Pre-service primary school teachers' statistical knowledge in data representation

In this article, we analyse if pre-service primary school teachers distinguish the categories or values of the statistical variable from the absolute frequencies when calculating the mode. Moreover, we investigate whether they select and construct the appropriate graph according to the type of variable and how they justify such choice. The work is carried out from an interpretative paradigm within mixed research, with a predominantly qualitative and exploratory nature, classifying and interpreting both the production of statistical graphs and the argumentation for their choice. The sample is conformed of 108 students in the third course of the Primary Education Undergraduate Degree of the University of the Basque Country. The instrument used for the collection of information is an ad-hoc questionnaire composed of 4 items. To analyse the information, two rubrics are presented. The first classifies the levels of correctness of the statistical graphs produced and has been compared with the work of Arteaga (2011). The second classifies the levels of justification for the selection of the graph and have been contrasted with the studies of Berciano et al. (2021) and Alacaci et al. (2011). In order to interpret the information a process of triangulation among researchers has been carried out. The results indicate that future teachers have difficulties in distinguishing the categories of a qualitative variable from its absolute frequencies. As regards the choice of graph, the most common errors are to make line graphs for qualitative variables and histograms for qualitative or discrete quantitative variables without grouping. The degree of correctness in making the graph and its justification is not sufficient, which shows the need to improve their statistical literacy. In conclusion, the following difficulties have been detected among pre-service primary school teachers: 1) Difficulty in identifying the type of variable: tendency to classify any statistical variable as discrete because absolute frequencies take natural values. 2) Difficulty in calculating the mode of qualitative variables. 3) Difficulty in the selection of the graph according to the type of variable. 4) Confusion between bar chart and histogram. 5) Absence of titles or labels, scales or values/categories on the axes. 6) Absence of values/categories or absolute/relative frequencies in the pie charts. 7) Difficulties in the use of statistical language: lack of precision and confusion in the use of the terms data, statistical variable, value and category. 8) Lack of statistical argumentation.