Introduction and Overview

Teaching Statistics using R and RStudio

Plan for today:

• Welcome: introductions and overview

• Vignettes and related resources: guide to additional materials

• Intro to RStudio: the basics

• Less Volume, More Creativity: making R and RStudio more accessible to students and instructors

• Markdown: reproducible analysis to facilitate projects

• First day: what to do on the first day

Welcome to the “Teaching Statistics using R and RStudio” shortcourse As we gather, please check out: http://www.amherst.edu/~nhorton/skidmore for useful information related to the shortcourse This has pointers to resources, vignettes, and other information.

I want to start by acknowledging:

• Albany Chapter of the ASA: for organizing the shortcourse
• Michael Lopez: for hosting
• Andy Choens: for agreeing to help shoulder surf
• NSF: DUE 0920350 for support of Project MOSAIC (Kaplan PI, Pruim co-PI)
• R and RStudio community: many people who have made R and RStudio what they are

Project MOSAIC (http://www.mosaic-web.org) is an NSF funded initiative to create a community of educators to develop a new way to introduce mathematics, statistics, computation and modeling to students in colleges and universities (Kaplan PI)

Our goal: Provide a broader approach to quantitative studies that provides better support for work in science and technology. The focus of the project is to tie together better diverse aspects of quantitative work that students in science, technology, and engineering will need in their professional lives, but which are today usually taught in isolation, if at all.

The name MOSAIC reflects the first letters — M, S, C, C — of key components of a quantitative education:

• Modeling: The ability to create, manipulate and investigate useful and informative mathematical representations of a real-world situations
• Statistics: The analysis of variability that draws on our ability to quantify uncertainty and to draw logical inferences from observations and experiment
• Computation: The capacity to think algorithmically, to manage data on large scales, to visualize and interact with models, and to automate tasks for efficiency, accuracy, and reproducibility
• Calculus: The traditional mathematical entry point
  

• Big data (or even medium data) is not going to be analyzed on TI calculators
• How do we build the capacity to make sense of the data around us?
• We need to start early (and make powerful technologies available to students and instructors)
• Our approach: make it easier to use R and RStudio (free software: less barriers)

To facilitate use of R and RStudio and the mosaic package for the teaching of statistics, we have created a number of online resources.

These can be found at the Vignettes link at http://cran.r-project.org/web/packages/mosaic

We welcome comments and suggestions for any of these materials (which are distributed under a Creative Commons Remix with Attribution License)

We will refer to these resources throughout the workshop.

• mosaic-resources: pdf file with links to other resources
• MinimalR: a minimal set of commands needed to teach using R and RStudio
• Compendium: a compendium of commands to teach an intro stats course
• Resampling: how to undertake randomization-based inference using R
• Modeling: a strategy for teaching modeling in intro stats using R

Other materials may be useful for examples or working code snippets

• Sleuth: provides the code and output for the 2nd and 3rd editions of the Statistical Sleuth
• IPS: provides the code and output for the 6th edition of Introduction to the Practice of Statistics
• Lock5: provides the code and output for Statistics: Unlocking the Power of Data (in process)
• fastR: Pruim's Foundations and Applications of Statistics: An Introduction using R
• Fresh Approach: Kaplan's 2nd edition of Statistical Modeling: A Fresh Approach

We will be using data from the Health Evaluation and Linkage to Primary (HELP) care randomized clinical trial (RCT) throughout this workshop (and in some of the vignettes):

• n=470 subjects without primary care enrolled in detox in Boston circa 2000
• followed for up to 2 years, every six months
• rich set of measures of behaviors, demographics and outcomes
• pdf of questionnaire available at http://www.amherst.edu/~nhorton/help
• more info available using “?HELPrct” after loading the mosaic packagen (see also HELPfull)

• R: free software environment for statistical computing and graphics (http://r-project.org)
• RStudio: powerful and easy-to-use free and open source user interface for R (http://www.rstudio.com)

RStudio is available as a client or server (Linux based) interface.

We find that this dramatically simplifies the use of RStudio in the first few weeks of a class.

Details can be found at http://www.rstudio.com/resources/faqs (under Academic)

Feel free to sit back and watch, or follow along If you want to install things on your own, you'll need:

• R version 3.1.1 or greater
• Preview version of RStudio 0.99
• mosaic package

• consistent interface
• improved interface

Let's take a quick tour of the RStudio interface…

• panels
• tabs

R (or its predecessor S) has been around for 30 years. Why do we now need the mosaic package?

R (or its predecessor S) has been around for 30 years. Why do we now need the mosaic package?

• simplify the learning curve (round off the rough edges)
• base R is now essentially frozen (so packages are even more important)
• scaffold their use of computing using the powerful concept of a modeling language (main focus for today)

• favstats()
• xpnorm()
• xchisq.test()
• mplot()
• googleMap() (see also STEW activity “Roadless USA”)
• rgeo()
• Cards

There are a number of ways to work in RStudio:

• Console
• Scripts
• Markdown (this is what we recommend)

See paper at http://escholarship.org/uc/item/90b2f5xh

Improved support in the Preview release

• outputs directly to pdf or word
• can be reformatted
• more later

Any questions?