Designing fonts with low-vision readers in mind: A reading acuity experiment
A lab-based experiment developed to empirically improve and test the distance-reading legibility of new typeface, tested on visually impaired readers.
Most new typefaces are designed to stand out – to explore the boundaries of new expression, to help create distinctive visual identities and to distinguish themselves from the competition.
However, there are situations where type needs to do its basic job: to be legible. Think of public signage, information interfaces, and health and safety information that, first and foremost, needs to be easy to read.
Inclusivity in type design
Typography, like other areas of user experience design, is currently undergoing a noticeable shift towards more inclusive design solutions. The focus is on creating typefaces that go beyond traditional legibility considerations, such as open counters or large x-height, and also address a wide range of user abilities and preferences, across age groups and individual needs.
An example of this is Apple’s expanding list of accessibility features across its devices, which already enable all UI text to be adjusted in width and weight, or Microsoft’s efforts, which range from an entire reading research department to its commercial solutions, such as its so-called Immersive Reader software, which allows users to customise their reading experience from within Microsoft’s apps.
At Typotheque, we’ve addressed inclusivity primarily by designing fonts for digitally disadvantaged languages. For example, we’ve designed fonts for the Indigenous communities of North America, which for the first time correctly render the sounds present in the local languages. We have also worked in India, with the aim of understanding user preferences, which in turn has allowed us to design fonts better suited to readers of Hindi, Marathi and Nepali.
The research you’re now reading addresses a different group of people – those with declining reading acuity. The present article describes a case study we developed to empirically improve and test the distance-reading legibility of our latest typeface, Zed.
Coincidentally with this release, Zed typeface has just been implemented for the way-finding system of the French 15-20 National Ophthalmology Hospital in Paris, planned and designed by Integral Designers, which is where this study also took place. For this research, we benefited from the support and expertise of Dr Sofie Beier, professor and researcher at the Centre for Visibility Design in Copenhagen.
What is accessible type?
Creating accessible type can be compared to the formulation of a pharmaceutical drug to treat a physical ailment. It is simply unrealistic to look for a universal solution that can address the needs of every patient, regardless of their disease; nor would you expect the same drug to work equally well for all patients with the same condition. Adjusting the dosage and combining it with other medicines is one option to increase the effectiveness of the treatment.
And something similar occurs in typography. While some might claim to have found a silver bullet, it’s very unlikely that their solution will really solve each and every problem.
It is crucial to our design process that we gain an understanding of who the intended reader is and the specific ways in which they will use the typeface we are developing.
Identifying our reader groups
For the purposes of this research, our target audience comprised readers of Latin script with a range of visual impairments. We focused on three different kinds of impairment that make up the population of ‘low-vision readers’, a well-established term in vision research. These readers had central vision loss, peripheral vision loss (tunnel vision) and/or severe focusing problems.
It was anticipated that each subject’s reading performance would differ slightly, in terms of how easy the text was to read, and we could then measure and comparatively rate the versions studied, across the participant group. In psychology, this is coined within-subjects experimental design. We also invited users with normal vision to take part, in order to form a more accurate picture of how our new type works.
Our experiment needed to evaluate various adjustments that could be made at the micro-typography level in order to maximise performance in distance reading.
Of all the attributes of text, the one that universally makes it easier to read is its displayed size. It's simple: larger text is easier to read than smaller text. However, size poses an inherent physical limitation when it comes to signage, where text needs to adapt to the constraints of the space. The adjustments needed to be made at the character level. Determining which could optimise reading in the distance was the first step to take.
Adjustments to improve legibility
After reviewing previous literature, we decided to focus our efforts on three typographic adjustments: letter width, letter weight and tracking.
In particular, a large body of research strongly supports the theory that width adjustments can improve legibility at critical point size. These benefits have been observed not only in healthy participants, but also in readers with the underlying conditions of interest in this research. However, the literature also suggests that these adjustments affect each reader in a slightly different way. This knowledge has led to a school of thought, supported by leading voices in accessibility design, that one size does not fit all. Hence, tools have been developed that enable type to be adapted according to the needs of the reader, such as Microsoft’s aforementioned Immersive Reader tool and web extensions such as Eye-Able for Chrome.
Semi-adaptive way-finding and the case of the 15‑20 Hospital
Although optimal for the reader, ultra-personalised type design is unrealistic in many cases, for example, for physical environments. A way-finding system can’t adapt its type to the individual. But it can certainly predict the needs of the reader, especially in a hospital where the majority of readers have a predictable type of low vision.
We realised early on that the layout of the hospital physically separates patients with different vision issues. For example, patients with glaucoma (who usually experience tunnel vision) will normally use a completely different area from those with keratoconus (who have extremely blurred vision).
We proposed a passive adaptation typographic system, which very slightly changes the typographic adjustments of all signs in each area according to the needs of its particular readers.
The experiment we conducted therefore focused on devising the optimal settings for each of the three groups of interest:
1 Patients with blurred vision – Seen in patients with keratoconus, irregular cornea, presbyopia, severe myopia and patients using orthokeratology.
2 Patients with central vision loss – Seen in patients with macular degeneration, with damage in the centre of the visual range.
3 Patients with peripheral vision loss – Seen in patients with glaucoma and retinitis, with damage in their visual periphery.
Research methods
Designing the materials
We created six variations of Zed Text, ranging in widths, with the weight and character spacing being adjusted for each version (see below). To achieve this, we used the data from a pilot study that was conducted before the start of the data collection for the main study, as well as practitioner knowledge.
The different variations of the Zed typeface were measured using a lowercase letter ‘n’. For the narrowest variation (100), the vertical/horizontal proportion was 0.93, and for the widest (200), 1.21.
Roberto Arista, Typotheque’s in-house font engineer, developed bespoke software to run the experiment on a macOS computer. The software ran on a 27-inch display, placed 2.5 metres away from the reader.
The participant had a controller within reach, which they used to interact with the software during the experiment. The room was a quiet space with constant artificial lighting, and the accompanying medic sat unobtrusively as the participant completed the experiment.
Demographic information was collected from each participant before the experiment began. This consisted of age, gender, attention deficit diagnosis and language proficiency.
All data was processed in accordance with the European General Data Protection Regulation (GDPR). The participant was then guided through a series of instruction screens by the accompanying medic. After the experiment had been explained, a trial round was used, to check that the participant understood the task.
Experimental design
The idea was to measure the threshold of acuity of each variation of the typeface. That is, the maximum distance at which reading is possible.
The experiment was based on the previously published experiment named Comparing typefaces for airport signs, by Robert Waller. Participants were asked to fixate at the centre of the display. Starting with a point size of 0, a string of letters would grow in size at a constant rate, reaching a maximum size of 250 pt after 15 seconds.
Participants pressed a key as soon as they were able to decode the string of letters on the screen. Two alternating ‘noise masks’ then appeared for 0.5 seconds, to remove afterimage. Following this, the participant was asked to choose the string they had just read from three different but similar-looking options, to validate their answer. You can try the experimental design yourself below:
Each participant completed this task 120 times. This repetition ensured that any variability caused by temporary distractions, environmental influences or user errors was reduced. Repeating the measurements also allowed outliers to be removed for more accurate measurements.
For each of the 120 tasks, the text was shown in one of the six typeface variations being researched, thus 20 times for each typeface variation. The four-letter strings were created manually, including and interchanging one or two easily mistakeable letterforms per string. The order of display was randomised for each participant.
Reaction times were used as the measure-ment of reading acuity. The answers to the similar-looking alternative questions were collected, but the real intention of these was to maintain attention throughout the experiment. The list of strings was shown at random, with no participant being presented with the same order of stimuli as another.
Pilot study
Before starting the data collection, we conducted a pilot with a representative sample of 24 visually impaired participants selected from the hospital’s patient pool. During the pilot we adjusted the typographical variables and improved the general flow of the experiment based on user feedback.
Because the previous way-finding system of the hospital was set in Helvetica, the decision was made to test its efficiency against that of Zed Text. The results showed that Helvetica scores significantly worse across all patient groups compared with any of the studied variations of Zed Text. Because of this, it was decided that Helvetica would be removed from the main study.
Participants and sampling
In total, 55 participants completed the experiment during September and October of 2023. The average age in years was 36.5 for the blurred vision group (N=15), 74.7 for the central vision loss group (N=11), 73.7 for the peripheral vision loss group (N=15) and 39.2 for the control group (N=14). The gender balance of the participants was 22:33, male to female.
The participants of the experiment were hospital patients that had appointments scheduled during the data collection period. Screening was carried out by Dr Marie Lang and Dr Malika Hamrani, who selected patients that fell into each of the four vision groups considered for this study on the basis of diagnostic data. Control participants were hospital staff.
Results
Participants in each group responded differently to each of the versions of Zed Text studied, as seen below. The Y-axes represent the average reaction times to each of the six versions.
After running the appropriate statistical analyses, a strong effect in terms of letter width was found for all test groups.
We observed a plateau in performance, which began when the character width reached a certain point. This point differed depending on the patient group: Version C for the central and peripheral vision loss groups (which has a relative width of 140 within Zed’s width axis); and Version D for patients with blurred vision and healthy readers (which has a width of 160).
This aligns with the results in earlier studies, for example by Dr Sofie Beier and colleagues, and suggests there is an effect exerted by width on the performance of reading in small sizes.
We recommended the above typeface settings to be used in the different areas of the 15-20 Hospital, but these findings go beyond this application, providing organic data on the performance of Zed.
Additionally, this experiment inspired us to design extremely wide masters of Zed Text, which are now publicly available, in order to increase the accessibility of the typeface.
The present study is the first of its kind to investigate the typographic adaptations of a commercial typeface, which was still in development, for people with low vision. The results of this research, in line with previous literature, have guided the design of the 15‑20 Hospital signage system.
We would like to thank Dr Marie Lang and Dr Malika Hamrani (Hôpital National des 15-20, Paris, France), who provided very useful feedback during and after the pilot and organised the collection of data used in the analysis. Additionally, we’re grateful to Benjamin Ribeau, David Thoumazeau, Morgane Pontis and Chloé Herbillon, from Integral Designers (Paris, France), who helped assemble the different parts of this project and assisted during the preparation and roll-out of the experiment at the hospital. Also, thanks to Jaro Kondela and Ondrej Jób, for developing the interactive experiment widget on this page, and Giammarco Gaudenzi who designed and developed interactive chart presenting data of four types of readers. Lastly, many thanks to Dr Sofie Beier, who has disinterestedly provided support and advice during the development and conducting of the experiment.
References
Apple. (2023). Apple previews Live Speech, Personal Voice, and more new accessibility features – Apple. Retrieved from Apple.com
Arditi, A. (2004). Adjustable typography: An approach to enhancing low vision text accessibility. Ergonomics, 47, 469–482. doi:10.1080/0014013031000085680
Arditi, A., Cagenello, R., & Jacobs, B. (1995). Letter Strokewidth, Spacing and Legibility. Vision Science and Its Applications, 1, 324–327.
Arditi, A., & Cho, J. (2005). Serifs and font legibility. Vision Research, 45, 2926–2933. doi:10.1016/j.visres.2005.06.013
Arditi, A., & Cho, J. (2007). Letter case and text legibility in normal and low vision. Vision Research, 47, 2499–2505. doi:10.1016/J.VISRES.2007.06.010
Baron, J., & Thurston, I. (1973). An analysis of the word-superiority effect. Cognitive Psychology, 4, 207–228.
Beier, S., & Larson, K. (2010). Design improvements for frequently misrecognized letters. Information Design Journal, 18. doi:10.1075/idj.18.2.03bei
Beier, S. (2016). Designing legible fonts for distance reading. 75–89.
Beier, S., & Oderkerk, C. A. T. (2019). Smaller visual angles show greater benefit of letter boldness than larger visual angles. Acta Psychologica, 199, 102904. doi:10.1016/J.ACTPSY.2019.102904
Beier, S, Oderkerk, C. A. T., Bay, B., & Larsen, M. (2021). Increased letter spacing and greater letter width improve reading acuity in low vision readers. Information Design Journal, 26, 73–88. doi:10.1075/idj.19033.bei
Beier, Sofie, & Oderkerk, C. A. T. (2022). Closed letter counters impair recognition. Applied Ergonomics, 101, 103709. doi:10.1016/J.APERGO.2022.103709
Sofie, B. (2019). Age-related deficits and their effects on reading. Association Typographique Internationale. Retrieved from YouTube.com
Bernard, J. B., Kumar, G., Junge, J., & Chung, S. T. L. (2013). The effect of letter-stroke boldness on reading speed in central and peripheral vision. Vision Research, 84, 33–42. doi:10.1016/j.visres.2013.03.005
Bernard, J.-B., Aguilar, C., & Castet, E. (2016). A new font, specifically designed for peripheral vision, improves peripheral letter and word recognition, but not eye-mediated reading performance. doi:10.1371/journal.pone.0152506
Bernard, M., Liao, C. H., & Mills, M. (2001). The effects of font type and size on the legibility and reading time of online text by older adults. 175–176. doi:10.1145/634067.634173
Bowma, H. (1971). Visual recognition of isolated lower-case letters. Vision Research, 11, 459–474.
Chung, S. T. L., & Bernard, J. B. (2018). Bolder print does not increase reading speed in people with central vision loss. Vision Research, 153, 98–104. doi:10.1016/j.visres.2018.10.012
Chung, S. T. L. (2014). Size or spacing: Which limits letter recognition in people with age-related macular degeneration? Vision Research, 101, 167–176. doi:10.1016/j.visres.2014.06.015
Darowski, E. S., Helder, E., Zacks, R. T., Hasher, L., & Hambrick, D. Z. (2008). Age-related differences in cognition: the role of distraction control. Neuropsychology, 22, 638–644. doi:10.1037/0894-4105.22.5.638
Foster, J., & Coles, P. (1977). An experimental study of typographic cueing in printed text. Ergonomics, 20, 57–66. doi:10.1080/00140137708931601
Garg, A., & Tsang, S. H. (2023). Retinal dystrophies. Microperimetry and Multimodal Retinal Imaging, 137–142. doi:10.1007/978-3-642-40300-2_14
Gary, G. E. L., Rubin, S., Pelli, D. G., & Schleske, M. M. (1984). Psychophysics of reading – II. low vision.
Geyer, L. R. (1977). Recognition and confusion of the lowercase alphabet. Perception & Psychophysics, 22, 487–490.
Kwon, M. Y., Liu, R., Patel, B. N., & Girkin, C. (2017). Slow reading in glaucoma: Is it due to the shrinking visual span in central vision? Investigative Ophthalmology & Visual Science, 58, 5810. doi:10.1167/IOVS.17-22560
Laubheimer, P. (2017). Typography for glanceable reading: bigger is better. Retrieved from NNGroup.com
Legge, G. E. (2016). Reading digital with low vision. Visible Language, 102–125.
Mansfield, J. S., Legge, G. E., & Bane, M. C. (1996). Psychophysics of reading: font effects in normal and low vision. Investigative Ophthalmology & Visual Science, 37, 1492–1501.
Mcleish, E. (2007). A study of the effect of letter spacing on the reading speed of young readers with low vision. British Journal of Visual Impairment, 25, 133–143. doi:10.1177/0264619607075995
Microsoft. (2023). How to enhance reading instruction: a guide to Immersive Reader for educators | Microsoft EDU. Retrieved from Microsoft.com
Minakata, K., & Beier, S. (2021). The effect of font width on eye movements during reading. Applied Ergonomics, 97. doi:10.1016/j.apergo.2021.103523
Oderkerk, C. A. T., & Beier, S. (2022). Fonts of wider letter shapes improve letter recognition in parafovea and periphery. Ergonomics, 65, 753–761. doi:10.1080/00140139.2021.1991001
Pardo, O. (Submitted for publication). The Ergonomy of Type: Visual performance-guided font design.
Reimer, B., Mehler, B., Dobres, J., Coughlin, J. F., Matteson, S., Gould, D., ... Levantovsky, V. (2014). Assessing the impact of typeface design in a text-rich automotive user interface. Ergonomics, 57, 1643–1658. doi:10.1080/00140139.2014.940000
Rumelhart, D. E., & McClelland, J. L. (1982). An interactive activation model of context effects in letter perception: Part 2. The contextual enhancement effect and some tests and extensions of the model. Psychological Review, Vol. 89, pp. 60–94.
Sawyer, B. D., Dobres, J., Chahine, N., & Reimer, B. (2017). The cost of cool: Typographic style legibility in reading at a Glance. 2017-October, 833–837. doi:10.1177/1541931213601698
Sawyer, B. D., Dobres, J., Chahine, N., & Reimer, B. (2020). The great typography bake-off: comparing legibility at-a-glance. Ergonomics, 63, 391–398. doi:10.1080/00140139.2020.1714748
Sheppard, S. M., Nobles, S. L., Palma, A., Kajfez, S., Jordan, M., Crowley, K., & Beier, S. (2023). One font doesn’t fit all: The influence of digital text personalization on comprehension in child and adolescent readers. Education Sciences, 13. doi:10.3390/educsci13090864
Wallace, S., Bylinskii, Z., Dobres, J., Kerr, B., Berlow, S., Treitman, R., ... Sawyer, B. D. (2022). Towards individuated reading experiences: different fonts increase reading speed for different individuals. ACM Transactions on Computer-Human Interaction, 29. doi:10.1145/3502222
Wilkins, A. J., & Nimmo-Smith, M. I. (1987). The clarity and comfort of printed text. Ergonomics, 30, 1705–1720. doi:10.1080/00140138708966059