Summary:
A new approach to teach basic statistics to young researchers is proposed. It hypothesizes two main causes for the wel1 known "mathematical trauma": a) lack of basic mathematical concepts; b) failure to understand the mathematical language. Related ancillary concepts, presented just before use, are adequate to overcome these limitations. Even more important, is to show why and how statisticians develop a technique or equation. This approach was tested on a class composed of 25 graduate students. Before the course, a pretest was presented. The final evaluation included a similar test, and an anonymous questionnaire about the teacher and course. Analysis of these data revealed an improvement of knowledge in statistics and approval of the classes. Subjective factors and difficulties in extrapolating the validity of the experiment to other student group are recognized. Nevertheless, the results are sharp enough to suggest reproducibility in similar situations.
Key-Words:
Learning Difficulties; Statistics; Medical Education