Week 19: Choosing Typography

Week 19

Typography research, trials and working decision

Research question. What typographic system can help an autistic adult traveller find the changed gate, distinguish it from the previous gate and choose a next step, while giving Grounded a recognisable adult identity?

Typography is both a brand decision and part of the wayfinding system. In an airport, the person may be looking between a phone, overhead signs, other travellers and announcements. The screen therefore needs an obvious reading order, recognisable numerals and text that remains usable when someone enlarges it or listens through accessibility features. Maia’s Critical Report gives the project an audience and a reason to investigate these needs; its ten self-declared autistic respondents do not establish a preferred typeface or prove that a design works for all autistic travellers. The design decisions below are research-informed hypotheses to test.

1 · What the audience needs from type

For this app, the most important typographic task is not to make every line equally prominent. A traveller needs to identify the current destination first, retain the previous gate as context, then understand the available action. For example:

A04 — the current gate; visually dominant and easy to distinguish.
Previous gate B12 — still readable, but clearly secondary.
Show next step — a plain action label, visually separate from decorative elements.
Supporting details — flight, source or illustrative-data notes; available without competing with the change.

This hierarchy responds to the product’s audience and job: it helps someone keep control of the next decision without assuming how they feel or what kind of help they want. It also creates a useful comparison test. If A04, B12 and the action are all bold, all caps or equally colourful, type is no longer helping the traveller decide what to read first.

The AASPIRE web-accessibility guidelines were developed with autistic end-user involvement and evaluated through usability work. They include plain sans-serif text, user-controlled font size, consistent navigation and read-aloud options. These are useful starting points, not rules that every autistic person prefers the same typeface. A 2025 study of adults with autism, ADHD and developmental disabilities reported that Arial was the most preferred of the tested font styles for 83.8% of its autism group, while 64.7% ranked 16-point as their preferred size. These are stated preferences within that study, not proof of better comprehension or a universal autistic preference. They make a plain sans-serif and 16-point body text useful comparison points for Grounded, alongside the iOS system baseline. AASPIRE guidelines · Adults with neurodevelopmental disorders and information presentation

2 · Overlapping access needs

Autism is not a single reading profile. Some autistic people also have dyslexia, ADHD, low vision, motor or learning disabilities, or other health conditions; others do not. A systematic review and meta-analysis of children and adults with autism included 340 studies and around 590,000 participants. Its pooled point-prevalence estimates included ADHD at 37% and anxiety disorders at 35%, but estimates varied by age and by whether the studies used clinical samples or population/registry data. These figures are context, not predictions about any individual traveller. They support designing for adjustable access needs without making a diagnosis or assuming a particular support need. Lai et al., systematic review and meta-analysis

This is an intersection-aware approach: an adult traveller with autism may also benefit from larger text, stronger contrast, more line spacing, reduced reading density, screen-reader support or an option to hear the update. It does not mean Grounded is being designed specifically for every one of those disabilities. It means a basic travel task should not become inaccessible because the designer treated autistic people as a uniform group or designed the interface around only one sensory or reading style.

There is also no strong basis here for declaring one typeface “autism-friendly.” AASPIRE provides guidance grounded in autistic participation, but the best option still needs to be checked with the intended users and task. Research on specialist dyslexia typefaces also cautions against treating a font label as an accessibility solution. A 2026 meta-analysis focused on children with dyslexia found no consistent or reliable effect of dyslexia-friendly typefaces on reading speed or accuracy. This is evidence about children with dyslexia, not autistic adults; it is useful here as a reminder that specialised-looking fonts still need task-specific testing. The practical priority is readable letterforms, adequate size, spacing control, text clarity and user choice. PubMed: Does font improve reading in dyslexic children?

3 · Research references and what they contribute

The earlier brand research already compared airport information products with wellbeing and companion-led apps. I am using those as different types of reference:

Apple system interfaces offer a benchmark for typographic restraint, hierarchy and platform behaviour. Apple describes San Francisco as its system type family, designed to adapt to interface roles, sizes and languages. Dynamic Type lets people choose a preferred text size across apps. This is a strong baseline for the operational parts of Grounded because the alert needs to behave like a usable iOS interface. Apple Fonts · Apple Dynamic Type
Headspace, Calm and Finch show how a wellness brand can use typography as part of a distinctive, friendly identity alongside illustration and colour. The reference board on the brand-research page collects examples from these kinds of interfaces. I can study how their headings, labels and body text create tone, but their visual personality is not evidence that a typeface is accessible for Grounded’s airport task.
The Grounded app itself needs more than a single mood. Its identity is warm, playful and adult, while the gate update must remain direct. A display style can express the name in a brand context, but the reading interface should not inherit decorative lettering if that makes A04, B12 or longer instructions harder to scan.

The reference board is a visual source, not a typography test. Its useful question is how each example separates identity from information: what is set as a heading, what stays quiet, and whether a character or colour becomes more visually prominent than the content. The complete app examples are not direct precedents for an airport alert, so I am carrying forward principles rather than copying their layouts or making claims about their users.

4 · Typography options for Grounded

I compared three system directions using the same brand name, slogan and gate-change content. The type families below are candidates for further specimen trials; none has been evaluated with travellers.

| Direction | What it could contribute | Main concern | Status |
| --- | --- | --- | --- |
| Native iOS system typography (SF Pro through system styles) | Platform-consistent roles, system scaling and support for Dynamic Type; keeps the gate facts practical and familiar. | It may give the wordmark less distinctiveness if the brand relies on type alone. Colour, illustration and a carefully composed lock-up can provide identity without changing the reading font. | Preferred for the app UI. |
| Open Sans | A custom sans-serif candidate to compare against the iOS baseline for body text and numerals. It has broad language support and is available through Adobe Fonts. | A custom face still needs deliberate Dynamic Type implementation, fallback and multilingual checks. Availability does not demonstrate accessibility. | Comparison candidate; not selected. |
| Proxima Nova or InterFace for the interface | Custom-family alternatives to compare for the same body, button and gate content. | Adding a custom family throughout the app gives another variable to maintain and test, without current audience evidence that it improves the task. | Comparison candidates; not selected. |
| Omnes for display with system UI text | A candidate for a more distinctive name or large editorial heading, while leaving task text in the system family. | A second family can make the brand expressive, but it could create a mismatch between promotional identity and an urgent update. Its numerals and enlarged-size behaviour still need inspection. | Display-only experiment; not selected for interface facts. |

The candidate families are linked to their Adobe Fonts records: Open Sans, Proxima Nova, InterFace and Omnes. Their inclusion records a shortlist, not a claim that Adobe classifies them as accessible. A font catalogue or specimen cannot tell me whether an autistic adult will find a particular screen easy to use.

What the wordmark and layout trials suggest

The image below holds the wording and palette direction steady while comparing sentence case, uppercase and lowercase. This helps me judge the brand lock-up separately from the in-app hierarchy.

Sentence case — “Grounded.” This is the strongest current arrangement. It is easy to pronounce and recognise as a name, and its scale leaves room for the full slogan, “Different needs. One connected journey.” It also balances the adult brand with warmth without requiring a playful font. This is a design judgement from the visual trial, not a participant finding.

Uppercase — “GROUNDED.” The capitals create a firmer, more sign-like presence and connect visually with airport wayfinding. The longer all-caps word spreads across the lock-up and makes the identity feel more formal. It is useful as a small eyebrow or label, but I do not want every app label in capitals.

Lowercase — “grounded.” This is a softer treatment, but it risks leaning too far towards a wellness or lifestyle brand. The project is a travel-information service with optional support; the name should not make it sound like therapy or imply that it can make someone feel grounded.

The same contrast appears in the interface specimen. A04 carries the largest typographic emphasis; “Gate changed” names the event; “Previous gate B12” stays visible as context; and the action is written in full. The earlier font specimen image used a generic static rendering, so it did not show an approved font or prove how iOS text would scale. The new board replaces it as a brand and hierarchy trial. It is still an illustration of a typographic direction, not a finished app screen or a real-device preview.

5 · Accessibility and hierarchy specification to carry into the prototype

Apple’s accessibility guidance gives iOS a 17-point default text size and an 11-point minimum for custom text styles; it recommends supporting enlargement to at least 200% and adapting the layout so text does not clip or overlap. For a gate-change screen, the minimum is not the target: the primary task details need comfortable size and weight because the user may be reading while moving or looking between sources. Apple accessibility guidance

| Role | Starting point for the prototype | Why it has this role |
| --- | --- | --- |
| Current gate number | 34–40 pt equivalent, bold/semibold; scales with Dynamic Type | The changed destination is the key travel fact. Check similar characters and numerals such as A04, B12, 0/O and 1/I. |
| Change title | 28–34 pt equivalent, semibold | Names the event before the traveller reads the detail or character. |
| Main action and instructions | 17 pt equivalent, semibold for the button and regular for the instruction | Matches Apple’s iOS default text size and keeps action wording explicit. |
| Previous gate and other essential context | 15–17 pt equivalent, regular or medium | Secondary in hierarchy, but still essential information and should not fade into decorative colour. |
| Supporting source or explanatory notes | Aim for 13–15 pt equivalent; avoid making the essential task depend on tiny type | Keeps secondary information available while preserving screen space. Test the smallest setting on a device and with enlarged text. |

These are starting roles, not fixed pixel measurements or a claim that this specific scale is proven for autistic adults. In implementation, use semantic iOS text styles and Dynamic Type rather than baking the mockup sizes into an image. Apple’s current HIG recommends using system styles, keeping a small set of consistent type styles, avoiding very thin weights at small sizes and maintaining hierarchy when text is enlarged. The system font is designed to work with platform text behaviour; a static image cannot show that behaviour. Apple typography guidance · Apple font overview

Weight, case and emphasis. Use regular for paragraphs and information that is read in sequence; semibold for a short title, the current gate and the action; reserve bold for the single most important value if semibold is not sufficiently distinct. Do not set every row in bold. Keep italics for occasional secondary editorial emphasis, not for gate numbers, action labels or status. Use sentence case for the title and action; reserve all-caps with letter-spacing for short category labels only. Type alone should establish hierarchy before colour or mascot decoration is added.

Line and letter spacing. WCAG 2.2 Text Spacing does not require every screen to use one specific spacing setting. It requires web content to remain usable when a reader overrides line height to 1.5 times the font size, paragraph spacing to 2 times, letter spacing to 0.12 times and word spacing to 0.16 times. This is a useful resilience check for Grounded’s longer instructions and route details: enlarge and space the text, then confirm nothing clips, overlaps or disappears. It is not a blanket instruction to use those values as the default for every short UI label. W3C WCAG 2.2: Text Spacing

Contrast and colour. Typography still depends on the selected light and dark surfaces. WCAG 2.2 gives a 4.5:1 contrast minimum for normal text and 3:1 for large text. The current colour roles are warm off-white and charcoal surfaces with lilac route framing and lime character/action accents; the exact tokens remain provisional. Essential text should use a foreground/background pairing that passes the relevant contrast calculation, while lilac and lime stay supplementary cues rather than carrying the only meaning. Contrast checks need to be repeated after exact colour values and font weights are chosen. W3C WCAG 2.2: Contrast Minimum · GDE751: Week 19: Brand Colours.

VoiceOver and audio. The gate number, change, previous destination and next-step control must exist as selectable interface text, not just as text drawn into an image. VoiceOver should announce the same information in the same order that sighted users see it. The separate audio option should read a concise update in a neutral, unhurried voice and allow replay or stop; it is an alternative access mode, not automatic reassurance. A person who chooses audio should not lose the visual message, and a person who mutes the app should still have the whole task available.

6 · Working type decision and next checks

Working decision: use Apple’s native system typography, retrieved through system text styles, for the app interface. Set Grounded in sentence case with a semibold/bold weight, and pair it with “Different needs. One connected journey.” in regular weight. Use the system family for gate facts, buttons, instructions and supporting labels; do not add a second display font to the operational screen at this stage. This is a provisional design choice based on the iOS target, the task hierarchy and Apple’s native scaling behaviour. It is not a finding that autistic travellers prefer SF Pro, and it does not lock exact sizes, colours or wordmark artwork.

Before treating the typography as final, place the same B12-to-A04 event in the prototype and compare the native system direction against one custom sans candidate. Check default and largest accessibility text sizes, light and dark appearances, long labels, number/letter ambiguity, increased spacing and VoiceOver order. Ask autistic adult travellers to find the current gate, identify the previous gate and locate the next action. Record different preferences and task errors rather than averaging away useful variation. This is the point at which the direction can be confirmed or revised.