Three aspects to be considered when teaching a one-semester beginning economics statistics course are coverage, mastery, and applications. There is a difference between coverage and mastery. Moreover, mastery is not an end in itself; instructors must consider how statistics courses will influence students' approaches to other subjects and applications. In principle, computer activities can be designed and implemented to improve any and all of these three goals. The HyperCard software for the Macintosh computer should result in an important advance in the interface between computer and user. This will be valuable for tutorial programs. Cognitive scientists are designing software which analyzes student solutions to standard problems by inferring a student's intentions from the details of her/his solution and then offering diagnostic assistance. Programs like "Stat Helper" (briefly discussed) for the Macintosh allow students to interact with the computer in solving a variety of problems. Students can learn about regressions better through hands-on experience on personal or mainframe computers. Computer experiments can exhibit a variety of points about regression applications. Computers can expand coverage and make applications more accessible to the average student. Students must develop some sense about what questions regressions can and cannot be expected to answer. Four examples, including graphs and statistical data, are given: automobile weight and fuel mileage, polynomial (quadratic), omitted independent variable, and logarithmic relationship. Nine references and numerous tables and graphs are provided. (GEA)
The CAUSE Research Group is supported in part by a member initiative grant from the American Statistical Association’s Section on Statistics and Data Science Education