This is basically a bubble chart that uses words instead of symbols. The advantage is the reader doesn’t have to refer to a legend to see what color bubble refers to what data. In the linked example I am comparing subjective values such as average lifespan of cars by brand (largest font = longest lasting) against their average sticker price (Y axis).
I don’t have a name for it, but IMHO, it’s an awful way to present those data. Why the distracting left/right positioning? If you are looking to show correlation, why not have a scatter plot, with x=life and y=price, with names next to each of the six points? If you are not interested in the correlation, why not a simple table?
(My apologies for not actually providing an answer to the question you asked. It’s just that data presentation is a big pet peeve of mine. Hopefully other posters can provide useful information. )
Because I don’t have exact data values for the chart I’m working with, but I do know the general relationships. The vague placement allows one to be very general with respect to actual data values. I wouldn’t feel good about using X-Y coordinate points unless I could precisely justify my source data.
Do you have a similar objection to bubble charts? They also have random X-Y placement of the data. My dislike of bubble charts is that they get too cluttered and the reader is forced to dart the eyes back and forth between the chart and a legend to find out which color represents what data.
I think that chart is great. Your data representations sound anal.
The chart communicates very intuitively by making natural use of fonts (you know… the way I’M COMMUNICATING SOMETHING HERE) and its roughness signals the right thing about the underlying precision (great point, patty). It takes a simple amount of information and does an excellent job presenting it. You have to be careful, though, that the font sizes are weighed correctly. This is highly subjective, and in fact intricate if the result is to be useful instead of misleading.
But a scatter plot?? You’re right, you’d be better off with a table.
btw, word clouds are dumb. They’re hyped bs that needs to evolve. The y-axis and the use of just a few words was an effective step beyond the rudimentary “hey let’s use fonts” idea.
Well, just so you don’t feel alone, Pasta, I don’t like the chart either. It’s not immediately intuitive to me, much because I’ve never seen a chart like that before and I expect the left-to-right axis to mean something when presented a chart in that manner. So, font size is important, left-to-right is not important, up-and-down is important. Seems like a confusing mess to me.
I’d call it a bubble chart with invisible/word-shaped bubbles.
I’m not a big fan of bubble charts, either. In particular, using bubbles to represent something like longevity that does not correspond to size in my brain does not work. I don’t think I’d ever choose a bubble chart for anything where bubble size represents something other than quantity, size, or money. Also, bubble charts shouldn’t be used to obscure the inaccuracy of the data.
If you really want or need to make this type of chart work for this application, I would eliminate the extra axis, plot different size bubbles along a single line (with branches if needed for legibility), and label the bubbles with the names (or logos) of the manufacturer.
For this application, I do think a scatterplot would be better. Data on car prices and longevity exist, just give a footnote explaining whose data you are using and any limitations (averages, different years, etc.)
I concur, and I did understand what it was showing me.
I’m also in the “word clouds are bs” camp, fwiw. Solution in search of a problem: the only people who “need” them are the nitwits who figured out how to make them.
For the one linked in the OP, what does it signify? Is a $21,000 Chrysler the same size as a $35,000 Volvo? $18,000 Fords are tiny?
They really are stupid - here’s one lifted verbatim from a website. This particular company sells physical things, and other than the “tools” at the bottom, none of this has anything to do with what they sell- it’s just people screwing around.
annswishes bd Beadles Beanie bisbee Buy cents chopchop
cool epyes ewm for fumehead homeschool Ilena jewelry john
Kitchen kmkm later laughter light maybe MissusWizard mommo Order **Peggy ** **qrxzzzz ** qwerty SallyB school Stuff
SuellenCraft SuellenKids toget tools uncouth want wenal05 willie
I appreciate everyone’s input here. Now that I know the name of my diagram, maybe this thread is better suited for IMHO.
Most charts and graphs present data without much insight. I guess I’m going for the reverse here - provide insight into some relationships without requiring the raw data to back it up. Such a thing is useful for executive summaries where Presidents and CIO’s probably wouldn’t understand the raw data even if it was presented to them.
Some executives need to be kept away from the ability to “drill down” when presented with a graphical analysis, lest they attempt to formulate incorrect conclusions based on a misunderstanding of the source data. What they are interested in is the general relationships between groups of data or even interpretations based on metadata.
Granted, because the diagram’s meaning may not be intuitively obvious, there would need to be some explanatory caption or better yet an oral explanation made during presentation.
I agree. Some charts visually represent numbers. Some charts visually represent thoughts. A lot of people want to communicate their thoughts, but get hung up on numbers.
Of course, although your observations about executives are apt, the executives probably don’t enjoy getting babied. In fact, I think it’s a good idea everywhere… science publications as much as corporate presentations… to do both. Express your analysis, but include the data as well. If people start questioning your analysis, it’s either you didn’t communicate well or in fact they might have a point.
)