Lane, D., Tang, Z. (2000). Effectiveness of simulation training on transfer of statistical concepts. Journal of Educational Computing Research, 22(4), 383-396.

The authors of this article conducted an experiment to determine if computer simulations could be as effective for teaching concepts in science education as they are in training people in strong perceptual-motor skills. The authors sought to evaluate the effectiveness of a statistical simulation of a sampling distribution in helping students understand statistical concepts. They wanted to determine if students would be more able to transfer what they learned about statistical reasoning to real-world situations if they were taught by the simulation as opposed to learning from the textbook.

The authors conducted their experiment using 115 undergraduate students at Rice University. Each student participant read two examples of statistical reasoning and three questions involving each example. The examples had to do with sampling from an urn of 33 gumballs. The students in the textbook-based groups then read about statistical sampling from a textbook. The students in the computer simulation group watched a computer simulation. After reading the textbook and viewing the simulation, each group was then given a set of questions to answer dealing with statistical probability. Their answers were rated according to a code developed to determine how much statistical reasoning they used in their answers. After studying the results, the authors found that the students who had watched the computer simulation answered the statistical problem questions more effectively than those who read the textbook. The authors discussed three factors that may have contributed to the better performance: the simulation was concrete rather than abstract; the simulation emphasized certain aspects of statistical probability that are more difficult for students to understand; and computer simulations are more interesting and thus more engaging than textbooks.

I thought this experiment was well conducted and the use of 115 students would allow for  significance in the findings. The authors made sure that each of the four student groups was randomly chosen, and that the questions given to all of the students were also randomly listed. The textbook section was actually composed of several well-written textbooks, and the computer simulation was videotaped so that each student group had the same presentation. The authors did a good job of making sure each stage of their experiment was as neutral as possible. This is another experiment that could easily be duplicated at another institution.

While the study demonstrates that computer simulations, even when they are merely watched and not actually utilized by students, can lead students to a better understanding of mathematical concepts, the authors stress the fact that computer simulations are just one way of teaching students about concepts, and that "an experienced instructor would employ every technique possible to ensure the quality of student learning" (391). As we have read in this class throughout this semester, computers can enhance the quality of learning in a classroom but should not replace all or even only the traditional methods of teaching. In a way, this provides teachers with some breathing space -- they don't need to rely completely on modern technology in their classrooms, but knowing various ways in which computers can be used in the classroom can increase their ability as teachers.