Comments on: Unemployment Rate Area Chart

By: Jason Fichtner

Jason Fichtner — Fri, 21 Feb 2014 21:46:45 +0000

My apologies for not responding sooner. Thank you all for the very helpful comments. I’ve been circulating your suggestions around the office for internal comments. We’re definitely going to incorporate some of your suggestions into our next version. I’ll post what we end up doing once we nail it down!
-jason

By: Jeff Harrison

Jeff Harrison — Sat, 01 Feb 2014 13:45:00 +0000

And here’s the line graph.

By: Jeff Harrison

Jeff Harrison — Sat, 01 Feb 2014 13:43:57 +0000

Hi, Jason.

You’ve chosen a good graph for the message you’re trying to convey. You could use a line graph for this but I think that’s a better choice when you want to plot each category separately (i.e., you were more interested in comparing the trends and quantities of the categories than in showing the cumulative effect). There’s no wrong answer; there may be useful insights from either view. I have an example of each that I will attach.

With area graphs, when the different areas represent ranges of a quantitative variable, I like to use different shades of a single color to reinforce this. This is not true in this case–the different strata are categories of unemployment, not degrees of unemployment–but it’s not hard to interpret the categories on a continuum from most to least severe, and I think this treatment reinforces that interpretation. In this view, in other words, we’re probably less interested in adding together the rate of discouraged and part-time workers and omitting the other two. What we want is to see how the rate changes as you start from the official rate and then expand our definition to include the other categories.

The rest of what I did was just cleanup and editing. There were a lot of visual elements that could be removed or made subtler without sacrificing meaning. These include the borders on the areas and around the outside of the cart, the months in the legend, and the grid lines, which tend to focus the attention on questions like “in what month did the rate hit 10%” instead of the bigger picture. I also moved the annotations on to the data to reduce cross-referencing and to expand the area available for the graphic.

Hope this helps.

Jeff

By: Ed Fitzgerald

Ed Fitzgerald — Sun, 12 Jan 2014 01:19:45 +0000

To better use the comment space.

By: Ed Fitzgerald

Ed Fitzgerald — Sun, 12 Jan 2014 01:16:06 +0000

Here it a line chart version with millions of people and percent change.
The percent change is different from what I could figure out from the original, this goes to the clarity.

By: Ed Fitzgerald

Ed Fitzgerald — Sat, 11 Jan 2014 15:58:41 +0000

I have 3 comments, first visually, make the areas a bit transparent, so the grid lines are visable, its difficult to make any assessment of values other than U6, or just use lines, not area.

Second, all 4 lines seem to follow the same pattern, what does it look like if the rate of change in shown?
Just from Jan 2005 to Dec 2013 they look like.
U6 – 42%
U5 – 28%
U4 – 30%
U3 – 27%
Maybe just show U6 and U3 numbers.

3rd, this is data about people, it would have more impact if you could show the number instead of or in addition to the percentage, so the reader is still connecting the data to human beings.