Speaker: Seeking Better Science
116 Responses
First ←Older Page 1 2 3 4 5 Newer→ Last
-
Sciblogs rounded up some highly diplomatic responses.
"The Sprout" at the Standard saw occasion for a depressingly partisan slagging of Gluckman.
And one other thing strikes me. I interviewed Gluckman shortly after his appointment and it was made clear at the time that his role would not extend to commenting on science funding. I don't object to him doing so now, but the line appears to have been pretty fuzzy.
-
a depressingly partisan slagging of Gluckman.
Good grief that's awful. Professor Gluckman may have his faults, who doesn't, but there can be no question he is a very good scientist and really does believe that science can and should contribute to society.
Science funding isn't a National v Labour thing. Note Labour only promised to deliver lots of money to science just before they were voted out. Science funding has suffered for 30 years. That's plenty of time for either major party to have changed things.
-
Long term science vs Short term engineering
Election cycle.
'nuff said.
Science in NZ is well and truly fucked, and this is deck chairs on the Titanic stuff.
-
Good grief that's awful. Professor Gluckman may have his faults, who doesn't, but there can be no question he is a very good scientist and really does believe that science can and should contribute to society.
Indeed - perhaps someone should man-splain to 'Sprout' that advisers do exactly that. They proffer advice, which their political lords and masters are under exactly no obligation to accept. Civil Service 101.
-
Thanks for that roundup Bart, most interesting.
There are good people in the labs and they do sometimes squeeze the time and money to do good novel work, but that work then gets ignored because there is no business plan with a clearly identifiable product.
I suspect this is most of our problem, really. For business to fund science, it either has to be a sure bet (in which case it won't be novel) or the business likes taking enormous risk (not so many of those here), or it's a huge business that can lower the risk by funding a lot of different research directions (again, not so many of those here either), for diversification. Which is why government support is needed if we want to attract big programs (which carry big risk). Yes, it puts the risk mostly on the taxpayer, and there is a strong case to ask why they don't get direct dibs on the returns, if there ever are any. But any model in which the government makes investment and expects a big cut of the return isn't one that businesses want to jump on, because they don't get their handout of the socialized risk. Which leaves it in the hands of the government to try to sell the technology, something that mythology suggests they are really bad at doing, either failing to sell, or selling for far less than it's worth.
It's a tough problem, the same problem NZ has with everything - scale. Do we just bite the bullet, up the government investment, and accept that it's a handout to business, with the return coming through tax revenues? I can't see it working any other way.
-
I can't see it working any other way.
It does, after all, seem to be the way that science funding works in the USA. For a huge amount of it, the research is hidden in the military budget, which is enormous. If/when they come up with amazing things, as they frequently do, it gets sold into private hands. This seems unfair and inequitable, but it does produce a shitload of science.
-
Yes, it puts the risk mostly on the taxpayer
And given the size of NZ and the fact that most big companies can more easily and efficiently do their R&D in Aus I can't see that changing. What this funding package is trying to do is encourage business to invest in NZ R&D, and it might work.
But the risk is not as bad as you think. Despite the fact that you can't predict which project will produce the new businesses you can say that if you fund enough good quality science the you will produce new businesses.
Yes the immediate payoff is for that business but if you look at data worldwide you can see that the country as a whole benefits as well. Hence the strong relationship between GDP and R&D spending. This is the point Professor Gluckman has made in most of his speeches.
The key is to fund quality science and fund enough. Where we've failed as a country over the last 30 years is by pretending we can do more with less money by picking winners. All we've proved is we can do less with less money.
-
Thanks for the summary, Bart.
I agree we just need to fund all the good science we can---and some of it---10% seems to be a figure that emerges from international experience over decades---will go on to make loads of money.
The rest, of course, will merely increase knowledge, educate more people better (by being a part of great teams, for example) and give NZ some standing in the science world (rather than being the cargo-cult we sometimes seem to be currently). All of this, I realise, is a waste of money :-)
I also agree with the point about putting more into Marsden. That is where we REALLY need the people (and therefore the money). But, some of the increased (in my dream-world) Marsden funding should be used to run it in a transparent and morally defendable way. Currently, it runs in two "rounds".
In the first round you write a page or two describing your idea, the team and why it needs funding. This gets looked at by a panel of our peers, selected by the fund. In the panels I've submitted to, there has only ever been one person remotely in my area, and NEVER anyone who is actually expert or an enthusiast for my area.
Each panel member (reports-back tell us) spends a few seconds or minutes on each application (the workload is so high). Then, a meeting of the panel is held, and applications get either selected or not. All this gets manipulated and scaled in ways I don't understand, to come up with final invitations to apply with a full application, or not. Full applications are long documents and get reviewed by experts.
Now, here's the worst bit---if you do not get through this round you get NO feedback (just a one liner saying sorry). So, from year to year there is no way of knowing if your idea is mad, bad or simply misunderstood. How a science system can work without the most primitive of feedback mechanisms, I fail to see.
So, tens of millions a year get distributed by a system which, at its first stage, is totally opaque. No accountability, no opportunity to learn from mistakes, no "improvement mechanisms".
If one enquires or complains, the only response, essentially, is: "we can't afford the cost of doing it better"!
-
I've spent almost all my career (30 odd years) working for startup companies, mostly in Silicon Valley.
For me R&D is largely engineering, not science - it's all about people having bright ideas and getting them to market before anyone else does. So scrapping Labour's $600m R&D tax credit and replacing it with $300m of largely science funding targeting at benefiting only medium to large companies (not startups at all) is, well, quite useless. It doesn't help me get my bright idea to market at a time when local funding has largely dried up.
I think that science funding IS important, but, in my industry at least, it's not the same as day to day R&D - good science takes time, years, in a world with product cycles that are measured in months these are apples and oranges.
So I'm eagerly waiting for National's plan to address engineering R&D ..... but not holding my breath
-
But, some of the increased (in my dream-world) Marsden funding should be used to run it in a transparent and morally defendable way.
I'm not sure that I agree. I think if they could fund a sensible proportion of applicants, there would be no need for pointless hurdles. It's rather like using academic grades to select medical students. It's not that it's a good selection tool, it's just an easy way to weed out the pool (some overseas universities actual use random ballots for people over a much lower grade threshold).
In fact, based on this model, perhaps they should have an annual lottery, and if your name comes out of the hat, you get to submit a full proposal. The longer your name has been in the pool without selection, the more entries you get. Or they could triple the Marsden fund. Obviously, I'd prefer the latter.
-
James, I agree.
In fact, everyone I meet and talk about Marsden with agrees that it is, in fact, just a lottery.
The trouble is, it sells itself as a sensible, defendable, reputable way to dole out public money for science.
So I guess the problem is the mismatch between what it tells us it is (mainly, I suppose, for the politicians' benefit, so they don't look too closely) and what we actually see---a lottery pretending to be a science system.
-
The worst thing about it being a lottery is that it is costly and inefficient. It's a massive blight on researchers' and panellists' time for little gain. Having an actual lottery would be cheaper, fairer, and probably have a pretty similar hit rate in terms of funding actually good projects versus crap projects.
-
The key is to fund quality science and fund enough. Where we've failed as a country over the last 30 years is by pretending we can do more with less money by picking winners. All we've proved is we can do less with less money.
If only we could treat that as a meta-science experiment.
I agree, of course, that the flow on benefit from more science funding does find its way into our society in many ways. But the links are not direct, and this nation has a habit of demanding such links. The user-pays mentality. People see it as corporate welfare. But they don't really have any realistic alternatives. I'd just as soon allow the corporate welfare, as yet another kind of social welfare, because, like social welfare, the flow on benefits are in so many dimensions towards the kind society I value.
-
everyone I meet and talk about Marsden with agrees that it is, in fact, just a lottery
Steve I understand your frustration but I do disagree. That first round of the Marsden is very hard because it is only one page and you need to excite most if not all of the panel. In a really narrow field that is really difficult. I know several people who have sat on those panels and they struggle with that first phase as much as you do. You need to convince (in one page) a dozen different scientists that your work as a good chance of producing a Science or Nature publication.
But as bad as that first phase is, the result is that most of the applications that go on to be considered as full applications are very good science. One of the messages I've heard many times is that the panels could fund twice as many grants without dropping quality at all and put twice as much money into each grant with no concern that it would be wasted.
BTW the Marsden process is neither costly nor inefficient. Panelists get a couple of hundred dollars for about three weeks work for the first phase and another 2 weeks work for the second round. It is vastly cheaper than the FRST system.
-
James
[Marsend's]
probably have a pretty similar hit rate in terms of funding actually good projects versus crap projects.
That isn't true. If you look at the results from the Royal Society who manage the Marsden's it shows they are actually very successful when measured in terms of papers, patents and even in generating new business ideas. Much more successful than any other funding system that has been tried in New Zealand.
-
The whole science pie needs to be bigger, but I reckon this is an improvement.
Prior to this announcement, the only government support for business-led R&D was via Tech NZ (tax credits only lasted one year). This was 7% of the total govt science budget. (Tech NZ’s $47.4m of the total $671.5m). The vast majority of funding is still going on CRI staff, contestable funds and and blue skies research as it should.
The new grants scheme promises less paperwork than TechNZ (and 3-year certainty without bureaucrats constantly checking to make sure the business R&D fits into their boxes). The changes to CRIs and funding earlier in the year tried to lessen paperwork for public science too.
Business-led R&D has a greater chance of (market) success because one significant part of the equation, is it able to be commercialised?, is already reasonably answered by the major funder of that research (the business).
While politically it's been sold as an investment in science, it's actually an economic development package.
-
our CRIs are now full of good
(but not great) scientists who are focussed on
taking ideas from overseas labs and applying
them to New Zealand crops/industries.
Which is fine. But it can’t and won’t create new
industries,why? it works well for China and Japan. Wasn't the LCD a British invention?
-
Bart, I agree.
My point is that if you fail at the first round you have no idea if your proposal was viewed as brilliant, good, rubbish, science fiction or what. In such an environment, how is improvement possible?
If you fail this year, how do you know what to do next: put in the same thing (assuming it was viewed as brilliant, but the money ran out)?; change it---then how?---you might make it worse rather than better, but in the absence of feedback, how would you know which direction to move in?
So, you may as well flip a coin---hence, the caricature of a lottery.
The way things were done in the UK when I got grants there was the opposite---perhaps things that did not "deserve" it got funded, but one of the big names in UK science was able to proudly tell me that we fund all the good science and don't miss anything, at the expense of perhaps funding some things that aren't worth it.
Here we seem to fund a few of the good things but are so afraid of funding bad ideas that we are content to miss good ones too. And apart from all those missed opportunities, it is hugely demoralising and makes science in NZ a very, very precarious thing to go into---none of which can be good for the country, surely?
-
why? it works well for China and Japan
Because if you are in that game you need to be faster than China and Japan. They can put 20 PhDs for every one of ours on any project we can think of and hence they are faster.
You can however, collaborate with chinese and japanese groups and that does work.
We are a small country, what works for China and Korea and Japan does not work for us.
That also implies China and Japan are not doing basic discovery science. That's a myth we in the west like to kid ourselves with ... that we are innovative and they are not, sadly it isn't even a plausible myth.
-
Actually given its demographics Japan may be starting to lag a bit - China (and India, don't forget India) are growing in leaps and bounds, still limited by funding in Universities though.
In some ways India's a bit ahead - half of the best chip designers I've worked with have come out of India, not so many from China - more because they've had closer ties to the west for longer
-
My point is that if you fail at the first round you have no idea if your proposal was viewed as brilliant, good, rubbish, science fiction or what
I agree Steve. The only advice I can give you is to ask panelists. they can't and won't be specific but they can tell you what things they are looking for in the first round.
I know from our experience that proposals that appeal to neurophysiologists and to bacteriologists and to developmental biologists and to biochemists are the ones that get through. Broad appeal. And then in the second round really high quality specialist appeal.
I'd love to have a system that had enough funding to let more different proposals through. But because of the effort they put into the second round they can't let too many through the first round and that means you need to impress all the panelists to get through that first round.
That's why I think we need to double or triple Marsden funding. They have been the most successful funding tool over the last decade. that success seems hard to ignore.
-
That isn't true. If you look at the results from the Royal Society who manage the Marsden's it shows they are actually very successful when measured in terms of papers, patents and even in generating new business ideas.
No no. I wasn't criticising the quality of Marsden funded projects at all. I've been to Round 2, and I'm absolutely confident that my project would have been amazing :)
What I meant is that if they made the first round a random draw, it wouldn't produce discernably different outcomes with respect to quality, and there would be a great time-saving for researchers and panellists alike (ie non-monetary cost) -
I know from our experience that proposals that appeal to neurophysiologists and to bacteriologists and to developmental biologists and to biochemists are the ones that get through. Broad appeal.
Marsden is like a two-phase wine judging.
Round 1 is like a wineshow (e.g., Liquorland Top 100). Winners are those that stand out from the 50-100 that you've already tasted. Gold Medal wines aint subtle, but.
Round 2 is like sampling a few more wines in depth. Finesse, elegance and structure become more important than brash. -
What I meant is that if they made the first round a random draw, it wouldn't produce discernably different outcomes with respect to quality, and there would be a great time-saving for researchers and panellists alike (ie non-monetary cost)
I'm not sure if many academics would be impressed if they were told that their application would be picked randomly to go on to the (much more extensive and hard work) second round. You might easily send the most brilliant and useful idea back because it never got looked at.
Open revolt might actually be the result.
-
Tim Groser on the "open source way" for climate change science.
http://www.3news.co.nz/Tim-Groser-17-April-2010/tabid/1356/articleID/151528/Default.aspx
Groser didn't necessarily do the best job at explaining the concept (under some surprisingly hostile questioning) but the idea is this, in order for our science investment to sustain we are, on balance, better off releasing openly, getting recognition and more funding from overseas and helping to ensure that the research reaches a wider audience.
Post your response…
This topic is closed.