Archive for the ‘publishing’ Category

Words, words, words…

February 8, 2017 Leave a comment

I have the good fortune of having three PhD theses on my desk at present, two where I am the external examiner and one where I am the internal. They are all on interesting topics, one very related to my own work and the other two a little more obliquely related. While reading theses for examination is a chore – you need to make lots of notes and make sure you have sensible questions to probe the student on, it is also a really good way to concentrate on a specific subject and hopefully learn something along the way.

Having picked up the first one to read my heart sank a little as it became clear that there were going to be lots of rather strangled grammar. If you are an aspiring or current PhD student, you need to know that the quality of your science is what will take you through in the end, but your viva is likely to be a hell of a lot easier if the examiner does not need to stop every paragraph to note how you have deviated from the norms of good English.

My strong advice: throughout your PhD, write as much as you can and get feedback on your work. This can be from your supervisor, lab mates or friends or through courses that I imagine all Universities (certainly in the UK) offer on academic writing. Or maybe by writing a blog. Also try, and this can be hard, to thoroughly proofread your final thesis, and get others to do so too. This will make your examiner’s job so much easier, and so much more enjoyable, allowing them to focus on what you have done. Words your means of delivering your message. Make sure they are your friends.

A different way

Like many physicists, I suspect, I grew up gripped by the developments in quantum mechanics that happened at the start of the 20th century. This is often portrayed as the work of lone geniuses: Einstein, Bohr, Schrodinger, Heisenberg and the rest. That this work was carried out in isolation is to some extent true, but there was a surprising amount of collaboration and certainly discussion between the big hitters of the time. This work, and related studies in areas such as radioactivity, ultimately led to one of the biggest scientific collaborations that had ever existed – the Manhattan Project. This was an altogether different beast: one goal, build a bomb. Many of the brightest minds, engineers, physicists and chemists came together to work out how to achieve what they viewed as something that could help to win the war.

In modern times we have our own parallels of such large scale collaborations, CERN being the most obvious example. These mainly occur because of the huge scale and expense of the projects under consideration. I do often wonder though if we wouldn’t be much better placed to carried out nearly all scientific research through such large ‘crowdsourced’ efforts.

I have a small research group, too small to easily carry out the various ideas that I might have, too small to have the resources to fund all the experiments I’d like to try. It may be that I can persuade a funding council to give me money for these ideas, but the odds are against me. I can then wait and see if we can do them on the fly somehow, or find, depressingly, that someone beaten us to it, a few years after my original thought. I suspect nearly every scientist has similar thoughts about work that just never gets done.

But there are lots of groups out there, lots of talented people, lots of equipment going spare – lots of slack at certain times within any research group, big or small – why don’t I just publicly lists all my ideas and hope someone else runs with it and sees if it’ll work or not? It doesn’t work like that of course. We are precious with our ideas as they define our careers, the funding that we do get, which in turn allows us to build our groups and justify the continued need to employ us. Even collaborations, which are a way to help realise ideas that often we can’t do ourselves can be difficult, time consuming and often not quite what you need if you team up with the wrong group.

This does, I suspect, also have the problem of massively slowing down progress. We all want to win the prize, get the plaudits, get the pay rise, and this stems from doing the work and having your name in the right place on the author list. In this day and age of open access publishing, open data and near instantaneous access to all knowledge it does seem that if the end goal, the experiments, the finding things out is what we want to achieve , that our current way of ‘doing’ science seems increasingly outdated.

Could we do things differently? Would it be possible simply to fund research teams that can then respond to new ideas – take the very best ideas and see them through – have secure funding for staffing and equipment at certain Universities and then let academics the world over provide them with the ideas? This would provide much greater focus and possibly much greater efficiency in how we spend research money. An example would be, say, a centre for optical microscopy in the life sciences, based, for arguments sake at Dundee. We fill it with 100 staff and then throw open to the world the idea to present us with the most pressing problems in the area. It may be that these ideas receive some peer review to set priorities and then we task the centre with solving the problems. The originator of the idea gets appropriate credit, and the centre works collaboratively with the research community to help it make progress. We set up these little ‘Manhattan Projects’ with stability for staff, enhanced training for students, and better opportunities to exploit the research through critical mass. In a sense it centralises the experimental skills and distributes the ideas. It is a model that appears to work for very large scale experimental work, but would it be more efficient than our current massive distribution of experimental skills?

As it happens I am reading J. Craig Venter’s most recent book ‘Life at the speed of light‘ which in a way promotes this idea – a highly skilled, well funded lab pushing for a clear and ambitious research goal. Admittedly he was (and is) in competition with other groups, but if that funding was more concentrated and the initial thinking open and free for wider input and discussion to happen, could things have gone even more quickly? Do we want to see the results and the progress and quickly as we can or keep all the glory for ourselves?

The answer is that I am not sure – the model would seem to work in some cases, but clearly has problems, and would more than likely have to be globally accepted to work in the way I think it could. But with new paradigms appearing in the field of ‘open’ academia very rapidly, maybe there is a different way that we could do science, and actually see more of the collective ideas of the research community come to light and bear fruit.

Media Interactions

This week I have a “News and Views” article published in Nature, which is a discussion of a research paper published in Physical Review Letters outlining an experiment in which a a mirror made from colloidal particles was trapped using laser beams. The idea is that this could be extended to much larger scale devices suitable for making space mirrors, for things like telescopes. Interestingly this article that I wrote about in Nature was then picked up by Physics Todayin a sort of degrees of separation from the original work game. I’m not quite sure what this tells me, other than the media is a bit different from academia – it’s rather faster paced, it scavenges material from where it can, and that I have the utmost respect for media people who can write quality material over a vast range of subjects with rapid turnaround times. I like writing stuff, but quick and accurate is not always my strong point (note the length of time since my last blogpost), but the communication of what we, scientists, do is really important. I’m glad that it is not always down to those at the coalface do that communication.

Note if you have comments on my article you can leave them at the bottom of the Nature page.

Open access publishing: fund it all

April 10, 2012 2 comments

There is growing discontent over the way in which scientific research is disseminated. Stephen Curry has a nice overview and call for action on the Guardian website today. The current system is largely based around the idea of a journal publication being written up by scientists and then sent to a publisher. The paper is then reviewed by some of your peers, who recommend what happens to it next: publish, revise, reject. If accepted then the paper is published. Broadly speaking the publishing model falls into two camps: one, in which the author pays no upfront costs and the publisher sells the article (either through library subscriptions, or on a pay-per-view basis) and two, in which an upfront fee is paid by the authors and the article is then free to read by all. There are variations on these themes, but essentially there is a cost in publishing and appropriately archiving: the question is who pays? The main argument against the closed publisher model is that it restricts access to papers which are taxpayer funded, and that the profits such companies make are at the expense of the taxpayer, Universities and academics who help with reviewing and other editorial duties for free.

I publish in both these types of journals. In optics there is an excellent open access journal called “Optics Express” which launched more than a decade ago and was at the forefront of the open access movement. It has a reasonably impact factor; the only real issue is that it publishes an awful lot of papers, many very good, but often pretty bad. There’s little discrimination. In addition all the papers are sent ‘camera ready’ in that they are published as the author submits them. There is no proof reading or typesetting as you might get with a more traditional publication. A paper in Optics Express will cost between $1000-1750 depending on the number of pages (and if it’s really long you charged per extra page above these fees). I would typically pay for this out of grant funding.

The problem comes when you do work that doesn’t have direct grant funding, or when a paper is published outside the grant funding period. How to pay then if you want Open Access funding? (I should note that often papers not in the Open Access model will still require some smaller fee). To my mind we could get bold about this – one of the main funders of research in the UK is the Research Councils UK, the umbrella organisation for the specific research councils (for Engineering and Physical Sciences, Biological Sciences etc). They could offer to fund all UK published research, including that from the big charities, and make it a condition that all work is published in a simple, immediately free to access for all, format. So how much might this cost?

According to Web of Science (a non-open access citation database) in 2011 there were just over 100,000 papers published by academics in the UK. This is a crude number, based on a search for papers with author addresses in the UK and keeping everything that Web of Science considers a ‘Research Article’ and a ‘Review Article’. If we then assume an average fee of £2000, say, for a journal article to be made open access, this is £200m per year. Which is a lot of PhD students…However it may be possible to arrange for bulk discounting, and possibly start to reduce University journal subscriptions. Money could also be stripped out of grant applications for such costs (I reckon a typical grant will have at least £5000 in pages charges on it. EPSRC funded 604 grants last year, so this would equate to only £3M in pages charges. Spread this across the other councils and it might get this up to £20M, still a long way from £200M, but maybe it could be achieved.

I think we need to be bold here. Centralize the funding and send the invoices to RCUK. Make non-allowable costs of any journal article published in a non-open access way. This has the benefit of giving RCUK full coverage of exactly what papers are published in the UK, and gives it better oversight of the research portfolio – and maybe it could be used as part of a better RAE type review system. Maybe it’s just a little too expensive at the moment, but this, I feel, is how we should be moving.

Note: Obviously there would be issues over publications between UK and foreign collaborators, but solutions could be found. We could also keep the costs lower by funding only research that results from RCUk funded work, but this would then require policing, and up the costs anyway.

Cash for Impact Factors

March 2, 2012 2 comments

If you are looking for an academic job in the UK then should be your first port of call. It lists most open postdoc and academic positions that are available in the UK as well as many academic related jobs and some PhD positions. I get a summary email for physics jobs in my mail box every morning. I’m not looking for a new job as such, but the summary gives you a little snapshot of what new projects are starting up, which departments are expanding, what the trendy areas are at present etc. This monring one of the jobs that caught my eye was for a position at Qatar University. So I followed the link and had a look through what was on offer (I was mainly curious to see what research funding provision they have available in Qatar).

One of the most interesting things that is on offer is a bonus for publishing papers, a “Research Reward Program“. You get a base bonus for a paper with an impact factor of 1.00 and then for every point abaove that you get the base multiplied by the IF, up to a maximum value. The money is shared between authors associated with the University, and authors from outside the University dilute your reward. I had a quick look at exhange rates and I think the maximum you could get is about £1700. The scheme came into effect in 2009; I should probably refrain for calculating my potential bonus to date, had I done all my work there.

So this got me thinking, could this work in our University system? If we set the bar at a different level, could such bonuses be used to try and raise our collective standards? Is there some sort of incentive (other than actually keeping our jobs) that with the REF deadline starting to loom, we could use to focus the mind on getting that precious 4* outputs. The Universities get a ‘bonus’ if we all do really well, so couldn’t they pass this on?

There are of course huge potential problems. Corruption of the system is one (as pointed out on twitter by Stephen Royle (@roylelab)), but I think with appropriate oversight this might not be such a problem. The other is that setting on impact factor favours some over others – your line of work will influence the types of journals you publish in, and you can be really brilliant and relevant but never have a Nature, Cell, PNAS, PRL etc paper.

Academics, I think, like working on their own problems. Often these are niche and small scale. We are poked, prodded and shaped, often cajoled into trying to work bigger: bigger challenges, bigger teams, more disciplines. This comes from funding councils and Universities alike. In reality though I see little to inspire me, other than through self-motivation. It might be nice to see some tangible reward if I really shoot for the moon (and get there).

The other point made on twitter about this was from Pete Watson (@Lardytugboat), who suggested that our bonus was promotion (or maybe a pay rise). This is true up to a point, but I’ve never worked in a University where anyone below a Professor really gets a proper annual pay review. A Professor can put the case that they publish lots of papers and brought in lots of grants, so give them a big pay rise. Below that level you might do a review, but unless you are actively applying for promotion, it’s more of a check to make sure you are on staying on track. And then we move up a point on the pay scale anyway.

Of course, we won’t be getting bonuses. Universities can’t afford it, and I doubt the public would stomach it very readily. Clearly if someone gave me more money for doing good work that would be great, but I do think we need to generate better ways to help people go for bigger gains, less incremental work and improve the quality over quantity ratio.

Sacked for not publishing enough

February 23, 2012 Leave a comment

The Nature News Blog has reported today that around 100 academics from the University of Sydney have been sacked because “they hadn’t published at least four “research outputs” over the past three years’. It seems that the staff had been warned that there was a publication level that they had to meet in order to keep their jobs, but that this was lower that the four within three years that ultimately they were judged on. In some ways this might seem reasonable – that if academics are not publishing, one has to question why they continue in post. This has resonances with the Research Excellence Framework exercise in the UK, where it has almost become untenable to support research that is not deemed sufficiently good enough, or which does not have enough ‘impact’.

The problem is, of course, that reasons for not producing good work are highly variable. It could be that you’re on the brink of some big breakthrough and the focus of this has taken many years to complete. It may be that you’ve had bad luck with papers getting accepted, or have been doing good work, but maybe aiming to high up the journal league table. It could be that you have had a run of poor luck on grant funding, and have not had the support to do your work and write up results, in terms of PhD students, postdocs, or simply a lack of equipment. Maybe a series of experiments didn’t work, but the next one will (and it will win a Nobel Prize). Maybe your research output is highly subjective (think works by artists) and the critics didn’t like your last show. Maybe your research rivals knew about your employment conditions and decided to reject your last paper (to make it to the magic four) to get you sacked. Or maybe, just maybe you didn’t feel the need to publish every last little bit of work to avoid saturating the journals and keep the overall quality of published work high enough to make it bearable to read them. Maybe you published one Nature paper a year over the last three years and nothing else, and they each got 500 citations. Sounds like that might not have been good enough.

These are just a few reasons you might not publish four papers or make four things in three years (any more?). If this report is true and this was the sole condition for dismissal, then shame on the University of Sydney, but maybe it’s a wake up call to the rest of us – coming to a department near you soon. Right, back to writing my next grant next paper.

Science and the press

September 29, 2011 Leave a comment

Some random thoughts. There’s lots on my twitter feed today about an article by Ananyo Bhattacharya (@Ananyo), the chief online editor of Nature, published online in the Guardian, about whether or not scientists should get to proof read articles about their work written by journalists. There seems to be a little bit of a fence being built with scientists on one side saying, “but journalists are a bit rubbish, they don’t understand our hard science, and they play up specific points that they feel make it all more exciting” and journalists on the other saying, “we’re good at writing interesting stories and explaining the relevant science in an informative manner, and we don’t need the direction of the players in the story; in some cases this could raise issues of conflict of interest.”

I’ve been quoted a few times in things like New Scientist (at least online) and in other publications like Chemistry World. I’ve never been asked for a quote for a national newspaper, but I have been close to work that has had a lot of press coverage, and the results are often pretty poor. One image of a 1800s microscope with a laser firing through it, in the Daily Record springs to mind…deary deary me. The issue is that when you are an expert and you allow a non-expert to write about your work, you will (almost by necessity) get something doesn’t look quite right. Imagine trying to explain your next conference talk to a bunch of 15 year olds:  I think this is the challenge that journalists face every day – and they haven’t spent years making the measurements. I think journalists need to do their best to explain, need to be confident that they understand what the hell it is they are talking about, and try their very best not to make ‘cure for cancer’ type claims.

However my real concern with this is that it must extend beyond science to pretty much everything. I am fairly confident that if you run most of the stories in an average newspaper past people who are either close to the story, or are experts in the area of discussion that they would find faults and misrepresentations. Maybe scientists just have more time to sit around and moan about it – but I don;t think it’s a particularly special case in this regard. All journalists can really do is try and report the truth as they see it and give a  balanced view of the issues. Based on the output and speed at which journalists must work, I am generally very impressed by the stories I read.

I have never asked to proof read any story I have been quoted in (although I think in most cases the journalists involved did ask me to look over it – I don’t think I had any need to make changes). I’m happy to keep it this way, and if I’m misrepresented, I guess I’ll just have to take it up after the fact.

Categories: press, publishing, science

Kinect optical tweezers

August 30, 2011 4 comments

Over the summer I have had a project student (BSc Applied Computing student, Matt MacPherson) working on developing a Kinect project to control an optical tweezers experiment. I keep meaning to write something on optical tweezers, but here I want to outline a little bit about this particular project.

In case you don’t know, Microsoft developed the Kinect as a control system for their XBox 360 games console. The idea is that your body becomes the games controller. So by kicking you can kick a football on screen, or by jumping you cause your on-screen character to jump. The system is pretty nice, and contains some simple optics, to give an idea of depth within a scene.

Not long after the Kinect was launched, open access drivers were released allowing anyone to hack together code to interface with the system. Recently Microsoft have made an SDK available. The range of things people have been using the Kinect for is huge and very impressive, from remote control of model aircraft, to a re-creation of the Princess Leia hologram message from Star Wars.

I have always liked to play around with high tech toys, and it seemed that the Kinect offered a really simple and intuitive way to interface with some of our experimental systems, especially our holographic optical tweezers. So I ran this idea as a summer project and we have got some decent results.

The initial ‘simple’ program took input from one hand and turned it into a special kind of hologram, called a kinoform (technically a phase only hologram), which changed as the user’s hand moved. This in turn was relayed through a microscope system to form a moveable optical trap, which can pick up and manipulate microscopic particles.

Example of a kinoform hologram

Intensity image produced using kinoform above.

Matt plays with the Kinect Tweezers

The second phase of work has been a bit more difficult, getting the kinect to control multiple particles. This has mainly been due to gestures clashes – but we have hopefully sorted this out now, and I hope to take the final set of videos we need to show the functionality I originally envisaged next week if I have time…

The plan now is to use the demos for things like open days (we are having a lab open day as part of the Dundee Science Festival) and possibily to develop outreach type projects at Sensation (but we’ll see).

The question now is whether to try and publish this work. I like a good publication and as there is an iPad controlled optical tweezers paper (by my Glaswegian colleagues), there is a precedent. I am torn a little though. The idea is not nearly as swish as people have demonstrated in their bedrooms and in a way doesn’t quite feel like proper science (although I have had some thoughts on ‘manual’ calibration of optical tweezers using the system). However I do feel like that a paper on this topic might generate some press, and this could help my group’s and my department’s profile. The flip side is that it could been seen in a slightly negative light. It’s an interesting question (at least in my head) but I’ll wait until we have the final ‘data’ before making a decision.

If anyone has any thoughts on what we could do with out Kinect tweezers, leave a comment, and I’ll see if we can try it out…