In this study, 48 subjects who had no previous exposure to probability or statistics read one of three texts that varied in the degree of explanation of basic concepts of elementary probability. All texts contained six formulas, each accompanied by an example as well as definitions and information logically required to solve all problems. The high-explanatory test differed from the low-explanatory and standard texts in that it emphasized the logical basis underlying the construction of the formulas, the relations among formulas, and the relations of variables to real-world objects and events. On both immediate and delayed performance tests, subjects in the low-explanatory and standard-text conditions performed better on formula than on story problems, whereas the subjects in the high-explanatory text condition did equally well on both types of problems. It was concluded that explanation did not improve the learning of formulas but rather facilitated the application of what was learned to story problems.
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