5  Other suggested methods of teaching statistics

This final section discusses some more issues dealing with statistics lectures - larger statistical projects, and reading statistical articles. I try to answer the question whether these teaching methods are suitable for a course like `EMa2 - Applied Statistics'.

5.1  Statistical projects

Work on larger statistical projects is sometimes recommended as a good way how to introduce `real-life' statistics in the lectures. Students are forced to use the methods they are taught on real data, and should attempt to proceed in the way statisticians do.

This thought might be interesting but has got severe disadvantages. Since the unit element `EMa2 - Applied Statistics' is limited by the number of lectures, introducing a project would necessarily lead to cuts in the amount of subject matter introduced in the course. Experience also shows that students working on such a project spend most of the time with data collection and filtering. Although this is a necessary part of practical statistics, this unit should teach `applied' statistics which does not equal `practical' statistics.

The idea is that students should first gain theoretical knowledge about statistical methods, and practise the knowledge intensively on simpler examples. Since this is an introductory course into statistics, it is not desirable to turn it into a practical course, unlike what is suggested in [15].

Although some arguments for the introduction of statistical projects in the courses include the fact that students should know real statistical work before entering a job (see e.g. [11, p.12]), or that students might be more interested in statistics if they feel its practical impact, I think that it might be rather contra-productive. They are very time-consuming, and dragging away from the original purpose of the course - to teach the students a whole range of statistical principals and methods which they might use in their future careers. Practical statistics is still better learnt by working experience.

5.2  Reading of statistical articles

Another idea is to develop understanding of statistics by letting students read and process statistical articles. This should introduce students to the statistical language and to the ways how statistical methods are implemented in practice, without the necessity of spending unnecessarily much time with collection and processing of data. Experience with technical articles is also valuable.

There are of course difficulties arising. Firstly, lecturer has to find articles that would be suitable for the students. Statistical journals dealing with applied statistics contain a whole range of diverse articles, some of which are definitely suitable for such a course. There are also collections of articles published as books - [16] is an example of such a collection of mathematically not very difficult articles, although these might be already slightly outdated. Finally, the Internet is an inexhaustible source, the only problems is to find suitable articles in the immense amount of data. The webpage created for the purpose of this project contains links to such articles.

In order to achieve the desired objective, there should be a way how to assess students' work with the article. This can be done in several ways - from writing a report, over presenting the article e.g. in example classes, or making an article examinable in the final exam. We have to remember that all these ways increase the marking load of the lecturer. Nevertheless, I consider scientific article reading as a good way how to familiarize oneself with the topic.