Design Exercise 2: Critique Practice

in Design Exercises

The goal of this exercise is to give you some practice at critiquing visualizations, and thinking about how that critique can help lead to better designs.

This exercise builds on Design Exercise 1: Questions from Visualizations and Data - so if you haven’t done that yet, you should. Reading this assignment might skew your thinking on that one.

When writing out critiques, it feels excessive to frame them in the “if the goal is X then Y” format - but I want you to take the time to do that.

Also, this early in the semester, we lack having a good background in principles and examples, so it might be hard for you to come up with principles or suggestions. Use your intuitions. And this will hopefully help you appreciate why we need to build up our toolbox of principles and examples.

For the questions here, we’re taking the view that a “critique” is a specific aspect that is identified; the “overall critique” of a visualization might involve many of these specific critiques.

This exercise will be turned in as a Canvas Survey: DE02: Critique Practice (due Tue, Sep 27). We recommend you type your answers offline and then upload them. The Canvas survey will give you a point for completing it. If you don’t give reasonable answers, or are excessively late, we may take that point away. Any grading will be scored separately.

Part A

Last time I provided some visualizations built from the Census Data Set data set. I showed you some visualizations, and asked you what questions they could answer. More or less, to guess what I was trying to do when I made those visualizations. Well, now I am going to tell you…

In general, I was trying to show how the level of education in a community impacts/correlates with the economic outcomes(we’ll use different measures). I am interested in the general trend across the country, but also curious about how this plays out here in Wisconsin and Minnesota - do these states have similar trends?

Put aside for a moment that we aren’t trying to do “good social science”.

For example,
  • We are using “counties” as the unit of analysis (equating it as a “community”) - in practice, counties vary in size, and are not homogeneous.
  • We are using the variables we are given as proxies for more interesting things (e.g., percentage of adults without a high-school diploma as a measure of levels of education in a community).
  • The real underlying phenomena are complex.
  • There are lots of outliers and special cases.

The visualizations I made were not great (which makes them easier to critique). I did not intentionally make bad visualizations - they were made quickly, using Tableau (and I am not an expert). I tried to use “standard designs” and did not nail all the details. And, I did not take the time to critique and re-design to refine them. That’s what you get to do.

We’ll start with the two visualizations that provide county level views from last time (these are education and unemployment - and location):

We’ll start with critiquing the map. (Census-22-maps.pdf 3.1mb)

Here’s one:

If the goal was to compare Wisconsin and Minnesota, providing contrast between these two entities can help the viewer see differences (or lack thereof); not showing the boundaries of the states makes it harder to make the comparison.

Notice that this critique tries to identify the problem. We separate critique from suggesting solutions. Critiques often (I think the book reading would argue that they should) focus on the problem; these lead to suggestions for things to try in a redesign.

A.1 If we were to try a redesign based on this critique, what change would you suggest trying? (you can give multiple possibilities)

But, for the map, let’s ignore the Wisconsin and Minnesota aspect (since the map wasn’t actually considering it). Think about trends across the country for the moment.

A.2 Let’s start positive: give a positive critique about something that works in the map. If you need a hint, think of something you can see in the map but not in the scatterplot and connect it to a reason.

A.3 Now, pick something about the maps to critique that suggests a problem we might try to fix. You can suggest a response to the critique (i.e., a design change to consider), but that isn’t as important as doing the critique to point out the problem.

Now, let’s critique the scatterplot. It would be too easy to say “it doesn’t enable looking at locations” - so don’t pick that. But do identify some aspect of the design that you think should be considered in order to improve it.

A.4 Give a specfic critique of something about the scatterplot to improve its effectiveness, given what you know of the overall objective. Again, suggesting a response to the critique (what design change to make) is optional. It is more important to consider how to make a critique.

Part B

Here’s another example… A visualization from the pages of the Sunday, September 11th Wisconsin State Journal: (voting-law-changes.PNG 0.9mb). If it helps, the headline was “Rules haven’t defined turnout”, and the article makes an argument that changes in voting laws haven’t had large effects on voter turnout (the point is not to agree with their conclusion).

B.1. Give at least one specific positive critique.

B.2. Give at least one specific negative critique. Preferably, something that would help identify something to improve to make it more effective.