Every department seems to have a different view on the purpose of journal club.
In one view, it's training in giving presentations,and the journal club may
be the only opportunity students have to give a seminar-style talk. In
another view, it's training in critical evaluation of papers. Some departments
try to cover both bases.
In my department, the emphasis is mostly on critical evaluation, but also on
organization of ideas. This year a student and post-doc came up with a scheme
that sound more complicated than it is, and has really worked for us. It does
require organization and forethought.
Essentially, whoever is 'it' picks a main paper. While reading that paper,
that person picks 4 key references. These additional papers are usually chosen
to give some background to the thought processes in the paper, or to explain
methods that may be unfamilliar. Then the cycle begins.
Week 0: Distribute main paper and background papers. This is done during
journal club. Each person takes only one background paper, and there are
always at least two people per paper. Week 0 is also Week 3 of the previous
Week 1: The person who is 'it' gives a short introduction. This is often
where terms are defined, or systems explained. For example, no one in our
department does C. elegans work, and a quick review of pertinent anatomy can be
helpful. Next, brief presentations of the background papers are given by the
readers. Ideally, you've skimmed the main paper and know what focus to take.
The order of presentation is chosen usually by the coordinator for that cycle
(whoever is 'it') to make the most logical sense.
Week 2: Discuss the main paper. Distribute papers for next cycle.
The co-ordinator for each cycle ends up doing a fair amount of reading,
but when it comes to the main paper, all participants are up to speed. It
really helps to stay in a system/subject for more than one week -- you get used
to thinking about an unfamilliar topic. Conversely, the timeframe is not so
long that it gets boring.
Given this scenario, let me address your questions.
In article <misandst.90.310105EB at postbox.acs.ohio-state.edu>, misandst at postbox.acs.ohio-state.edu (misandst at magnus.acs.ohio-state.edu) writes:
> 1. In order to get discussion moving so that many people can
> participate, is it acceptable to engage in what (to the experts) may seem
> like wild speculation?
During my qualifying exam, I made an overhead that said "Wild Speculation". I
used this to preface remarks and answers which I knew were tainted by my lack
of experience. It worked. Anyway, this depends on your colleagues and your own
level of self confidence. I've worked in labs where you couldn't speculate
over a beer without being asked for evidence to back up your statements.
However, meeting oposition with an (undefensive), "Hey, I'm just a student!
Teach me something. Tell my why I'm wrong" can be helpful.
> 2. If a paper has been chosen and the authors are less rigorous about
> defining their hypothesis than the standards of faculty present would
> usually allow, is it still possible to discuss interpretation of data in a
> meaningful way?
Sure. Discuss what the authors intended to do, and then if it's not 'rigorous
enough', discus how they should have done it.
> 3. If the terms used in a paper, which have been well defined in the
> literature prior to this paper, begin to be used with definitions which veer
> away from the strict operational parameters to which they were once
> associated, is it better to discuss/teach the former literature which
> these authors seem to discount or to question the previous definitions?
If your journal club is for seminar practice, yes. You can generally do this
in about 5 minutes. Everybody loves a good controversy. Besides, any critical
analysis of a paper is aided by knowing what has gone on before. These
specific issues are part of what our Week 1 is used for.
Credit where it's due. The scheme described above was the brainchild of Dr.
Loren Haarsma and Tarik Alkasab.
Hope this was helpful.
M. S. AtKisson
Tufts University School of medicine