Overview
In the domain of typography, the choice between Serif and Sans Serif fonts is more than just a matter of taste; it influences how content is received and perceived by readers. And with digital communication growing ever more important, understanding the visual appeal and readability of different font styles has never been more critical.
Serif fonts, with their decorative markings at the ends of certain characters, are the classic choice for print media such as newspapers or books. San Serif fonts, on the other hand, are minimalist and often used as defaults for emails and business reports. While there’s currently no consensus on which font style is better, it’s broadly assumed that Serif fonts are more aesthetically pleasing while Sans Serif fonts are easier to read.
But there’s also the question of how font choice impacts the message itself. For example, might serif fonts, with their decorative tapers, make a passage seem more creative?
To investigate these questions, we designed a textual experiment pitting Serif and Sans Serif fonts against each other along three dimensions: Visual appeal, readability, and creativity. While many might assume that Serif fonts enhance visual appeal, our findings challenge this notion, revealing that Serif fonts may actually just be harder to read compared to their cleaner, more straightforward Sans Serif counterparts.
The Experiment
For our experiment, we had 1,203 people from the research platform Prolific read a short story about a rock. Unbeknownst to participants, we randomized which of six fonts (three Serif, three Sans Serif) was used for the story each participant read. The fonts for this study were chosen based on their popularity, and included Times New Roman, Garamond, and Georgia for the Serif fonts, and Arial, Calibri, and Helvetica for the Sans Serif fonts.
The story participants read is as follows:
“Perched atop a windswept cliff, an ancient rock bore the scars of countless storms. Over millennia, it witnessed the relentless dance of the ocean waves below, carving stories into its craggy surface. In its silent vigil, the rock stood as an eternal sentinel, guarding the secrets of the deep.”
Participants then answered a multi-part survey question to measure our outcomes of interest. Participants were asked “To what extent do you think each of the following words describe this text?” with answer options given on a 1-7 scale (1 = Not at all, 7 = Very much). The words included “Visually appealing,” “Easy to read,” and “Creative.”
Results
Overall, there was only one recognizable difference in serif vs. sans serif fonts—ease of reading. Serif fonts were rated about 5.4% lower on our ease-of-reading scale (p = 0.006). However, this difference is quite small in statistical terms, only 0.16 standard deviations, which many social scientists would consider trivial. Nevertheless, this difference can add up when considering the number of digital communications we send on a daily basis and the growing number of people who read them.
In particular, a few specific fonts may be better choices for readability than others. For example, across our six fonts, Calibri scored highest in readability (avg. = 4.97), whereas Garamond scored about 10% lower (avg. = 4.48). But overall, most of these fonts were within the margins of error to each other.
Conclusion
This study shows that while many think Serif fonts look better, Sans Serif fonts are actually easier to read. Among the fonts we tested, Garamond was found to be the least readable, while Calibri was the easiest to read.
These findings suggest that for digital content, where quick and clear reading is important, choosing Sans Serif fonts like Calibri can be more beneficial. This is especially true for emails, websites, and other online materials where readability is key.
Content creators and designers should keep this in mind when selecting fonts to ensure their messages are easily understood by their audience.
Methods Note
We used ordinary least squares (OLS) regression analysis to test for significant differences in outcomes between our texts using serif or sans serif fonts. For significant differences, the coefficient would be large, and the corresponding "p-value" would be small (p < 0.05). We also conducted OLS regression analyses with interaction terms to assess whether these differences varied by participants' gender or age.
Data and survey materials for this study are available upon request.