RA Tutorial, Fall 2020

Schedule

Week 1: Review of R basics

Readings

• R for Data Science, Sections 1-4

Tasks

• Make sure you have the latest versions of R and RStudio installed
• Complete R Programming course in swirl

Week 2: Data wrangling & exploration with the tidyverse, part I

Readings

• R for Data Science, Sections 5-8

Tasks

• Complete Getting & Cleaning Data course in swirl
• Develop a list of questions about your dataset (e.g. R4DS, Section 7)
• Write an R script which loads your dataset

Week 3: Data wrangling & exploration with the tidyverse, part II

Readings

• R for Data Science, Sections 12, 14 & 15

Tasks

• Write an R script which transforms your data (i.e. using filter(), select(), & mutate(); R4DS, Section 5). For this task, keep the following in mind:
  • Data should be tidy (see R4DS, Section 12)
  • Data should be transformed so as to answer the questions you developed last week
• In the script you create, include steps to explore and handle missing values (see R4DS, Section 12.5). Key questions to consider for this task are:
  • Are there missing values in your data? Are missing values represented in a consistently, or are there multiple placeholders for missing values?
  • Can you replace missing values with NA in an efficient way?
  • Are there patterns of missingness that you can detect? Might these be meaningful?

Week 4: Data visualization with ggplot2, part I

Readings

• Data Visualization, Chapters 1 & 3

Tasks

• Complete Exploratory Data Analysis course in swirl (stop at “Hierarchical Clustering”)
• Develop a list of visualizations you would like to create to illustrate effects of interest in your dataset
• Write an R script which creates relevant summaries of your data, using the group_by() and summarize() functions (i.e. R4DS, Section 5)
• If you have time, try creating a few basic visualizations using ggplot2

Week 5: Data visualization with ggplot2, part II

Readings

• Data Visualization, Chapters 4, 5 & 8
• R for Data Science, Section 28 (review)

Tasks

• Apply what you have learned from the readings on ggplot2 to plots for your dataset to improve aesthetics of your plots and add functionality such as grouping and facets
• Try using ggplot2 to create different kinds of plots from those which you have made already, using a dataset at this site

Week 6: R Markdown for reproducible reporting, part I

Readings

• R for Data Science, Section 27
• Reproducible Research with R and RStudio, Chapters 1, 2, 3 & 10

Tasks

• Create an R markdown document (and knitted html report) which organizes and displays all the code and figures you created for your datase

Week 7: R Markdown for reproducible reporting, part II

Readings

• Reproducible Research with R and RStudio, Chapter 12
• R for Data Science, Section 29

Tasks

• Review this tutorial to make sure you are comfortable with (a) markdown syntax and (b) R markdown basics
• Finalize your markdown document: apply a (non-default) theme to your R markdown report, be sure that each chunk of code is visible (along with its output), organize your document into logical sections labeled with headers and subheaders, and add (markdown-flavored) text to describe what each chunk of your code does. Be sure that the figures you create are a reasonable size.

Week 8: Distributions & descriptive statistics

Readings

• This tutorial on the normal distribution
• This tutorial on measures of central tendency
• This tutorial on measures of spread

Tasks

• Write some R code to apply methods from the readings to visualize and understand distributions in your data, and to compute measures of central tendency and spread
• Incorporate your R code into your R markdown document

Week 9: Correlation & sampling distributions

Readings

• This tutorial on correlation and covariance
• This tutorial on sampling distributions and standard errors

Tasks

• Write some R code to apply methods from the readings to (a) create scatterplots reflecting associations between variables in your data, (b) perform correlation analyses reflecting those associations, (c) compute standard errors/confidence intervals on the mean values for some variables in your data
• Incorporate your R code into your R markdown document

Week 10: Comparing two means

Readings

• (If you need a refresher on hypothesis testing & p-values) LSR Chapter 11
• LSR, Chapter 13
• The first three sections of this chapter on comparing means

Tasks

• Write R code to compare two groups or conditions in your data using appropriate t-tests. Evaluate and summarize what the test results mean, and add relevant code, output, and summary text to your R markdown document

Week 11: Comparing three or more means & factorial ANOVA

Readings

• LSR, Chapters 14 & 16
• The final section of this chapter on comparing means

Tasks

• Write R code to perform factorial ANOVA on your data, to compare groups and/or conditions. Evaluate and summarize what the test results mean, and add relevant code, output, and summary text to your markdown document

Books & Resources

Gandrud, C. (2020). Reproducible Research with R and RStudio (3rd edition). Chapman & Hall.

Healy, K. Data Visualization: A practical introduction. https://socviz.co/

Navarro, D. Learning Statistics with R. https://learningstatisticswithr.com/book/

Wickham, H. & Grolemund, G. R for Data Science. https://r4ds.had.co.nz/

Demos

example script

example R markdown document

example script: t-test & ANOVA

Useful links

swirl: learn R interactively

mycodingclub: variety of R tutorials

UCI ML database: variety of sample datasets