I went to Australia in search of the fully integrated teaching of statistics and research methods as part of the PBL tutorials. I did not find it anywhere. I did find three different models:
In addition, one school (New South Wales) did not use problem-based learning and one (James Cook) taught virtually no statistics. More information about the individual medical schools is available [see Additional file 1].
Material integrated but separately taught
This approach was used or planned at the University of Sydney, the University of Melbourne, and the Australian National University. However, of these only the University of Sydney had actually put this into practice. The University of Melbourne and the Australian National University were about to implement what was essentially the Sydney model.
In this model, statistics and research methods are taught by PBL and the PBL triggers are integrated with the PBL problems for other parts of the course, but the material is not taught in the same tutorials or by the same tutors as anatomy, biochemistry and physiology. There will be a separate set of triggers for the statistics, etc., presented in a separate tutorial, and by separate tutors.
I asked why the main PBL tutors could not do this. As I understood it, the function of a PBL tutor is to facilitate and guide the group, not to impart knowledge. I had no problem, at least that I was aware of, in acting as PBL tutor when students were working with triggers designed to elicit questions about anatomy, biochemistry, and physiology, subjects of which I know virtually nothing. Besides, many of these tutors must routinely read journals which bristle with P values, t tests, correlation coefficients, etc. They must be familiar with the terms, if nothing else. Answers to this included:
-
the tutors themselves refused to do it,
-
in the early years of the course we need expert tutors,
-
many tutors are still rooted in the old paradigm and are reluctant to embrace EBM and it was the view of my informants that experience round the world shows that it is difficult to teach EBM principles.
The team at the University of Sydney were on the whole positive about their course.
A parallel course
This approach was used at Monash University, the University of Queensland, Flinders University, the University of Newcastle, and the University of Adelaide.
In this approach, a non-PBL course is given separately from the main PBL course. This may consist of any combination of lectures, seminars, practicals, web pages, or text handouts. Usually there is an attempt to link this to the PBL cases by using examples related to the case of the week. For example, the case of the week might be asthma and the parallel course could include a critique of a paper reporting a trial of a treatment for asthma.
There are several problems with this approach to teaching statistics and research methods. As noted in the Background, the subject may seem peripheral to the main thrust of the medicine course. Student feedback tends to give a much lower approval to parallel courses than to the main PBL teaching. Finally, teaching is dependent on the cases chosen by the PBL teachers, who may change the cases or reorder them at little or no notice. This can make statistics teaching, which is much more dependent on the order of presentation than most subjects in the medical curriculum, extremely difficult.
Most people involved in these courses were unhappy with them, the exception being a group project in the Flinders course.
PBL used only for 'public health' related subjects
This approach was used at the University of Western Australia.
This was an unusual model, found at only one university. PBL teaching had been initiated by an enthusiast, after a period spent at McMaster University. She was a member of the public health group and persuaded her colleagues to introduce PBL. However, only about one third of the course is taught this way, anatomy, biochemistry and physiology are taught traditionally. The consequence of this approach is that the tutors are drawn from the population medicine area and so are quite happy to teach statistics, research methods, and EBM. The triggers can be chosen as population-oriented problems, rather than being restricted to the patient case.
People I spoke to were very positive about this course, not surprisingly as they saw themselves as the educational leaders in their institution.