Abstract:

We have introduced statistical problems, to be solved using the R software, into a Biostatistics course, in order to increase motivation for the field that usually requires not so basic mathematics to students that are not always inspired by it. The other goal was to complement the tradicional class style based on powerpoint presentation followed by pen and paper exercises with calculator. Also, this students have little or no experience in software as a professional tool since they are in the first year. Approximatly, one third of the two hundred students have been retained, from previous year, and have been exposed to a different mechanism of practical that was not working anymore. Students came from Biology degree, Teaching of Geology and Biology degree and Marine Sciences degree.

Each of the 200 students were presented with a total of four problems, during the semester, in the topics of Descriptive Statistics, Inference in One Variable, ANOVA and Simple Linear Regression. Students were requested to solve them at home and answer them in a form available in the Moolde Inquiry tool. Each student has his own different sample and also, questions were parameterized. For example, questions about Confidence Intervals were posed with different conficence levels (90%, 95% or 99%). Each problem has more than ten parameterized questions related to the same dataset. Moodle doesn’t do this type of tools so a small Python library was used to generate different problems and evaluate each student answers (numerical, textual or multiple choice).

To evaluate our methology, the request “Share ideas, thoughts and constructive judgments about the Problems and also about the course” has been posed while students were working in the third Problem and also after the First Written Evaluation. The last and fourth Problem has been answered in class and students were requested to grade sentences in a five item Likert scale. Questions were about effort, time, help from toher students and help from teacher.

The analysis of answers suggest that the methology of Problem Solving should be used again, with improvements, given the motivation and enthusiasm it promotes.

Keywords:

Problem based learning, statistics.