Hacker News Reader

I don't think anyone disagrees that the raw data should be eventually made public, and preferably sooner rather than later. But initially it's all about the message: you need time to properly analyse the data and release the results responsibly. If you just released the raw data first, how long do you think it'd be before all manner of crackpot headlines popped up? There'd be a race to be the first publication to get something interesting (and likely wildly incorrect) from the data. The fact is, most people aren't qualified to interpret the data.

As for the prestige? That's important too. Not because people care about the prestige in and of itself, but because it enables further scientific research by helping to acquire funding. It's a very principled view to take that scientific research, in theory, should be separate from ego, but in the real world the fact is that reputation matters because not everyone has time to independently verify everything. Ask any practicing scientist and they'll tell you that they're far more willing to trust certain research groups than others, and that's ultimately what the "prestige" is all about.

> The fact is, most people aren't qualified to interpret the data.

so what if most people isn't qualified? Data is data, and can be used - even for crackpots who wants to use it. If these crackpots publishes something wrong, i m sure they'd be pointed out and either ignored or shunned by the publisher(s) anyway.

> The fact is, most people aren't qualified to interpret the data.

All too often, when that is given as an excuse, the people analyzing the data turn out to have made mistakes and the general public does have qualified individuals who can, and will, point out the errors. Alongside this, frequently this raw data is not shared with the reviewers, who should be qualified and should need the data to do a proper review. Unfortunately, this often results in a rubber stamp culture of reviewing rather than thorough peer review to validate the results.

I agree with you in general terms (quick release is best).

But I think your post is in error because it implies that the raw data is not eventually made public. To my knowledge, this is not true -- especially for Earth missions, I think all that data is public. Perhaps the level 0 data [essentially, raw telemetry] is not, but everything from level 1 up will be public.

The data-release culture is very discipline specific. In planetary (e.g. Mars), there are embargo periods -- partly to ensure the systematics are understood. In cosmology, a four-year embargo period for Planck was ended just this summer. This kind of makes sense -- they didn't want half the astrophysicists in the world off chasing a phantom caused by their very sensitive noise-suppression tools. But I understand the community was getting pretty impatient. So, there is push-back.

For Earth missions, after an initial calibration period, I think all the data is made public as fast as it can be processed. Indeed, MODIS even broadcasts the raw data to whomever wants to pick it up (http://modis.gsfc.nasa.gov/data/directbrod/), like disaster responders.

Like I said, it's discipline-specific. For disciplines where time matters (Earth, space weather, seismicity/geodesy, time-domain astronomy) there is a real advantage in releasing data fast, and that's what has happened. I could give websites in each case where data is released in near-real time, say, less than 1 hour after taken.

I'd love some examples of where to find this data?

Someone asked for links, here are a few:

geodesy: http://pbo.unavco.org/data/gps/realtime

Earth: http://rapidfire.sci.gsfc.nasa.gov/realtime/

space weather: http://www.swpc.noaa.gov/SWN/

time-domain astronomy: http://voeventnet.caltech.edu/feeds/Catalina.shtml

The last one is pretty neat. It contains an XML event summary that can be used to target a telescope automatically, if you think that event is of interest to you. The events are automatically found by an image-processing front end which taps into a near-earth-objects survey. In other words, they did the hard work already to find unexplained brightness variations.

Don't forget about PDS! The Planetary Data System is where all NASA-funded data ends up eventually: http://pds.nasa.gov/

You're right, PDS is very germane to the current topic. I was thinking real-time sources only.

I worked on the LRO project for 4 years (on the LROC science operations team, we ran the cameras). LROC released, and continues to release, all of the data to PDS where it is publicly available. A release happens every 3 months, with a six month lag. The data are released as a set, with raw[1] and calibrated data released concurrently. So the raw data actually are available, just not immediately.

Now there's some difference between LROC and Curiosity. A Mars rover is much higher profile, of course. But it's also a snail's eye view. I imagine timeliness is a much bigger deal with rover data. Whereas with LROC (and the other LRO instruments) there's a massive data set that can fuel interesting research for years to come. So who comes to get the raw data? Very few people, as it turns out. Who but a planetary scientist getting paid to do Lunar stuff has the time, ability and inclination? Also, DLR (German space people) managed some really awesome original stuff with our data, and not having it as soon as it came down didn't seem to hamper them.

Notes:

1. By "raw" I mean EDR (Electronic Data Record), which may be reorganized to facilitate access and conform to PDS format, but contain unaltered, unprocessed measurements.

But, considering this research is tax-funded (that includes my taxes), shouldn't I, a funder of this research, be allowed to access the raw results?

(disclaimer - I'm a former NASA contractor and current federal employee, though not at NASA)

Yes - and, in fact, the Administration released a memo on this topic earlier this year ("Increasing Access to the Results of Federally Funded Scientific Research" - http://www.whitehouse.gov/sites/default/files/microsites/ost...). However, each agency is responsible for developing their own individual plan on how to carry this out, and I don't believe NASA's has been published yet.

There is a balance between accessing raw results and everyone accessing the means by which those raw results are accessed.

I say this as an Information Wants To Be Free t-shirt wearer.

No one hates someone who holds something if they know the holding of it is something everyone wants.

But why should I, as a non-American, have access to that same data? Or this documents made free in the OP?

the fact is, as a non-tax paying non-american you don't have the privilege to access it, but for the "generosity" of the US Gov't (and the fact that enforcing copyright for these things is both costly and not worthwhile).

Surely the solution to this problem is to put the data into the public domain after a reasonable embargo period of maybe a few months or a year?

This seems fair to me, but I'm not the one who's being told, "Look, find a breakthrough in a year or you're going to get scooped by an amateur."

It would be ironic and painful if open data policies resulted in scientists leaving the public sector, and I'd really like to see more open data.

Seems like the guy requesting that certain data be collected is due some measure of academic credit for having the foresight to know what question to ask, even if someone else "scoops" them on the analysis.

Sure, sometimes the right question will be starkly obvious, but no system is perfect.

> Seems like the guy requesting that certain data be collected is due some measure of academic credit for having the foresight to know what question to ask, even if someone else "scoops" them on the analysis.

I'd agree, but my knowledge of the academic community is that credit goes by published papers and, secondarily, how often you get cited. This seems like one of those things where we just don't have a great system for giving people credit, and the one we do have is problematic.

(Wealth is like that, too. Also, resumes.)

Which is where things like figshare [0] work well. You can publish datasets/figures and you get a citable DOI for free. I published code there that my wife cited in her PhD, without having to do the usual computer science thing of writing a paper about it and having people cite that.

Disclaimer: I work for a partner company.

[0] http://figshare.com/

I would recommend: http://www.gigasciencejournal.com/

This is exactly how it works and that data can be found here on the MSL PDS Archive: http://geo.pds.nasa.gov/missions/msl/index.htm

I think the standard release delay is six months. Though I'd say its less about exclusive access to data than it is about ensuring the consistency of that data. MANY things change in the early days of a mission.

I personally feel most government scientific data(non- defense) should be released immediately.

I have a feeling there's more than a few Einstein's out there delivering pizzas. Go ahead and laugh, but I feel some big mysteries will be solved by dropouts, and the self taught.

> The reason the data is not released immediately to the public is both understandable and frustrating: scientists around the world compete for control of Curiosity's precious mission time.

The same applies for the giant telescopes (KECK, VLT, ALMA, etc). A scientist applies for some time doing observations with the telescope, and gets exclusive access to the data for 2 years before the data gets released to the public.

But I agree with you too, all scientific data that is acquired with taxpayer money should be released to the public at once.

A related paper, "The Virtues of Frugality - Why cosmological observers should release their data slowly": http://arxiv.org/abs/0909.2649

All government data, excluding true national security issues and PI data, should be open and free by default. You shouldn't have to file a FOIA request for anything (in the US), it should be publicly accessible.

Prestige? Prestige?

Science is not about prestige. It's about science. If you want prestige, go become a jockey, or the world's best hooker. Too many "scientists" are in it for the holidays, money, and "prestige", rather than the, y'know, SCIENCE.

I'm afraid it's not "understandable", it's nasty, selfish, and contrary to scientific advancement.

By not releasing data, they're preventing the internet hivemind from doing its thing - and it has managed some pretty impressive things (NEAR target spotting, anyone?).

I agree with your message in idea but sadly in the world we live in; without prestige you don't get founding.

well we don't get paid jack. 99.9 % of the science undertook is boring, inconclusive or negative results.

The only carrot is prestige, and the only perks are holidays ...

I dunno why you suggested money because there is none (in wages).

I expect a big reason for not releasing the data is ... who is gonna do the necessary steps to release the data? You want to take a scientist off his role to build a website and database?? We have little administrative support (probably JPL has better, but still, yet another distraction that eats time from our main job that pays too little). The people that make websites for a living get paid better the scientists, so its a bit harsh to expect scientists to constantly be building website to share data whilst paying them under the industry norms.

In his article, the OP mentions "$2.5 billion taxpayer dollars that made it possible". Sorry you don't get paid as much as you'd like -- that doesn't give you the moral high ground. Doing it for "prestige" still isn't a good reason.

At a cost of 2.5 billion, I don't think it's too harsh to expect a small budget for releasing the data publicly.

I sympathise with your goals but you're being a douche. Science is a crap career if you're chasing money. You're a smart college graduate with excellent grades who spends 4-7 years doing an apprenticeship for piss poor money, then a postdoc, for three years, maybe a second postdoc, and if you're lucky you then get a tenure track position. Now you have seven years of earning high school teacher money but working 60-80 hours a week and at the end of it there's an excellent chance you'll be fired.

People do science for love or glory, not money. (Except people from 2nd or 3rd world countries studying or postdocing in industrial nations. For them it's a sweet deal.)

(Except people from 2nd or 3rd world countries studying or postdocing in industrial nations. For them it's a sweet deal.)

Reminded me of this article:http://wuphys.wustl.edu/~katz/scientist.html

"I expect a big reason for not releasing the data is..."

No, no, no. The data are released. The mission pays people to properly calibrate, process, and format the data. NASA then pays to host the data indefinitely. The basic link for planetary data is http://pds.jpl.nasa.gov . There are other sites for Earth, solar, astrophysical, etc., data.

Source: I know the people who run PDS.

The kerfuffle in the thread is about derived papers, in some cases by non-US collaborators, not the data itself. To the extent it is about data, it's about the post-capture embargo period of a couple of months (in the case of initial Mars data).

> Too many "scientists" are in it for the holidays, money

Yeah no. Prestige within the profession yes, these two you probably get more of (and less politics and having to fight for grants) piloting garbage trucks.

And talking about grants, that's a big part of what "prestige" is about. There is also an ego-stroking component but the scrap-fighting of obtaining grants (and keeping them) requires publication and forces many into "prestige-seeking" not because they want to do it but because without it they get no grants and without grants they get to do no science.

> I'm afraid it's not "understandable", it's nasty, selfish, and contrary to scientific advancement.

Fix science funding and 90% of the "prestige" issues will be fixed.

I think the answer to this question is better left to those doing the actual science.

I agree with your sentiment, but sadly science (or rather, the doing of science) is about prestige.

Put yourself in the shoes of a funding agency. If you have two grant proposals, one from somebody at an institution in the middle of the country you haven't heard of, or someone with a strong publication record in big-name journals --- who are you going to award the money to?

A scientist in academia needs funding and publications for tenure. To get publications, funding is needed for equipment, postocs, etc. To get funding, a list of publications is needed so that people recognize your track-record. It keeps on spiraling. To paraphrase from Pratchett in Carpe Jugulum: "Human families raise their successors, but a [s/vampire/professor] is raising competitors." There's a limited funding pool that the researchers are fighting for.

It doesn't seem much better in the industry side of things. For instance, Google [X] is an extremely prominent research lab. However, why? They work on ??? something that no one really knows about for years and don't publish anything. The scientists there may come out in 5 years with something really cool, but if it doesn't pan out, the Google machine will eat it up and the public won't see. Science there is more about the company's prestige than science.

Perhaps prestige needs to be replaced with something else like commercial value. Whilst this will not work for all topics, perhaps it could be one variable in a group of others that should be taken into account.

Many scientists hate the commercial part of their research. In fact that's sometimes the main reason why some of them choose to stay in (relatively) poorly paid academic positions

Well, prestige is used as currency precisely because we as a society need a mechanism to generate knowledge that isn't (immediately) commercializable. I, for one, would find it tragic if we lost that.

The problem is that if you aren't guaranteed some payoff (in terms of publications that will make it possible for you to get more grants and continue your work) from your work developing the experiments and getting the data, what's the incentive? It's not about being greedy (though there are certainly examples of that) but about encouraging good science.

And I don't think people are worried about the internet hive mind, they are worried about their competitors who did not get the grant to actually collect the data.

That's prestige as in your boss says, "Okay, you published in a prestigious journal, which means you must be doing very good work -- which I can't judge easily -- so we're not going to fire you." Prestige is how scientists can be evaluated by people who are not part of the same very narrow speciality. Scientists who don't get prestige don't get to be scientists very long.

I don't understand your argument at all---scientists are in it for the money? Really, what money is that?

But what I find really interesting from your post is the sense that scientists are supposed to be saints. You don't say this explicitly, but I get that sense from you implying that it is somehow bad for a scientist to be selfish or to want to be paid. (Or to want to be a scientist at all, since if you can't get funding then you can't work anywhere.) Are scientists all supposed to be volunteers?

Myself, I have no problem with professional scientists who selfishly want to get paid for their work. In most areas of science, the internet hivemind has had zero impact. Scientific advancement would stop dead if we stopped paying scientists.

> no problem with professional scientists who selfishly want to get paid for their work.

nor do i, but if their only method of getting paid is to restrict others from using raw data till they've made their break through "discovery", then they are not worth what they are paid!

Scientists with "prestige" get positions at nice universities, labs, raises, smart people who want to talk to them, smart PhD students who want to work with them, etc.

Scientists without "prestige" end up being poorly paid, teaching intro to chemistry at a community college to students who just want a piece of paper to help them find a job, and unable to get the funding to even buy a laptop, let alone a lab.

So prestige matters a lot to scientists, whether they like it or not.

The only way you have a chance at solving this system is by changing the reward metrics; find a way to make it more rewarding to release the data then to keep it for yourself. But if you just mandate the data be released, then the scientists will be actively discouraged from collecting data, and probably move away from the field.