Recipe: Distributing papers for a journal club

Problem
You want to distribute papers for a regular journal club in your department.

Solution
Create a group for your journal club in FriendFeed. You can create a either a private group, where only group member can read and post messages, or a public group that is open to everyone. Then invite all regular participants of your journal club to FriendFeed and make them join the group.

Announce the papers that you want to discuss in the journal club via a FriendFeed message. For this go to the webpage for the paper you want to discuss (e.g. this paper) and then use the FriendFeed bookmarklet to announce the paper (and additional information such as the date of the journal club) in the FriendFeed group. If the copyright of the paper allows this, you could also post the fulltext PDF of the paper to the FriendFeed group.

Use FriendFeed comments to capture the discussion about the paper in the journal club. The comments can also contain links to other relevant papers and the slides you may have prepaped for the journal club. This is helpful for those unable to attend the journal club in person, or to look back at the journal club a few months later.

Nature Network forums, CiteULike, Labmeeting and Basecamp (and probably some other tools) offer similar functionality, so use the service you are most comfortable with. Of the tools mentioned, FriendFeed for me is the easiest to set up and use.

Discussion
Email is not a good solution for regularly sending around large files. And discussions among a larger group of people (i.e. all members of a journal club) are difficult to follow via email. Google Wave is a good alternative without these limiations, but is not yet publicly available.

A Wiki or blog could also be used to organize a journal club, but requires a larger effort to set up and maintain.

Many reference managers allow their users to create private groups for sharing references. But in order to work as a tool for a journal club, we also need messages/comments. Not only to discuss the paper, but simply to provide the date and presenter for the journal club or other organisational information. But I wouldn't be surprised if more reference managers besides CiteULike add these features in the future.

If all journal club papers should automatically be stored in a reference manager, use either CiteULike or put the papers for the journal club first into the reference manager and then export them via RSS feed into FriendFeed. This step can be automated if you create a group/folder for the journal club in your reference manager.

Related Posts Plugin for WordPress, Blogger...
This entry was posted in Snippets. Bookmark the permalink.

22 Responses to Recipe: Distributing papers for a journal club

  1. AJ Cann says:

    This solution assumes the target audience will frequent the Friendfeed group regularly and so be aware of what’s being posted – which may or may not happen. There is no external notification strand with this proposal (unless Group members choose to subscribe to the Group’s RSS feed, and of course, check their RSS reader regularly :-)

  2. Martin Fenner says:

    You are right. People that use FriendFeed, Nature Network, Basecamp, etc. infrequently should set up notifications of new messages/comments via email.

  3. AJ Cann says:

    So the first journal club should contain a presentation on “How to set up notifications from Friendfeed” :-)

  4. William Gunn says:

    I have first-hand experience trying to get a group transitioned from emailing PDFs to a more sensible method. I think the first thing you have to be aware of is that even the slightest change will make some people grumpy. The first change I made was from emailing the PDF itself to emailing a link from which it could be downloaded. Believe it or not, many people had problems doing this, ranging from not knowing where their downloaded files get stored to problems with email clients/antivirus programs doing things to the link preventing the link from working.
    Any better approach requires either signing up with a service or installing new software, sometimes both. A fairly easy thing to do is share a Dropbox folder with everyone. All they have to do is look in the folder on their desktop, and new item notifications are on by default. However, his does require installing something, and some people might have institutional restrictions on installing programs or limited write access.
    Probably the best thing to do is get everyone signed up for a bookmarking service. “Mendeley”:http://mendeley.com I’ve invited Martin and you to a example “Journal Club” shared collection on Mendeley to illustrate how it would work. This also requires people to download and install software and sign up, but the advantage is that you can share *annotated* PDFs through the shared collection.
    There are many, many ways to solve the problem of emailing PDFs around to people, but it remains a difficult task.

  5. Martin Fenner says:

    William, thanks for the invitation to the Mendeley “Journal Club” shared collection. You can add tags and notes to the references in the collection, but it isn’t quite the same as the messages/comments mentioned above. But I wouldn’t be surprised if a future version of Mendeley does just that.
    William, the experience you describe sounds very familiar. So I think that we have to walk a very thin line for any new workflows we propose. If they are much better of what we currently do and not too difficult to learn, people might actually use them. That#s why I stressed the comments/messages in the recipe, this makes it much more interesting than a shared folder in a reference manager or a dropbox folder.

  6. William Gunn says:

    Martin, I think we agree that Friendfeed has the best discussion features and requires the least overhead in terms of setting up and using, but you’re still required to handle the actual file distribution some other way, since you can’t attach the file directly to the group.
    Each approach has its advantages and disadvantages. The best solutions, from my experience, place the article on discussion somewhere everyone would be able to get it and also track the discussion around the paper. The synchronized folder approach has some important advantages in terms of usability, but it does require a little more setup.
    Some enhanced features that would take things to the next level would be the ability to share annotations and notes directly within the PDF, so the presenter could highlight items of interest or call attention to figures as the paper was being read. Slideshare provides a different and more labor intensive way of doing something similar. Certainly, Mendeley wouldn’t be appropriate for all situations, because it limits the number of people who can belong to a shared collection, but also because it’s beta software. If you’re trying to make the initial transition from email to something else, perhaps it wouldn’t be the best option at the moment.

  7. Bersenev Alexey says:

    Martin,
    Distribution the paper for journal club online is not a problem. The real problem is critical mass of people for productive discussion. I’ve been thinking about the best way to do it for a while and I realize we have enough good tools online, but we don’t have people who really want to do it. Let’s take one typical academic lab as an example. Let’s say 10 people in the lab (postdocs + grad students).
    If you ask all of them whether discuss 1 particular paper online (your example HIV paper – if lab study HIV is good) with whole world (all professionals around the world) is good thing – everybody will say YES! Ok. Next you tell your labmates – hey guys, I upload paper here, and made a room (group) – so you can easily download, read and discuss in comment section. How many of your labmates will really do that? 1 out of 10? How whole entire scientific world will get to know about it?
    In my building (12 floors, more then 100 labs and thousand of scientists) – nobody knows what FriendFeed is. And labmeeting.com and so on…
    Most of people on FF – bioinformaticians and sci-web geeks, who use a lot of online tools and realize how kool is that discuss about it with people around the world. These kind of people represent 0.5% – 1% of my building – typical academic institution.
    Can you give me an example of good discussion on paper online? On FF? NN? I’m talking about paper from particular scientific field – leukemia, stem cell, HIV, cancer research… but not about general scientific problems such as impact factor metrics or online tools.

  8. Bersenev Alexey says:

    I think blog is still the best way to discuss paper online so far. If I’m blogging about a paper (let’s say bone marrow stem cell niches), I have chance to get 2-5 people to discuss in comments, but on FF it’s usually 0-1. Even I know that there is an army (thousands) of stem cell researchers online every day searching filtered information.
    Another way i like – http://bit.ly/18LQMt

  9. Martin Fenner says:

    Alexey,
    I would think that many labs/research groups have regular journal clubs. Even though most scientists probably don’t use FriendFeed or the other online tools mentioned, they will probably do so if the person organizing the journal club (especially if it is the principal investigator) can convince them to at least try these tools.
    You touch on another important aspect: there is no reason to limit the discussion in a journal club to the people in a particular lab. I agree with you that blog posts are probably the best tool for this online discussion, and “Research Blogging”:http://www.researchblogging.org is a good website to find these discussions. Most blog posts at Research Blogging (e.g. “this one on stem cells”:http://observationsofanerd.blogspot.com/2009/07/live-mice-from-stem-cells.html) currently are probably not written for a journal club audience (where most people know much more about the topic than the average blog reader).
    I would love to write and read more about specific research topics, but as you have experienced with “your blog”:http://hematopoiesis.info/ about stem cell research, there probably isn’t yet a large enough audience. I also have had a hard time to start a regular dicussion in a more general online journal club (the “Good Paper Journal Club”:http://network.nature.com/groups/goodpaper/forum/topics here on Nature Network discusses examples of well-written papers). But maybe we need a little bit more patience and soon will have enough scientists online to discuss interesting papers, at least in some larger research areas such as cancer research.

  10. Bersenev Alexey says:

    Martin,
    I agree with you – PI should take initiative and introduce new tools for his/her students, but is it reality now?
    What is % of academic PI’s (who lead hardcore benchwork lab) among all NN users? What is % of PI’s among “Life Science” room (1000 members) in FF? What is % among them really encourage their students/postdocs to use online sci- collaboration tools?
    Unfortunately you can see opposite situation on US academia (I can’t tell about others because I work in US) – students and postdocs start to use these tools more and more but none of PI’s go online to network. They are too rigid.
    Also if our lab 5 people + PI email is good enough way for us to distribute paper for JC. If we have very good and productive discussion about paper on JC in the lab, many people don’t feel like they should go online and continue discussion with other folks around the world. Scientists scientifically asocial. I don’t know any nerd around me who will go to bar or go home after work and start discuss paper.

  11. Heather Etchevers says:

    Alas, all advice to use FriendFeed must now evolve.
    I hope we can use the tool Google Wave that made you so enthusiastic, to equal effect.
    In my personal situation, journal club generally is reduced to my sending a link or a PDF to a student or a colleague or two, and perhaps our discussing it over coffee or even in group meeting. But that’s as formal as it gets. No one likes to type as much as I do.

  12. Richard P. Grant says:

    _snicker_

  13. Martin Fenner says:

    For those that haven’t seen the news yesterday: Heather is referring to the fact that “Facebook is buying FriendFeed”:http://blog.friendfeed.com/2009/08/friendfeed-accepts-facebook-friend.html. Cameron Neylon wrote a very nice “blog post”:http://blog.openwetware.org/scienceintheopen/2009/08/11/the-trouble-with-business-models-facebook-buys-friendfeed/ of why many scientists (including myself) like FriendFeed and why we are worried that this aquisition means an uncertain future for FriendFeed.
    Although there probably will be no immediate changes to FriendFeed in the near future, we have to see how the services evolves in the coming months. I would very much like Nature Network to pick up some of the nice features of FriendFeed, e.g. the streamlined interfaces, aggregation of other sources or the “Like” button.

  14. Eva Amsen says:

    What do you mean by “streamlined interfaces” in the context of FF vs NN? Things like threaded comments on snapshot page, maybe?

  15. Martin Fenner says:

    Eva, I know that “you are not a big FriendFeed fan”:http://network.nature.com/people/eva/blog/2009/06/26/from-the-vault-the-friendfeed-attitude, but it is much easier to follow several discussions at once in FriendFeed compared to the way Nature Network works. The Snapshot page is really just a starting page and requires a lot of going back and forth.

  16. Eva Amsen says:

    Yeah, but how would that work on here, practically? Do you imagine a screen that displays the blog posts and forum threads started by people in your network? Or something for all NN members? Or a page with all the recent blog posts and their comments and another with forum threads?

  17. Richard P. Grant says:

    MT4.

  18. Martin Fenner says:

    I would leave blog posts as they are now, hopefully soon using Movable Type 4. But forum messages and all the comments from the people in your network could be in one big stream similar to FriendFeed. Forums could work similar to FriendFeed groups, i.e. as a way to organize your stream. And of course I would want to integrate outside information into this stream, e.g. my Twitter posts. Or the papers I publish (automated using some sort of author identifier).
    FriendFeed is a very clever service. But what it lacks are some science-specific features, e.g. clever handling of links to papers in the form of DOIs.

  19. Eva Amsen says:

    Twitter integration is probably problematic NPG would be republishing things that you say on Twitter. In your case, that’s the kind of stuff you say on here anyway, but other people would integrate their Twitter streams and fill it with things like “Shakespearean pants quotes”:http://twitter.com/#search?q=%23shakespeareanpants and that’s probably not the image NN is after.
    But author identifiers to automate papers-by-author in profiles would be good. They already show up in contact’s snapshot pages when you add a paper, so it would just mean that you wouldn’t have to think about adding it to your profile.

  20. Maxine Clarke says:

    The discussion touches on an issue that comes up many times: technology start-up (eg FF, now a subsidiary of FB) vs publisher (eg NNetwork). NN is a free platform for scientists to interact, provided by a publisher (Nature Publishing Group). It would be lovely to have lots of IT and web type functions on N Network, but they all cost a lot of money to develop and then maintain. Personally I am all for them and would love NN to develop in ways mentioned here and in other ways, but a publisher will always be looking at the business case. (This is always what happens to editors when they ask for things for their journal, eg colour images years ago – and all sorts of “nice things” that authors and readers would like and/or are used to in other contexts. But the publisher always asks – why make the investment? This is a neutral question – the publisher (in the case of NPG) is always eager to invest and develop, but wants to know the “why” first.)

  21. Maxine Clarke says:

    PS those Shakesperian pants quotes are *so* annoying, even though it is August.

  22. Martin Fenner says:

    Maxine, you remind us that developing and maintaining these social networking sites costs money, and (as far as I know) nobody has yet figured out how to earn money from these things. And social networking sites, including Nature Network, are probably judged by the people using them and the discussions taking place, and not primarily by the technology.
    FriendFeed is a service that is not targeted primarily at scientists. The advantage is that the service is used by many more people, but it makes it more difficult to add features required only by scientists. I still believe that there is a market for social networking sites for scientists. Like many other people I use Facebook primarily to interact with friends and family and not with other scientists, we’ll see whether this will change when the FriendFeed/Facebook team moves forward.