Working research, shown honestly. See what is verified and what is not →Working research · inspect evidence →
Sign First

Learner experience

Why “ages 5+” cannot mean one AI interface

A beginning reader and a teenager should not perform the same confirmation work. Age changes presentation and support—not intelligence or signer authority.

By Sign First Research6 minute read
Four differently sized interfaces orbit a common center
Age-adaptive concept art; it does not depict a tested child interface.

Direct answer

What the evidence supports

The academic task can remain the same while the interface changes how much text, explanation, choice, and provenance it presents. Sign First currently implements four presentation targets; none has yet been validated with children.

Decision trace

How can the interface change with age without taking authority away from the learner?

Implemented Sign First presentation targets · cited Deaf literacy and AI studies

Observed

Transcript review adds its own access burden.

Dense written-English debugging can turn an ASL access path into another language task.

In brief

Three things to keep straight

  1. Reading load and decision density should change across developmental stages.
  2. The learner’s right to reject a candidate must not shrink for younger users.
  3. Age bands are hypotheses to test, not evidence of usability.
01

The same confirmation screen is not neutral

If every signer must read and debug a dense English transcript, the access tool creates additional written-language work. Research with Deaf and hard-of-hearing users of text AI reports both real benefits and challenges involving ASL support and Deaf cultural understanding.1

Research on LLM-generated learning questions likewise points to potential benefits while emphasizing accessibility across a diverse Deaf and hard-of-hearing population.2

02

Four implemented targets, zero child outcome claims

Sign First currently defines visual-first ages 5–7, guided ages 8–10, independent ages 11–13, and advanced ages 14+. The band changes copy density, step size, and explanation—not the underlying academic meaning.4

The age range is selected rather than inferred from a birth date. The current system does not establish that these bands are optimal, understandable, or effective for Deaf children.4

03

What a real age-specific study must measure

The next evidence needs to examine comprehension, correction success, fatigue, attention, trust, ability to stop, and whether the AI answer supports the actual lesson. Accuracy alone would miss much of the experience.

The Clerc Center’s reading principles reinforce why visual attention, keeping ASL and English visible, following the child’s lead, making implicit meaning explicit, and connecting concepts matter in literacy interactions.3

Questions answered

The short version

Does Sign First collect a child’s exact age?

The current interface uses a selected presentation band and states that it does not need an exact birth date.4

Are the four age modes research-validated?

No. They are implemented design targets that make future Deaf-led evaluation concrete.

Should younger students have less control?

No. The interface can simplify language and choices while preserving the same right to reject, repair, or stop.

Source register

What this article relies on

  1. “We do use it, but not how hearing people think”ACM CHI EA 2025 ↗

    Mixed-method study of 80 Deaf and hard-of-hearing participants plus 11 interviews.

  2. “Real Learner Data Matters”CHI 2025 research preprint ↗

    Prototype study on LLM-generated questions for Deaf and hard-of-hearing learners.

  3. 15 Principles for Reading to Deaf ChildrenLaurent Clerc National Deaf Education Center ↗

    Practice guidance based on research into how Deaf adults read with Deaf children.

  4. Evidence registerSign First ↗

    Public, bounded record of internal tests, failures, and current product limits.