Data Visualization Unit (5 min)
For the next (roughly) 2 weeks, we will be focusing more explicitly on data visualization. We have been talking around data visualization throughout the term, as you have been including them in your writing thus far but also in terms of more general points about writing with data:
- how do you find patterns or stories about your data (visualizations help at stage of invention!)…
- and tell a story about your data? (visualizations can be more efficient in telling non-linear stories!)
- How do you communicate non-intuitive information? (related, do visuals ever help with complex calculations or datasets)
- How do you compose visuals ethically? Ways to misrepresent your argument or data? (e.g., scales on axes, choices of comparisons, units of measurement)…you’ll find similar issues as we have dealt with in prose, but there are also new problems here, as well.
You have two final projects for the course, both due at the end of the term (by December 12 at 11:59pm). One will be a reflective paper, a sort of beefed up version of a Learning Narrative. The other will be a revision of a past Data Visualization or a new Data Visualization. I’ll get more into the nuts and bolts of that assignment next week.
Today we will talk tables, next class we will talk charts, and then next week we will talk about non-traditional visualizations as well as start to work with free software from Tableau where you can do some cool things to make some visualizations.
Tables (30-45 min)
Look at this report on “championship windows” for major league baseball teams. Skim each IMRaD section to get a sense of what the argument is (the introduction and methods section might be particularly helpful for this activity).
There are three tables in this report. I want you to focus Table 1 on National League teams and compare it to my prose example below (forgive the brevity, but you get the idea).
Prose Comparison:
The Atlanta Braves have had 7 championship windows of 3, 1, 8, 1, 2, 17, and 4 years, averaging 6.80 years (excluding single-year contending years). The St. Louis Cardinals … averaging BLANK years. The Chicago Cubs … averaging BLANK years. The Los Angeles Dodgers … averaging BLANK years. The San Francisco Giants … averaging BLANK years. The Philadelphia Phillies … averaging BLANK years. The Pittsburgh Pirates … averaging BLANK years. The Cincinnati Reds … averaging BLANK years.
This reveals that every NL team has had at least one single-year contention event. This phrase means that there was a single year where the team contended, sandwiched between two (or more) years of non-contention. But for purposes of this analysis, the question is: Should a single-year contention event be considered a window? Although this is debatable, single-year contention events will not be considered a window in this analysis, because the concept of a window closing implies an already open window. Therefore, a single-year contention event cannot be considered a true window. In terms of this paper, this means that single-year contention events are not included in the calculation of the average length of windows or the number of windows for each team in both tables.
So what can we learn from the data? In terms of number of windows, it’s not surprising that the expansion franchises have had fewer windows historically than the franchises that have been around since 1903. Among the latter, the Cardinals have had the most separate windows with 10, while the Phillies have had the fewest with three. The three longest windows are 22 years by the Giants (1917–38), 17 years by the Braves (1991–2007), and 16 years by the Dodgers (1970–85). The Giants have the longest average window duration at 8.75 years, while the average window duration of the 1903 teams as a group is just over six years. The average window duration for the expansion teams is 3.7 years, and the overall NL average window duration is 5.5 years. Seven of the NL franchises had windows that were still open after the 2016 season.
What kind of “reading” do you do with and without a table? How is the reading experience different?
For the article with the table, how does the text refer to the table? How is it integrated into the text? What does it bring to the…(here it comes, it’s bad, I’m sorry)…table?
Different functions of a table
On CourseWeb, there is an article from Research in the Teaching of English. As with the previous text, skim the IMRaD sections to get a sense of what the article is about.
What is the function of each table? What does one table do that the other tables do not do? How do they contribute to your reading experience?
Considerations for making tables
- What is your table? Is it a database (which is fine for a text where this makes sense as a function, like some government reports or factbooks)? Or does it bring together information surrounding one topic or sub-topic relevant to the flow of your paper?
- What is the order or organization? How does it mirror your writing?–alphabetical, high/low, theoretical/argumentative, order of questionnaire, etc.)
- Are you referring to the table? How so? And when?
- What is your title? Does it name each of the major components of the relationships in your table (e.g., the variables, how you are comparing the variables, the key statistics)?
- Are there layers to the information you are reporting where you’ll need to use things like spanners, panels, or indentations? (see p. 124 in Miller for example of table with spanners, panels, and indentations)
- Do you need to note things at end of table to reduce confusion and increase likelihood that table can stand on own (e.g., notes about symbols for various statistics like statistically significant at a certain alpha level, expanding on definitions of column titles)?
- What design and accessibility concerns do you need to address? (e.g., would colors or shading help draw attention in helpful ways?)
A table, like prose and like other figures, is emphasizing something. Some pattern or something notable in relation to the overall story you are telling. What is included and why?
Let’s try to make a table with some data
Go to CourseWeb to download the earthquake dataset (these are the past 7 days of earthquakes of 2.5 or higher in U.S.).
Make a table! Look at above considerations and from Miller. What is the story you want to tell?
For tips on making a table in MS Word, check out this tutorial.
This resource also has some common formatting help about tables. It does not cover things like shading and line point help, but these are valuable, as well, to create color contrasts between every other row or to distinguish parts of table by making lines thicker or thinner. Another helpful thing is removing or adding borders (something important for having a cleaner look or to meet requirements of documentation, like for APA format).The table menu bar (or table properties, too, when right clicking) is your friend for this stuff:
Leftovers: Survey, LN2/3, PW (5-10 min)
Writing is hard, man: Tone/style? Revision? How to do this, where to start? Don’t be shy about this stuff, I’m happy to help! Writing Center is also great, too.
LN2/3: I don’t like the “what did you learn so far” framing that I provided in LN2, that was a lame transition question that one student says is nebulous and vague, which I agree.
Focus more on the details, the in-between spaces of what is going on in your writing and how that relates to the things we talk about in class and to things you are theorizing about quantitative and data-driven writing. Read your writing with a new eye. What is going on there? What is happening? What do you admire? What is still to be worked out?
PW and Next proj: what data do and don’t say, contradictions, limitations…and how you still make an argument through all of that? What ethical considerations need to be more foregrounded when it comes to the possibilities and limitations of data-driven writing? Even for driest of dry papers, this is a consideration I want you all to sit with as you write with data here and elsewhere.
Stats: don’t forget assumptions behind methods. Please seek me out if you have questions about using stats in your analysis! Here to help. Also, see #3 under the questions for Sci/Tech Writing Project Index for more on this.
In-Class IMRaDing (20 min)
Return to questions from last class.
Focus on Intro/Results/Discussion VS. Methods. We have a sense of trying to make a recipe, of how to provide enough information so someone else could carry out our analysis and learn the necessary context to understand our data.
Next Time (2-5 min)
-ch. 7 in Miller
-Journal 6
-Work on projects