Descriptive Statistics and Linear Regression

A Compendium of Reproducible Research about Descriptive Statistics and Linear Regression
Abstract This document can be used as an illustrated tutorial about Wessa.net and FreeStatistics.org. It illustrates how Reproducible Research can be integrated in Statistics Education based on a Compendium of Reproducible and Reusable Research. 1

1

Introduction

This case study illustrates how free online statistical software (R modules) can be used in data analysis. This tutorial can be used at an undergraduate level the prerequisites are as follows: • a basic understanding of Descriptive Statistics • an interest in scientific data analysis • basic ICT skills All statistical techniques are based on the online “e-Handbook of Statistical Methods” [16] that is available online as HTML and PDF documents. The structure of the e-Handbook is like a cookbook: it contains a list of statistical methods in alphabetical order. There is also an overview of all methods that is ordered by type of analysis. In my experience with students it is best to use the electronic PDF version of the e-Handbook because it can be easily browsed and searched. Each method is explained in simple words and with a minimum of mathematical background. In addition, the documents contains numerous illustrations and examples which may also be beneficial. The case-based approach of this tutorial and the technology that allows to reproduce/reuse every aspect of the research, is consistent with the pedagogical paradigm of (social) constructivism. The fact that this tutorial has been designed as a compendium makes every aspect of the portayed research reproducible which empowers anyone with a healthy interest in statistics to interact (and conduct experiments) with the underlying statistical science. From the statistical point of view, the following concepts are used in this tutorial:
1 Acknowledgements: The Compendium Platform for Reproducible Research was partially funded by the OOF 2007/13 grant of the...