We have just finished another semester, and before my mind completely turns to rubble, I want to share what I believe to be a fairly good assignment. What I present below was parts of two separate assignments that I gave this semester, but upon reflection I think it would be better as one.
Read the article Let’s Practice What We Preach: Turning Tables into Graphs (full reference given below). In this article, Gelman, Pascarica, & Dodhia suggest that presentations of results using graphs are more effective than results presented in tables.
Gelman, A., Pasarica, C., & Dodhia, R. (2002). Let’s practice what we preach: Turning tables into graphs. The American Statistician, 56(2), 121–130.
Find an article in a journal that presents results (or data) in a table. Re-create the data in a tabular format using R (or Excel).
- Use the functions in ggplot2 to produce a plot that conveys the same message as the original table.
- Include the original table (this can be a screenshot or web-link) and citation, along with your plot.
- Write a few sentences describing why the plot you produced provides a better presentation of the results or data (be sure to use recommendations from the article in making your case).
In the second part of this assignment, you will write a tutorial for the process you followed for turning a table into a plot using R Markdown and will publish that tutorial on RPubs.
There are several resources for learning R Markdown.
- RStudio’s [documentation] for writing a document with R Markdown
- Yihue’s [screencast] introducing R Markdown
- An [example/tutorial] from PSU
Your tutorial should be written so that a student who was just learning ggplot could follow your directions easily. Include instructions for obtaining the data, getting it into a useable tabular format, manipulating the data so it can be used with ggplot, and well-commented instructions for creating your final plot. (Think of the level of detail you would want in a tutorial when you were first learning ggplot!)
It should also include:
- a citation or link to the website/journal that published the original table
- a view of your final data (full or a subset depending on size)
- all commands necessary to create your final plot (with appropriate explanation), and
- the final plot
When you knit the .Rmd document it should compile without errors.
Students commented that they learned a lot about the use of ggplot during the initial assignment (this was the second assignment in the course). The Markdown part of the assignment I gave as an extra credit assignment at the end of the class, but in retrospect, I should have made it required and done it very early on.
Here are a couple of the tutorials that I have received so far:
- These students took a table of characteristics of survey participants published in the Journal of Ethnic and Cultural Diversity in Social Work and turned it into a bar graph. http://rpubs.com/TSK_2012/3184
- These students took data about trends and topics discussed in Seventeen Magazine‘s Traumarama articles from 1994-2007 and turned it into a line plot. http://rpubs.com/opalc123/3155
- These students took a table of data related to approval ratings and turned them into a box-and-whiskers plot. http://www.rpubs.com/GeorgeBrisse/3217
- These students’ work depict a great example of how data initially presented in a table is much easier to process in a graph. The data, from a table published in the Journal of Deaf Studies and Deaf Education, show the academic status and progress of deaf and hard-of-hearing students in general education classrooms. http://rpubs.com/mens0055/3211
- These students used a stacked bar chart to show data about the sample sizes for different stages for 12 problem behaviors published in Health Psychology. http://rpubs.com/nikedenise/3256
- These students created a line graph representing pre- and post-training scores for consonant, vowel, sentence, and gender perception scores in cochlear implant users to examine whether an auditory training program improves performance. http://rpubs.com/koern030/3255