Brian M post for 28 Feb

Brian Melton

My article this week is titled “VISA:  Reducing Technological Impact n Student Learning in an Introductory Statistics Course” and is found at http://escholarship.org/uc/item/1gh2x5v5.  Although not published in my usual journal choice, Journal of Statistics Education (JSE), it was found via a link in the JSE.  In choosing this article, I was considering the question, “How can technology be used effectively to teach introductory statistics?”  This would help guide me in the future in developing a course; whether I wanted the students to use more hands-on learning through the personal computer, or instead focus on concepts which are then reinforced through a statistical software package.

In this study the statistical software package used was Visual Interactive Statistical Analysis (VISA), an Excel-based analysis package designed to support student learning though a user friendly application.  The researchers compared the use of VISA through a computer lab environment, where each student learns how to use the tool on their own computer, to a classroom where the instructor only has the use of the program to demonstrate while using it as a tool for teaching the concepts.  The study was two fold.  The first was to see if there was a difference in student performance between the two groups.  The second was to argue that an introductory statistic class should be geared towards learning how to do decision-making using statistics, while downplaying the details of specific statistical tests.

The researchers discuss their literature review, provide a description of how the VISA tool is used, and conduct their experiment on the two learning groups.  The result of the study showed there was no difference in the level of learning between the two methods, and that the tool was relatively easy to learn and implement.  This would lead to more time in the classroom for teaching statistical concepts, rather than spending much time teaching how to use the software program.  The authors thus conclude that the use of the software tool can be an effective teaching tool in an introductory statistics course.

For me, in developing an introductory statistics course, the article is directly applicable and provides ideas worth considering.  If there is a software tool that can assist in the teaching of the course subject, it could prove very valuable.  However, there is one huge issue I have from the larger teaching perspective. 

The authors argue that an introductory statistics course should concentrate more on the role of statistics in decision making and less on the details of specific statistical tests.  Although it can be time-consuming up front, understanding the underlying math gives the decision-maker confidence in knowing what the statistics mean, and allows for a deeper understanding of the underlying assumptions and input parameters, thus improving the confidence in the decision.  The individual may not remember the underlying math, but will be able to go beyond just accepting the statistics at face value.  In this case I would be reluctant to use a software tool that, even though it may be user-friendly, gets away from achieving this deeper understanding.  Even though the authors conclude that learning statistical concepts in this manner is just as effective as using traditional methods, they admit that it may not help in developing statistical reasoning.  For me, the authors have it backwards.  A decision maker should understand the mathematical reasoning that underlies the statistic, and the use of software statistical programs could then be used as a time-saving tool.