Credibility of information given font choice

Thanks to the Monkey Cage blog (a very good read if you’re a fan of statistics and politics as I am) I’ve come across an opinion article on the New York Times’ page that discusses whether the font or handwriting style that is used to convey information has any impact on whether we believe that information. It’s worth a read, as is the follow up and the preceding article (an essay and a quiz).

There’s a lot of really interesting stuff in there, but one that really leaps out is the ATLAS slides from the presentation about the Higgs Boson. Comic Sans is used to convey fun, light-heartedness, etc. and as far as I’m concerned it simply does not belong in academia. The article goes on to do some analysis of the previous article’s quiz and shows that Baskerville is the font with the most gravitas.

Rather than having me summarise the article here (it’s quite long) it’s probably better to just go and read it. If you’ve ever wondered about whether you should use a particular font (personally, Times New Roman is boring and Cambria is ugly) to convey a particular feeling in your article then this is food for thought. I use LaTeX to write papers and read a lot of papers typeset in LaTeX. As such, my eyes tend to see a lot of Computer Modern, which has a bit of a reputation as being an academic font. I wouldn’t publish a school newsletter in Computer Modern, of course, but for “serious” writing, CM is it.


7 thoughts on “Credibility of information given font choice

  1. Luisa

    I read that article (and actually participated in the original study, although I can’t remember what my impressions were at the time). I thought it was interesting, but… kind of obvious? I do find the psychology of font use pretty fascinating, especially what you mentioned: some fonts are too boring, some too ugly, some too frivolous, some too passĂ©. It’s amazing that tiny little curls and crossbars can inspire such strong emotions.

    1. Sam Clifford Post author

      I also find the font stuff a bit obvious but there are people out there who are unaware of the impact of font choice in the reception of the message. Like the ATLAS group. I think you can get very ridiculous with the elements of design in choosing how you present your message but it’s definitely something that people should keep in mind when writing. We can’t all be design gurus and typography nerds but we can all learn how to choose a font that conveys something other than “Comic Sans always screams fun”

      1. Luisa

        Yeah, I guess I really didn’t consider that it might not be so obvious to everyone but you’re right, if that were the case we wouldn’t have gaffes like the CERN presentations.

        Maybe data presentation should be part of our dream communications unit that should be compulsory for all science degrees everywhere.

      2. Sam Clifford Post author

        As much as it’s easy to go “I have this great idea, everyone needs to learn it”, I think you’re right that data presentation needs to be part of a communications unit within the B Sc. Helen MacGillivray did a bit of data presentation work in MAB101 when I took it with her, but I don’t know that she went far enough. I’d be interested to see Sama Low Choy’s notes from this semester’s MAB101. I know Sama’s a chart nerd and puts a lot of effort into visualisation of her own results but I don’t know that putting it in an optional core first year unit is enough, particularly when the focus is on learning the t, F, chi-squared and other similar tests.

      3. Luisa

        I don’t think we did any data presentation when I did MAB101, but I’m fairly sure that at some stage in my studies someone said “3-D bar graphs for 2-D data are ridiculous. Don’t do that.” My snobbishness about fonts is all my own.

        The new Science Communications co-major might have some data presentation in it (yet to be seen), but I doubt it will turn up elsewhere in the B. Sc.

      4. Sam Clifford Post author

        I remember that Helen MacGillivray really railed against the use of jargon in reports, e.g. “bathtub distribution” for a beta(a,b) (with a, b < 1), and the use of pie charts and 3d bar charts, but I think that might've been in MAB210 rather than 101. Part of the problem there is that almost all the data analysis people do in a B Sc is done in Excel and the default graphs are ugly, so rather than fixing up the colour scheme and axis labelling they opt for a flashier style of graph (3D line plots, for example) that look "impressive" but actually don't contain any extra information.

        Even if we have students doing their analysis in Excel, there are programs out there like Igor and Origin that can be used to build nice graphs from scratch. A few people in my group use these programs to make their graphs and I wouldn’t consider them data/stats nerds.

      5. Luisa

        I’d never heard of either of those! They could be useful tools to pass on to people in my lab (you know, since they’re definitely going to resist using MATLAB or R).

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s