Benefits from Journal Club Style Seminars

Good and Easy

What if I told you that I know about a type of university course that is both good and easy to organize? Good in the sense of both lecturer and students improve their knowledge and easy in the sense that it does not require extensive preparations and organization. Well, this type of course exists an I call it a Journal Club.

Some background first. Students of math and computer science are not well known for their love of reading papers initially. The more if they are of a very practical type as is frequently the case at universities of applied science. As I have pointed out in Flipped Concurrency , flipped classroom approaches can lead them towards better reading of papers and books. And the trend towards more theory and less exercises is strong and continuing. Once students are more familiar with learning from papers and books, the question arises what should happen then? As a lecturer you are used to reading lots of papers and books anyway with blog entries becoming more and more important in computer science. But you know very well, that you get the most from a paper when you have a chance to discuss it with others. In Germany, universities of applied science are organized a bit differently than regular universities with their huge numbers of assistents and PhD candidates. At the computer science and media faculty we are 16 profs and about 16 lab engineers and PhD candidates and we cover a wide range of topics. This means that you are on your own mostly! There aren't four other profs with the same research and teaching areas like yours and yours are anyway much wider than theirs.

So what can you do? You can offer a journal club style seminar. And here is how to do it:

  1. Grab a generic container in your faculties curriculum. If you don't have this type of class ready for guest professors or to try new topics without immediately changing the curriculum (which is now a major task as everything is IT based (:-)) your should talk to your dean immediately. Generic class containers allow flexibility and experiments. Call it "current topics on [whatever]". Two ECTS are enough for a journal club seminar except if you expect a larger scientific paper at the end. If your are lucky, a PhD candidate or lab engineer in your faculty is interested in running the club with you. (In my case I have to say thanks to Stephan Soller for helping me in many ways)

  2. At the beginning of a new term announce that you will be offering a journal club seminar. You might wait a little so most students already picked their classes to avoid too many applications. Absolutely limit the number of participants to eight or ten. Eight is ideal. Everything larger will bring down the intensity of the course and allow students to hide behind others.

  3. Explain the journal club style and make it absolutely clear, that everybody will be required to read the same papers and that this will cost you an evening bevore the next session. Reading the papers at the evening bevore makes sure that you know about the content the next day.

  4. Explain the grading by pointing out that active participation is absolutely required to pass the course. And that at the end a blog entry in our Computer Science Blog at HdM is expected.

  5. In the first session all participants suggest papers to read and a list of papers is put in the owncloud repository. (You should encourage students to use the journal club to improve on a scientific paper they are currently working on)

  6. In every session questions, statements, critique etc. are collected at a whiteboard and discussed. Every participant should add something. One after the other the topics are discussed in the group and hopefully clarified.

  7. Fifteen minutes before the end of the session discussion stops and the participants select a new paper for the next session.

  8. At the end of the term, all participants create public blog entries about one of the topics handled in the papers.

Lessions learned

Let's start with what a lecturer learns from a jounal club. Well, it's fun and low overhead and really helps in understanding complicated things much better than without the discussion in a group. It is amazing what even an experienced lecturer will miss in a deep and compressed paper. Another thing that a lecturer realizes is the amount of know-how that is available in a group of 8-10 master students. And the most gratifying thing to learn is to notice how the students grow and change in such a club. The blog entries - possibly because they are public - turn out to be very good usually and both students and the faculty benefit from the PR gained by those publications.

The lessons learned by students start with recognizing that reading papers really takes time and effort. Even chosing papers is an important activity which leads to reasoning about ones goals. And then they realize, that everybody in the group has only a partial understanding of the paper which improves much during the discussion. Adding your own ideas turns you into an equal between equals and you relize that it is the things you know that make the difference to before. Suddenly students realize that they are equals, capable and that reading papers gives them an advantage that it recognized by the group. A true academic feeling shows up in the group and students, when they meet greet each other or the lecturer with the words: "did you see the new paper on xxxx by yyyy?

It does not happen with every student but I have seen it happening often enough to be convinced about the positive effect on self-confidence and paper-reading attitude. And I think that is a lot you gain for so little effort, not to mention the fun you can have in those discussions. So if you haven't tried a journal club, just follow the steps mentioned above and you should be OK.

If your are coming from a regular university and perhaps from a liberal arts faculty you will probably wonder what is new about a journal club. There, journal club like courses happen frequently and are called "Oberseminar" etc. and are sometimes on invitation only. But I believe that journal clubs are not so frequent in universities of applied sciences. Something that should change I believe.