ABACUS: Apps Based Activities for Communicating and Understanding Statistics

ABACUS, developed under the R Shiny framework, aims to communicate basic statistical theories in a user-friendly and interactive way. ABACUS implements multiple options to convey the concepts: it takes advantage of statistical simulation to simulate varying datasets, creates graphical interfaces to visualise the statistical concepts, and integrates the interactive features using the power of R Shiny. ABACUS incorporates the framework of frequentist inference: that a given experiment is an instance of an infinite sequence of possible repetitions of the same experiment under identical experimental conditions, each capable of producing statistically independent results. The current version of ABACUS includes properties of Normal distribution, comparison of Normal and t distributions, properties of the sampling distribution, one-sample z and t-tests, two samples unpaired t-test, and analysis of variance. ABACUS provides a framework for essential statistical tools that educators can employ in a teaching environment with a range of learner-driven activities. The flexibility of input values allow integration with relevant examples from broader subject areas. Combining the simulation scenarios and graphical interfaces, it provides a platform for the learner to understand the basic principles of statistical methodologies, identify best practices for the presentation of results and acknowledge the correct interpretation of results often inconspicuous in standard statistical software outputs. The context of ABACUS also aligns with the ASA’s statement on being thoughtful and open to the interpretation of research outcomes beyond the statistical significance. ABACUS is currently available as an R package (cran.r-project.org/web/packages/ABACUS) that users can run with minimal R knowledge. A dedicated web interface is also maintained to support learners unfamiliar with the R environment (shiny.abdn.ac.uk/Stats/apps/). The web version also includes some additional applications, including the comparison of frequentists and Bayesian inferences.