Home » Blog

Open Science and the developing world: Good intentions, bad implementation?

14 January 2008 18 Comments

I spent last week in Cuba. I was there on holiday but my wife (who is a chemistry academic) was on a work trip to visit collaborators. This meant I had the opportunity to talk to a range of scientists and to see the conditions they work under. One of the strong arguments for Open Science (literature access, data, methods, notebooks) is that it provides access to scientists in less priviledged countries to both peer reviewed research as well as to the details of methodology that can enable them to carry out their science. I was therefore interested to see both what was available to them and whether they viewed our efforts in this area as useful or helpful. I want to emphasise that these people were doing good science in difficult circumstances by playing to their strengths and focussing on achievable goals. This is not second rate science, just science that is limited by access to facilities, reagents, and information.

Access to the literature

There is essentially no access to the subscriber-only literature.  Odd copies of journal issues are highly valued and many people get by by having visiting positions at institutes in the developed world. I talked to a few people about our protein ligation work and they were immensely grateful that this was published in an open access journal. However they were uncertain about publishing in open access journals due to the perceived costs.  While it is likely that they could get such costs waived I believe there is an issue of pride here in not wishing to take ‘charity’. Indeed, in the case of Cuba it may be illegal for US based open access publishers to provide such assistance. It would be interesting to know whether this is the case.

Overall though, it is clear that acccess to the peer reviewed literature is a serious problem for these people.  Open Access publishing provides a partial solution to this problem. I think to be effective it is important that this not be limited to self archving, as for reasons I will come back to, it is difficult for them to find such self archived papers. It is clear that mandating archival on a free access repository can help.

Access to primary data

Of more immediate interest to me was whether people with limited access to the literature saw value in having free access to the primary data in open notebooks. Again, people were grateful for the provision of access to information as this has the potential to make their life easier. When you have limited resources it is important to make sure that things work and that they produce publishable results. Getting details information on methodology of interest is therefore very valuable. Often the data that we take for granted is not available (fluorescence spectra, NMR, mass spectrometry) but details like melting points, colours, retention times can be very valuable.

There were two major concerns; one is a concern we regularly see, that of information overload. I think this is less of a concern as long as search engines make it possible to find information that is of interest. Work needs to be done on this but I think it is clear that some sort of cross between Google Scholar and Amazon’s recommendation system/Delicious etc. (original concept suggested by Neil Saunders) can deal with this.  The other concern, relating to them adopting  such approaches, was one that we have seen over and over again, that of ‘getting scooped’. Here though the context is subtley different and there is a measure of first world-developing world politics thrown in. These scientists are, understandably, very reluctant to publicise initial results because the way they work is methodical and slow. Very often the key piece of data required to make up a paper can only be obtained on apparatus that is not available in house or requires lengthy negotiations with potential overseas collaborators. By comparison it would often be trivially easy for a developed world laboratory to take the initial results and turn out the paper.

The usual flip side argument holds here; by placing an initial result in the public domain it may be easier for them to find a collaborator who can finish of the paper but I can understand their perspective. These are people struggling against enormous odds to stake out a place for themselves in the scientific community. The first world does not exactly have an outstanding record on acknowledging or even valuing work in developing countries so I can appreciate a degree of scepticism on their part. I hope that this may be overcome eventually but given that the assumption of most people in my own community is that by being open we are bound to be shafted I suspect we need to get our own house in order first.

The catch…

All of this is well and good. There are many real and potential benefits for scientists in the developing world if we move to more open styles of science communication. This is great, and I think it is a good argument for more openness. However there is a serious problem with the way we present this information and our reliance on modern web tools to do it. Its a very simple problem: bandwidth.

All of our blogs, our data, and indeed the open access literature is very graphics heavy. I actually tried to load up the front page of openwetware.org while sitting at the computer of the head of the department my wife was visiting (the department has two networked computers). Fifteen minutes later it was still loading.  The PLoS One front page was similarly sluggish. I get irritated if my download speeds drop below 500K/second, at home, and I will give up if they go down to 100K. We were seeing download rates of 44 bytes/second at the worst point. In some cases this can even make search engines unuseable making it near impossible to track down the self-archived versions of papers. Cuba is perhaps a special case because the US embargo means they have no access to the main transatlantic and North American cables, in effect the whole country is on a couple of bundles of phone lines, but I suspect that even while access is becoming more pervasive the penetration of reasonable levels of bandwidth is limited in the developing world.

The point of this is that access is about more than just putting stuff up, it is also about making it accessible. If we are serious about providing access, and expanding our networks to include scientists who do not have the advantages that we have, then this necessarily includes thinking about low bandwidth versions of the pages that provide information. I looked through PLoS One, openwetware, BioMedCentral, and couldn’t find a ‘text only version’ button on any of them (to be fair there isn’t one on our lab blog either).  I appreciate the need to present things in an appealling and useful format, and indeed the need to place advertising to diversify revenue streams. I guess the main point is not to assume that by making it available, that you are necessarily making it accessible. If universal accessibility is an important goal then some thought needs to go into alternative presentations.

Overall I think there are real benefits for these scientists when we make things available. The challenges shouldn’t put us off doing it but perhaps it is advisable to bear in mind the old saw; If you want to help people, make sure you find out what they need first.


  • That may be one advantage of using Wikispaces for a lab notebook. It is pretty much bare-bones: most pages have no embedded images, just explanatory text and links to the raw data.

  • That may be one advantage of using Wikispaces for a lab notebook. It is pretty much bare-bones: most pages have no embedded images, just explanatory text and links to the raw data.

  • Interestingly I checked some page sizes (probably should have done that before)

    UsefulChem front page is 91.1 kB
    PLoS ONE is 43.5 kB
    BioMedCentral 49 kB
    OpenWetWare 22.2 kB

    None is particularly big, but at the end of a long pipe each of those pictures can end up causing trouble. For instance half of the PLoS ONE front page is the nice grey shading frame that the articles sit in. I couldn’t figure out why UsefulChem was so big but I guess it shows that its not immediately obvious how much space these things take up. Slightly embarrasing is that our Lab Blog pages come up as ‘unknown’ size which seems to be something to do with the gel images (which are pretty big).

  • Interestingly I checked some page sizes (probably should have done that before)

    UsefulChem front page is 91.1 kB
    PLoS ONE is 43.5 kB
    BioMedCentral 49 kB
    OpenWetWare 22.2 kB

    None is particularly big, but at the end of a long pipe each of those pictures can end up causing trouble. For instance half of the PLoS ONE front page is the nice grey shading frame that the articles sit in. I couldn’t figure out why UsefulChem was so big but I guess it shows that its not immediately obvious how much space these things take up. Slightly embarrasing is that our Lab Blog pages come up as ‘unknown’ size which seems to be something to do with the gel images (which are pretty big).

  • All the images, scripts, etc. on the OpenWetWare homepage total 229KB! So I think we can see where the bandwidth all goes!

    Firefox has a useful feature that allows you to change stylesheet. You can select “No Style”, which means it in effect shows a plain text version of the page. As long as the page uses correct HTML (and I mean semantically correct, not just that it validates) the site should still be usable. I’m viewing your blog like that now, and it’s fine (I would probably put the navigation after the content, but otherwise it’s OK).

  • All the images, scripts, etc. on the OpenWetWare homepage total 229KB! So I think we can see where the bandwidth all goes!

    Firefox has a useful feature that allows you to change stylesheet. You can select “No Style”, which means it in effect shows a plain text version of the page. As long as the page uses correct HTML (and I mean semantically correct, not just that it validates) the site should still be usable. I’m viewing your blog like that now, and it’s fine (I would probably put the navigation after the content, but otherwise it’s OK).

  • Sharon

    I think this insightful posting offers an opportunity to bring up another issue to consider with ONS accessibility – accessibility right here at home (in the developed world, that is). My progress towards creating my own Open Notebook is coming along slowly (though thank you for all the information posted previously), so I have not had the opportunity to evaluate existing platforms for usability and accessibility. But based on today’s posting and comments, it seems that there might be room for improvement. In addition to standard usability metrics (e.g., http://www.useit.com), which make for a better web experience for everyone, especially those with slow internet connections, there are also accessibility guidelines that expand usability approaches to include people with disabilities, such as those with vision and dexterity limitations. The World Wide Web Consortium (W3C), an international body that sets web standards, is a good resource for accessibility guidelines (http://www.w3.org/WAI/gettingstarted/Overview.html). It is my hope that open science and accessible lab notebooks will also make it easier for persons with disabilities to become active, contributing members of our scientific community.

  • Sharon

    I think this insightful posting offers an opportunity to bring up another issue to consider with ONS accessibility – accessibility right here at home (in the developed world, that is). My progress towards creating my own Open Notebook is coming along slowly (though thank you for all the information posted previously), so I have not had the opportunity to evaluate existing platforms for usability and accessibility. But based on today’s posting and comments, it seems that there might be room for improvement. In addition to standard usability metrics (e.g., http://www.useit.com), which make for a better web experience for everyone, especially those with slow internet connections, there are also accessibility guidelines that expand usability approaches to include people with disabilities, such as those with vision and dexterity limitations. The World Wide Web Consortium (W3C), an international body that sets web standards, is a good resource for accessibility guidelines (http://www.w3.org/WAI/gettingstarted/Overview.html). It is my hope that open science and accessible lab notebooks will also make it easier for persons with disabilities to become active, contributing members of our scientific community.

  • Sharon, that’s a good point. I am pretty lazy I know at giving alternative tags to links (haven’t actually put any pictures up here yet) and I haven’t even got as far as thinking about how these things would look to people with various disabilities. The problem of course is that in the rush to get anything up we, or at least I, don’t take the time to ‘clean up’. If you watch my posts in detail over time you’ll realise I end up proof reading them properly only once they are posted. But I agree we should be working to incorporate features in our systems that make accessibilty to everyone as high as possible. Feel free to nag about any specific points which I need to clean up.

  • Sharon, that’s a good point. I am pretty lazy I know at giving alternative tags to links (haven’t actually put any pictures up here yet) and I haven’t even got as far as thinking about how these things would look to people with various disabilities. The problem of course is that in the rush to get anything up we, or at least I, don’t take the time to ‘clean up’. If you watch my posts in detail over time you’ll realise I end up proof reading them properly only once they are posted. But I agree we should be working to incorporate features in our systems that make accessibilty to everyone as high as possible. Feel free to nag about any specific points which I need to clean up.

  • M.C.Arunan

    For effective implementation of open science in developing countries lack of access to broad band is only one of the problems. The other is inaccessibility to internet itself. I have in my department four computers, but none connected to the internet. The institution provides two terminals with internet connections to the faculty numbering over a hundred. Mind you, mine is one of the reputed colleges in Mumbai, the richest metroplis of India.

    Some of us are spearheading a movement asking the university to make it mandatory to provide a certain minimum number of terminals with internet connections, now that open access journals are vailable at the click of a mouse.

    Thank you for highlighting that making available is different from making the same accessible. Third world citizens always new this. For, millions of people in India starve not because we do not have food, but because the grain which is rotting in godowns are not made accessible to them.
    M.C.Arunan

  • M.C.Arunan

    For effective implementation of open science in developing countries lack of access to broad band is only one of the problems. The other is inaccessibility to internet itself. I have in my department four computers, but none connected to the internet. The institution provides two terminals with internet connections to the faculty numbering over a hundred. Mind you, mine is one of the reputed colleges in Mumbai, the richest metroplis of India.

    Some of us are spearheading a movement asking the university to make it mandatory to provide a certain minimum number of terminals with internet connections, now that open access journals are vailable at the click of a mouse.

    Thank you for highlighting that making available is different from making the same accessible. Third world citizens always new this. For, millions of people in India starve not because we do not have food, but because the grain which is rotting in godowns are not made accessible to them.
    M.C.Arunan

  • Jean-Claude Guédon

    The issue of bandwidth has to do with electronic publishing, not open access. Open access is necessarily affected by bandwidth because open access relies on the non-rivalrous nature of bits to exist. Any electronic publication, toll-gated or not, will suffer from the same problem. However, anyone familiar with Third World conditions will not entrust diffusion of materials to paper and local post offices either.

    Bandwidth is indeed an issue, but it is a general one and this is why it has to be solved in general for the Third World. There, the issues are related to the commercial strategies of major carriers and these are sometimes tied with the geo-strategies of major industrialized nations.

    But, I repeat, the bandwidth issue is not specific to open access. As all of scientific publishing is being digitized and made accessible through the net, the same problem will occur, be it Elsevier, Springer or Wiley.

    Finally, Open Access repositories would have an interesting project in transforming their files into formats that would be frendllier to countries or groups with limited bandwidth.

  • Jean-Claude Guédon

    The issue of bandwidth has to do with electronic publishing, not open access. Open access is necessarily affected by bandwidth because open access relies on the non-rivalrous nature of bits to exist. Any electronic publication, toll-gated or not, will suffer from the same problem. However, anyone familiar with Third World conditions will not entrust diffusion of materials to paper and local post offices either.

    Bandwidth is indeed an issue, but it is a general one and this is why it has to be solved in general for the Third World. There, the issues are related to the commercial strategies of major carriers and these are sometimes tied with the geo-strategies of major industrialized nations.

    But, I repeat, the bandwidth issue is not specific to open access. As all of scientific publishing is being digitized and made accessible through the net, the same problem will occur, be it Elsevier, Springer or Wiley.

    Finally, Open Access repositories would have an interesting project in transforming their files into formats that would be frendllier to countries or groups with limited bandwidth.

  • @Arunan, Yes, bandwidth definitely includes the issue of having to walk to the computer that has some. I think your movement sounds like a positive step. Are there things we can do to help?

    @Jean-Claude Guedon. I take your point at a technical level and agree that it is definitely a much more general issue tham just Open Access publishing but from my perspective as a scientist who would like to make their science available to the developing world I essentially have to reject the subscriber access publisher as totally irrelevant before even starting.

    My point was more that, if you are motivated to make your science accessible to people in the developing world, which means by default you will choose an open access publisher, you have to do more than just make that choice and then polish your halo. You have to think about whether the journal is really accessible to your target audience or not. Or whether perhaps you should re-publish the article in a small size format etc etc. There was a lot of interesting discussion on this in a session of the ScienceBlogging meeting in North Carolina. Some links can be found here.

  • @Arunan, Yes, bandwidth definitely includes the issue of having to walk to the computer that has some. I think your movement sounds like a positive step. Are there things we can do to help?

    @Jean-Claude Guedon. I take your point at a technical level and agree that it is definitely a much more general issue tham just Open Access publishing but from my perspective as a scientist who would like to make their science available to the developing world I essentially have to reject the subscriber access publisher as totally irrelevant before even starting.

    My point was more that, if you are motivated to make your science accessible to people in the developing world, which means by default you will choose an open access publisher, you have to do more than just make that choice and then polish your halo. You have to think about whether the journal is really accessible to your target audience or not. Or whether perhaps you should re-publish the article in a small size format etc etc. There was a lot of interesting discussion on this in a session of the ScienceBlogging meeting in North Carolina. Some links can be found here.

  • Open Science is that it provides access to scientists in less privileged countries to both peer reviewed research as well as to the details of methodology that can enable them to carry out their science. There is essentially no access to the subscriber-only literature. Odd copies of journal issues are highly valued and many people get by by having visiting positions at institutes in the developed world. Of more immediate interest to me was whether people with limited access to the literature saw value in having free access to the primary data in open notebooks.so really its nice article. i like it so much.thanks for the informative article.i appreciate to your efforts.

  • Open Science is that it provides access to scientists in less privileged countries to both peer reviewed research as well as to the details of methodology that can enable them to carry out their science. There is essentially no access to the subscriber-only literature. Odd copies of journal issues are highly valued and many people get by by having visiting positions at institutes in the developed world. Of more immediate interest to me was whether people with limited access to the literature saw value in having free access to the primary data in open notebooks.so really its nice article. i like it so much.thanks for the informative article.i appreciate to your efforts.