U.S. Science and Engineering Research Flattens

Invisible Pink Unicorn writes "The National Science Foundation is reporting that the number of published U.S. science and engineering articles plateaued in the 1990s, despite continued increases in funding and personnel for research and development. This came after two decades of continued growth. Since then, flattening has occurred in nearly all U.S. research disciplines and types of institutions. In contrast, Asian and EU research had significant increases in this period. They do point to one positive for the US, however: article quality. According to one of the researchers, 'the more often an article is cited by other publications, the higher quality it's believed to have. While citation is not a perfect indicator, U.S. publications are more highly cited than those from other countries.'"

Re:Fuck the USA

[Xiph]

I'd make the guess that language matters more for citation than for acceptance.
Acceptance only evaluates the scientific merits, citation requires the paper to have given the citing person insight.

I'd love to see this compared with british statistics, and possibly french (since the majority of non-english journals i know are french)

Output of papers isn't too useful

[ejito]

There's still growth, the article just doesn't state exactly how much - probably ~0.65% annually, from the information given..

Anyways, as the article states, papers aren't always the best indication of actual information output. It's common practice for researchers to "recycle" papers, adding a bit of new information on top of the bulk of previous published work. It's in a researcher's best interest to limit the amount of both papers AND information, as to keep a steady stream of output (and keep their job). Tracking citation count seems more accurate in representing useful information output. It'd be even more accurate if we could somehow track actual implementation and use of the information.

Re:Output of papers isn't too useful

[15Bit]

Citations aren't really a great way to monitor quality either, cos each of those "recycled papers" will cite previous ones. So you should look carefully and see WHO is citing the paper, as often it will be the authors themselves citing earlier work. This is not necessarily a problem (it may be that there are very few people working in a field, and self-citation is unavoidable), but some scepticism is required.

A much better indicator of paper quality would be a weighted combination of the Journal quality (impact factor is a common measure), citation count and something like the number of different citing authors. You can probably add some other factors too.

How does funding factor in?

[moosesocks]

I wonder what it would look like if we also plotted the funding allocated to the NSF alongside the number of papers published.

The NSF has had some serious funding woes since the 90s that very well may be causing this "draught" -- I wouldn't even go as far as to completely blame it on the Bush administration either (although they certainly did contribute).

As far as physics research goes, Clinton's cancellation of the already partially-constructed SSC easily set the entire field of particle physics back by 20 or so years. The LHC, which is being constructed in Europe as its "substitute" isn't even remotely as big or powerful as the SSC was originally planned to be.

Re:Who's wondering why?

[gtall]

In addition to those problems, science these days is funded by business school product who have understanding of science and don't ever expect to. They won't fund anything unless you can point to an application. That means pure research isn't being funded. The result, as the article points out, is that we are leveling off in publications. No pure research is being done at the high end and hence it never gets developed into the mid and low end where the applications are. To put it another way, the mid and low end are simply rearranging the pure research deck chairs handed to us by previous generations.

It isn't just business school product that have this attitude towards pure research. A fair number of slash bots also have it. One gets the feeling they believe research is something written up in books and there isn't any reason to write any new books.

Gerry

Citations

[jsse]

We love to cite US research paper because they can be searched electronically, while the others might be required getting down to the microfilms, or worse, papers.

And yes, this is the quality that counts - the quality of storing and indexing research papers.

The truth is ...

[jopet]

that there is no objective way to measure the quality of research. For this, one would have to know what "quality" means and already there, opinions are highly divergent. But of course the beancounters of the money-giving institutions need some yardstick and so there are and have been different yardsticks in different countries and at different times. Scientists will quickly adapt to any yardstick: if you get money and jobs by publishing a lot, they will publish a lot. If you get it by getting cited, they will get cited. If you get it by not publishing and having lots of patents or company cooperations instead, this is what will happen. None of this will ensure research though, that will advance the state of the art. Most of these regulations and rules imposed by beancounters will simply take the time and energy away from scientists who want to do research.

Ultimately, science, like art, often has to be useless to be good. In many cases however, useless science might eventually and surprisingly turn out to be quite useful indeed, practically. Take number theory: what beancounter of the world would have guessed that this esoteric branch of pure matematics would once become the fundamental force behind e-commerce, authentification and authorization systems and other applications of electronic cryptography?

Non English language research material

[Anonymous Coward]

I wonder if the higher citation rate of US articles is due to the fact they are written in English and therefore more accessible to a higher percentage of the scientific community? Presumably a lot of a country's scientific publications will be in the language of that country and it would be reasonable to assume that an article published solely in, of example, Russian or German, would be less widely cited outside the Russian or German speaking communities...

This is actually a phenomenon in historical research. Over the last few decades there has been much written about various major events of the 20th century, such as WWI, WWII, the Korena and Vietnam wars etc... and most of it has been from the American/UK point of view citing mostly American/UK sources but not French/German/Japanese/Vietnamese ones since the historians who wrote this material were often unable to read these languages. Instead they often relied heavily upon abundant US/UK sources but only upon foreign sources where English translations of these were available. Unfortunately these translations were often limited in scope and accuracy. The result has been that academic flamewars have sometimes ensued when historians from these non english speaking countries started mixing in citing various sources that had never been translated into english and in the process often overturned versions of events that had previously been accepted as patriotic gospel by historians in the US and UK. To few historians seem to bother to take all sources into account before arriving at conclusions, impartial historians seem to be as rare as honest politicians.

Immigration Issues

[widman]

Even the 50s were better on treating foreign teachers and researchers. Now you get, with a lot of luck, a non-citizen Green Card. You are constantly bullied by random uneducated locals. And, if you are lucky there are many others in your same situation around you, you end up in a virtual ghetto. F*** that. Europe has it's priorities better now. USA lost it. Go build your racist and unfair wall to keep off the real native North Americans from their own land (check the "Mexican War".) Mod me down if it touches your right-wing heart. But that doesn't make truth go away. Read the title.

Re:We always used foreign scientist/engineers

[moeinvt]

"No child left behind?"

I think a major part of the problem is that the U.S. public education system has an overwhelming focus on bringing the slow learning underachievers up to par. Far too little is done to accelerate and unleash the potential of the best and brightest. Raise your hand if you were in the public education system and got all 'A' and 'B' grades while rarely or never bringing a book home with you. I susepct that most of the readers here were taking AP or college prep classes as well.

With idiotic programs like "No child left behind" the entire herd has to move at the pace of the slowest member. For example, I know an elementary school teacher that has a small group of students who are children of recent immigrants. They barely speak English, yet the school is supposed to make sure that they don't get "left behind"? Where do you think she needs to focus all of her extra effort? The phrase is emotionally pleasing, but the implementation has serious negative consequences (I HOPE they are unintended, but I'm not sure). I think that kids SHOULD be left behind a lot more frequently than they are.

I'd be in favor of getting the Federal government out of the public education system entirely. We should eliminate the Dept.of Education and distribute the entire department budget as block grants to the states for the next couple of years.

This is a network effect, not a sign of quality

[budGibson]

The statistic they're citing, number of citations, is the same statistic that underlies page rank. As we've discovered from the industry that has grown up trying to game google search rankings, being well connected by citations is really only a sign that you are well connected.

In a system where your prestige depends on being connected to well connected others, being among the first to be connected has its advantages. Others will want to be connected to you in order to show that they are also connected. It should be noted that after WWII, the US was really alone in the western research world. It's still accruing benefits from that.

I wouldn't be soothed by the citation statistic. At this juncture, it's an historical artifact.

Re: Citations

[TheRaven64]

Review is almost always done anonymously, and by reviewers who aren't assigned until the paper is received. How would you know who to suck up to?

Typically, there are only a dozen or so people who are qualified to review a given paper. Each conference or journal will be for a particular area within a broader field, and each paper will be within a narrow specialty within this area. There will usually be one reviewer doing a general 'is this interesting to the community at large?' review, and one doing a specific 'is this novel?' review. The latter will be one of the dozen who is qualified to know the answer. Of this dozen, you can usually narrow it down to two or three who have some history with the conference / journal, and so will be asked for review. It's not 100%, but you can usually make a pretty good guess about who will be reviewing the paper, and gratuitously cite their work.

Paper acceptance these days reflects a lot more on the ability of academics to play politics than on the quality of research. I've seen papers rejected from well-respected conferences that were far more interesting and novel than several that were accepted.

My feeling is that peer review should happen after publication, not before. I believe the physics community does something like this. I would like to see any paper anyone wants to write made available online, and then commented on and ranked by others in the community. The top few papers in each field as a result of this could then be collected together and released in print.

U.S. Should Separate Research, Universities

[littlewink]

Research and teaching should be separated. That is, rather than maintain the current U.S. educational system, wherein universities do most of the research, we should move to a separate system:

• Universities - teaching only,
  
• Research institutes - research only.
  

Why do this?

Because the current system sub-optimizes both the researchers' and the teachers' time. Excellent researchers are busy instructing students rather than working in the lab; excellent teachers are writing grant proposals rather than focusing on the classroom.

Researchers as a group don't want to teach nor is teaching an optimal use of their time. The best teachers are usually not excellent researchers. The two fields are so different that it is uncommon to find an individual excellent in both.

Re:US tends to attract top researchers

[Alomex]

The US continues to be attractive because it tends to offer the best facilities (laboratories, datasets, computers, funding) in the world.

Actually with the latest cuts in funding this is no longer the case. Since GWB took office, the NSF funding per researcher has gone down and nowadays there are top notch american based scientists whose last n funding requests have been declined. In the past requests from such researchers had much higher rates of success.

Re:Who's wondering why?

[Otter]

You should hear what scientists have to say when their work is linked here and you idiots helpfully "critique" it...

Re:We always used foreign scientist/engineers

[mh1997]

Actually, go to http://www.johntaylorgatto.com/underground/ [johntaylorgatto.com] and you will see (very well referenced) that the American education system is designed to fail on purpose. John Gatto's book, Underground History of American Education is free at the site.

Re:Output of papers isn't too useful

[tmosley]

The British are still at the top table, the Spanish and French were military superpowers, not a scientific ones, the Chinese succumbed to monoculture, and the Persian empire was destroyed by invading armies long before the invention of the atom bomb. If ICBMs had been invented by the persians, we would all still be speaking Farsi.

No nation that is well armed with nuclear counterstrike capabilities will ever be invaded. The political boundaries of the world will remain as they are until a highly effective nuke shield is developed.

That being said, why the outburst? You seem to have some sort of inferiority complex, lashing out like that for no great reason. No, the US wasn't "chosen by God", it was chosen by circumstance. And the conditions that exist in the world right now tend toward unshakable stability for the US and all other great nuclear powers. The only way to reduce the power of one of these nations is for their economy to collapse in a spectacular fashion (a la the Soviet Union), but even with that, Russia remains a great power with "a seat at the top table".

Re:Also

[timeOday]

There's a lot of truth in that joke. I recently attended a conference in Atlanta where not one of the presenters was a native English speaker. I couldn't understand them and they couldn't understand each other, but it made me think about how much harder it will be for us Americans if we lose the "home team" advantage of having so many important conferences relatively nearby and the most prestigious conferences and journals in English. It's amazing how well non-native English speakers do despite the additional challenge of a language barrier. Could we do the same? I find getting published hard enough, what if I had to write in Chinese? Success is a self-reinforcing thing (I guess that's why we have words like "hegemony" and "monopoly") and this implies there is a tipping point - the leader won't fall behind until they're truly inferior, but then the fall will be quick. I hope we're not that close to the brink but this is no time to rest on our laurels.

Re:We always used foreign scientist/engineers

[vtcodger]

***Far too little is done to accelerate and unleash the potential of the best and brightest.***

Y'know what, That wasn't true in the 1950s when schools had to deal with the baby boom that caused huge classes and a relative shortage of classrooms and teachers and it's not true now in a lot of schools. I've actually worked recently in a rural school (K-8 -- 300 students). Yes, a lot of effort is expended on the low end. But the really exceptional kids on the high end aren't ignored. They get guidance, special tutoring, access to advanced materials etc. The kid that had real acting talent ended up at one of the top musical theatre schools. The kid who was good with computers got special tutoring, administrator access, and a recommendation to the IT folks at the high school (He's come in first and second in two annual statewide tech support contests since he got to the HS). A lot of that is informal and wouldn't show up if you looked at statistics. But I really don't think the really top flight kids are ignored. I'm not so sure about the merely pretty good. They indeed may be being shortchanged.

I do think there may be a problem in the US with the culture rather than the schools. It's hard not to agree with Bob Parr (Mr Incredible) "They keep finding new ways to celebrate mediocrity," We don't seem to care much about superior performance except in athletics.

***With idiotic programs like "No child left behind" ... ***

With that I couldn't agree more. NCLB may help with the really awful schools ... for a while. But it's going to be a disaster for the schools that were doing pretty well and that'll get worse. For a while teaching kids to pass tests will help to show annual improvement. But it won't be long before the requirement to show improvement every one of a dozen or three category/age group combos one year out of two will result in good schools being ranked as failing. (At least that's the way I'm pretty sure that's how NCLB works).

***I'd be in favor of getting the Federal government out of the public education system entirely. We should eliminate the Dept.of Education and distribute the entire department budget as block grants to the states for the next couple of years.***

Amen Brother!!! I'm a liberal in modern terms, but I don't think the central government has much place in either education or law enforcement. Yeah, some states and local governments may not be much good at either, but at least we'll have some decent schools and police forces in some places and people can still vote with their feet. In education in particular, there is a tendency for large organizations to create monumental bureaucracies that devote an inordinate amount of effort to creating and 'evaluating' piles of incomprehensible, jargon laden, basically silly documentation. Bad idea overall. I submit that any organization that can't (or won't) communicate in simple English should be phased out ... sooner rather than later.

Re:We always used foreign scientist/engineers

[Anonymous Coward]

> But that is definitely changing. Corporations don't value research now. Some do - biotech is
> probably now the best example - but by and large, US companies are obsessed with acquisitions,
> splits, layoffs and wage cuts in order to provide the fastest, greatest profits to
> shareholders.

And that is catching up to them, look at Motorola. The current CEO has gutted R&D, relying on slick marketing, cost cutting, acquisitions, and repackaging of the Razr. But it isn't working, one can find any number of articles detailing the woes of Motorola. However, this is nothing to gloat over, major companies failing because of their mistakes won't bring anything good to our country. And they aren't likely to change course either, consider, the Motorola CEO just believes he hasn't been aggressive enough with his approach, not that his approach is fundamentally wrong. I doubt most CEOs are any different in attitude. I think our best hope for the future of research will lie with smaller, more dynamic businesses that focus on emerging technologies like biotech.