Autor: Gaspar Berbel  
Doctor UAB 
Trainner and Investigator
(ORCID ID: 0000-0002-9042-2922)

Reference book:
“Paola aprende estadística. Desde un entorno personal de aprendizaje”
(3ª ed.)

Gaspar Berbel (2020).

ESTATest shows which statistical test to perform for each combination of two variables.

ESTATest is an intuitive and easy-to-use mobile application. Designed for any student, teacher, researcher, or data analyst who needs to relate variables (perform hypothesis testing, bivariate relationship tests).

The app is available for Android and iOS devices.


The app guides the decision-making, realization, and interpretation process in a very simple and intuitive way. First it will ask you to mark the variables that it intends to relate to each other, their typology (categorical of two categories or more, or numerical or metric–quantitative), once the combination has been entered, the app will ask you some filter-type questions to determine the application conditions, according to the type of test, in relation to the supposed monitoring of normal law by the distributions or metric variables at stake , or the number of cases expected in a relationship procedure between categorical –Chi squared. In some cases, the app will ask you for the type of design of the variables to continue determining, with precision and accuracy, the type of test appropriate to carry out the requested relationship. Once all the filters have been applied –which can be passed by the user by default, marking the most common or expected situation in case the application assumptions are unknown- the app will indicate the test or statistical procedure to apply to see the possible association (relationship) between the variables. The tests that it will propose when the application conditions are met are the following: Chi square, t-test independent measures, t-test related measures, anova, correlation … In this case we are dealing with parametric tests – when the metric variable follows normal law , or we suppose it. In the event that this is not the case, when the metric variable in play does not follow the normal law – in case of non-parametric tests – the app will propose the following tests: U by Mann Whitney, T by Wilcoxon, kruskal Wallis test, correlation of Spearman. 

The app not only proposes the test to be performed, but also allows you to watch a short video explaining how the statistical procedure is performed and how it is interpreted, carried out with the IBM-SPSS statistical package. Videos last between one and two minutes.