Practical Projects in the Teaching of Statistics at Universities

Since this is a very personal view on the teaching of statistics, I shall unashamedly use the first person. I first learnt statistics at one of our ancient universities. The theoretical lectures varied from the stimulating and exciting to the frankly incomprehensible. These were supplemented by weekly practical classes supervised by a very young demonstrator (now a distinguished, if controversial, professor of statistics). We were handed out a duplicated sheet of data, and given step by step instructions on how to turn these into an analysis of variance, correlation coefficient or whatever was the topic for the week. Being reasonably dextrous, I was able to turn the handle of the Brunsviga provided in a minimum of time to produce the "right" answers, consequently obtaining good reports on my ability for "practical statistics". Apart from the skill in turning a handle quickly and learning a few computational tricks, these classes did little to increase my knowledge of applied statistics.