Developing Statistical Concepts for Engineering Students Using Computer Packages

The idea of using the computer as a teaching tool is not new. It has been used as an effective aid in mathematics lectures for a number of years (see, for example, the report produced by the Shell Centre for Mathematical Education, 1986). Hunka (1988) reports on a computer-assisted instructional course which has been used for fifteen years at the University of Manitoba for a first course in statistics for graduate students in educational psychology. While such a system has the advantage of allowing students with diverse backgrounds to progress through the course at their own pace, it is difficult to update and take advantage of modem developments such as new techniques and easy to use statistical packages. Another approach is described by Harrington (1988) who discusses the dilemma which statistics educators face in deciding where the balance between hand and computer calculations should lie in statistics courses. She advdcates the use of a spreadsheet in which the students can enter data using "macros" as a means of achieving this balance. They can also investigate "what if" questions by altering data values and observing the effect on the statistics being studied. Like the calculator, many mathematics and statistics educators see the computer basically as a computational tool. Joiner (1982) stated that the goal of a basic statistics course should be to help students understand as much as possible of the statistical analysis process. He listed a number of steps which are involved in the statistical process and explained how the use of a computer can help free the student's mind from the need to worry about calculation details and concentrate on the process. In recent years the computer has opened the way to readily handling large amounts of real-life data, giving students the opportunity to investigate many situations for which the calculations would have previously been too tedious. Although the computer has proved most useful in facilitating virtually any form of statistical analyses, it has not necessarily greatly improved the understanding of