An Experiment In Teaching Statistics At Primary School In Italy

We report an experiment which involves teaching statistics at compulsory school in Italy (145 primary school teachers and 2130 pupils, 6-10 years old, living in six Italian regions from north to south, giving a total of 132 primary classes were involved) where we evaluate the different efficiency of two different methods of teaching: by Objectives or by Concepts. In this experiment the fact that both these approaches anticipated working with real data reflecting a phenomenon of daily life is of importance. For testing this hypothesis we used multidimensional permutation tests by non parametric combination of dependent tests. Overall the results show that the method by Objectives is better than the method by Concepts.