Beymer, Russell, and Orton: "An Eye Tracking Study of How Font Size and Type Influence Online Reading"

Beymer, D., Russell, D., & Orton, P. (2008). An eye tracking study of how font size and type influence online reading. In People and computers XXII : culture, creativity, interaction : proceedings of HCI 2008, the 22nd British HCI Group annual conference (Vol. 2). Presented at the 22nd British HCI Group annual conference, Liverpool John Moores University, UK: British Computer Society. Retrieved from





This experiment was developed to respond to the following question: "In order to maximize online reading performance and comprehension, how should a designer choose typographical variables such as font size and font type?" (p. 1). While several previous studies have been conducted to attempt to determine visual response, these studies did not bring together online reading with modern eye tracking. Beymer et al. (2008) developed an eye tracking study to measure online readers'' responses to the variables of font size and font type (specifically, serif vs. sans serif).




The researchers collected data from 132 participants (employees for a major computer company) for each typographical issue (between-subjects design); a base set of 114 had usable eye tracking data. The authors stated that they had a good distribution of ages and gender in the base set: 74 male, 40 female; 2 in their 20s, 63 in their 30s, 28 in their 40s, 18 in their 50s, and 3 over 60. The researchers chose 82 (method not stated) for each final sample.


Participants were presented with a single screen story with black text on a white background and a single graphic (not wrapped in the text, but aligned to the right). Line breaks were designed to ensure that they would occur at the same location in each story. For the font size task, the font type was Verdana; for the font type task, font size was 12pt. Participants were instructed to read the stories for comprehension, then to complete a questionnaire that asked for the participant''s name, first language, and self-estimate of web use.




The researchers found that the greatest influence on reading time was whether or not the participant''s first language was English. While Beymer et al. did note that 14pt font was read 12.6% faster than 10 pt and Georgia (serif) was read 7.9% faster than Helvetica (sans serif), they explained that these percentages were not statistically significant. A significance was found, however, in the rates of re-reading among the first-language English thirty-somethings (32%) versus the first-language English fifty-somethings (24%), although the retention rate was the same.




The researchers acknowledge that there is a lack of significant difference in reading time across font sizes, but they note that the participants'' response to the 10 pt font "was fairly negative" (4). It is their belief that the best choice would be 12 pt font.


Suggestions for Further Research:


Beymer et al. argue that their findings are very much in line with those of their predecessors, which in some ways sets up their research as the "confirmation study." That being said, I find issues with the  sample which would point to the need for further research.




The idea of this study is an excellent one, especially for technical communicators who are developing online help guides, web sites, and the like. However, I would be rather hesitant to accept the findings and interpretations without further research. First, I was surprised that this recent study references participants as "subjects," which holds some negative connotations as to the researcher/researched relationship. I disagree with the researchers'' claim as to a "good" distribution, as the sample was nearly 2/3 male, the age distribution corresponds to neither population nor online users, and there is no indication as to the language systems that the readers did come out of if it were not English, so it is impossible to determine how many of the total sample were non-native readers of English.  The explanation of the sample group is one that I found rather confusing and I''m still trying to puzzle through their math; although the design is between-subjects for the font size, it appears that the same group completed the font style study, but that is not made clear. Furthermore, all of the participants were employees of a specific computer company; organizations have a way of training/organizing to normalize certain reading processes and environments (such as font size and choice) to keep with corporate branding.  Finally, Beymer et al. continually reference the font as "san serif."