Sunday, February 19, 2012

Deeper Understanding Through Student Discussion

In "Peer Instruction: A Teaching Method to Foster Deep Understanding" (Communications of the ACM, February 2012), Beth Simon and Quintin Cutts suggest that computing educators should use discussion techniques already used for physics. Curiously, given the digital revolution the products of computer science has unleashed on society, computer students at university are still largely taught in old fashioned lecture theaters. Simon and Cutts suggest using a technique of "Peer Instruction" (PI), where students discuss topics in groups. This is used for physics classes in the USA, but is not that different to the way many secondary school classes are conducted in Australia.

Simon and Cutts describe peer instruction as a four step process, which is designed around a set question for the student to consider. The instruction ideally takes place in a specially designed "cabaret" style room with tables for groups of students and provision for short lectures by the tutor:
  1. Each student records their answer with an electronic "clicker" device.
  2. Students discuss the question at each table.
  3. Students can then revise their answer.
  4. Selected students explain their answer to the entire class, with the tutor moderating the discussion.
The session might be divided up into topics with the tutor presenting a mini lecture between each.

PI was developed by Eric Mazur for teaching physics at Harvard University.

Part of the problem in introducing such techniques may be that university lecturers receive little formal training in how to teach. Early career academics, who are trained in new techniques, may be hesitant to propose them to older colleagues, especially as education is seen a secondary role for academics, after research.

Also the educational techniques are built into the terminology and renumeration structures of the academic profession. A university "lecturer" is expected to "lecture" and their workload is measured by how many "lectures" they give. The workload of a student is similarly calculated by the number of "lectures" they are required to attend. The quality of the lectures is not measured (and research shows has little to do with student results).

One way out of this may be by making new teaching techniques not sound too radical. Students already attend "tutorials" and "workshops". These can be remodeled to the new format. The "lectures" can be remodeled to be audio visual presentations.

No comments: