Here, we summarize the takeaways for the investigations presented in this article.
• Our results show that some fonts, including EB Garamond and Montserrat, tend to increase the reading speeds of readers older than 35 years, on average; however, such recommendations can not be made for the general population.
• While fonts and typographic considerations generally matter more for older participants, starting at age 35 in our data, we observe larger reading speed gains when accounting for individual differences per font per participant.
• Different fonts are effective for different people, leading us to believe that custom reading experiences can help people read more effectively. By designing tools to match individuals to their best font, individual readers can experience large gains in reading speed.
• Preference ≠ effectiveness. People do not know what is good for them in terms of font choice for reading.
• Preference for fonts is personal. People differ in what they prefer.
• Familiarity with a font does not drive preference, and has a small effect on reading speed. A font need not be chosen for an application just because people may be used to seeing it.
• A single size does not fit all fonts. If an application has a few font options for the same piece of text, then each font needs to be adjusted in size according to the font’s characteristics.
• When normalizing fonts based on font attributes, x-height is not the only option, nor was it the most common option selected based on participant perception.