Uso de árboles de decisión para la enseñanza de la estadística inferencial

Abstract

The article shows the usefulness of decision trees to provide students with the choice of appropriate confidence interval, being able to estimate the mean, proportion or population variance or difference of means, proportions or ratio of variances. This choice can be a tough lot, and it depends on many factors, among which are: ●The number of samples to be compared. ●The normality of the populations from samples. ●The size of the samples. ●Knowledge of the population variances. ●The dependence of the samples. ●The equality of population variances. The authors have set 3 decision trees taking into account these factors and other theoretical and empirical considerations: 1. A general tree that helps to establish appropriate confidence interval for the estimate. 2. A tree that is used when we want to estimate the population mean. 3. A useful tree to choose the confidence interval when you want to estimate the mean difference. The article describes the use of decision trees with an example to illustrate its ease of use, speed and thus enhance their usefulness in teaching. One of the main virtues of the decision trees is that they have fully configured all the options available to make the estimate. Compared with other methods of assistance to the election, show the following differences: ●They are easier to use and require less time and space that the flow diagrams. ●Are more complete than the summary tables. This method can be used for other areas of statistics. On the other hand the decision trees have implications for use in computing, to systematize the process of choice, and can be used for configuring a statistical software more friendly than at present.