2025 ·Research, Prototyping ·[Lab or course]

Frame by Frame

Pacing studies for digital reading

A study of how paragraph timing — line length, leading, scroll cadence — shifts comprehension and dwell time in long-form reading interfaces.

HCITypographyReading

The question

If reading is paced like film is paced — frames per second, beats per shot — what happens when we treat scroll as a timing instrument rather than a navigation tool?

Method

We ran a [N]-participant within-subjects study comparing [three or four reading conditions]. Each participant read [text source] under each condition and answered a comprehension survey, then a self-report on attention.

Conditions

  1. Baseline — standard scroll, default leading.
  2. Paced — scroll velocity capped at a “reading speed” the system inferred from the first two paragraphs.
  3. Frame-by-frame — text revealed in fixed beats, like subtitles.

Findings

  • Comprehension was [comparable / higher / lower] under “paced” — the win was smaller than expected but consistent.
  • Self-reported attention was significantly higher under paced.
  • Frame-by-frame felt cinematic in pilot but tested poorly: readers reported feeling “controlled.”

The take: pacing helps when it’s a suggestion, not a constraint.

What I’d build next

A reading mode for long essays that nudges scroll cadence without locking it. A subtitle for the act of reading itself.


Sample images, charts, and a link to the working paper would live here.