just to unlock the articles for anyone to download. It doesn't solve the underlying problem but rather just throws fuel on the fire of publishers and their mighty monopoly. They are for-profit entities.

A truly effective mandate to shift the paradigm of publishing would be to mandate publication in completely open access journals, not ones that allow it only with an extra fee. Then more and more scientists would shift to these journals inherently raising the influence of completely open access journals.

> So the publishing house (likely) recoups their costs of editing and publication with the publication fee paid by the scientist.

This is not universally true, I just checked Nature Genetics (as it was mentioned before) and it appears there are no publication fees unless you have colour figures.

Do you have any statistics on this? I can only quickly find APC data for open access.

Though the charts don't break it down by each journal specifically, here's a decent survey of the prices from a few years ago: http://www.nature.com/news/open-access-the-true-cost-of-scie...

Another fix for the second problem is a institution-wide pot exclusively for OA fees, so that they don't come from your own pocket. Have only seen that once.

All you need to do is email the corresponding author and they'll send you a PDF. (Works a bit less well when the author is dead.)

You usually don't have to go that far.

Why don't we talk about the elephant in the room, that is, Sci-Hub?

This is the best method - but many don't check email often, don't respond, or aren't at that university anymore. My success rate is ~%50, which probably says something about the age of research I'm looking at.

IanCal's and a_bonobo's answers are correct, but to go further back along the line of causality: all of this is a holdover from the pre-internet days, when publication in paper journals was arguably the only or most efficient way to disseminate research results. Journal publishers were thus able to establish themselves as the gatekeepers in all the ways IanCal and a_bonobo described.

Why isn't it obviously motivated by money? Is evidence lacking?

Because if "the public" (people without a vested interest in perpetuating a given field of research) had easy access to the data, methods, etc rather than needing to rely on press releases, they might come to different conclusions.

Reply on news.ycombinator.com