Gesture Elicitation Study

Description

Shows participants the effect of each command and asks which gesture they would use, then finds consistent pairings across people to build a user-defined gesture set.

Rather than designers choosing gestures alone, elicitation draws on the intuitions of intended users. In the feature-to-gesture form, participants see a command's effect (the 'referent') and propose a gesture to trigger it; in the reverse form, a gesture is demonstrated and participants say what it should do. Proposals are grouped into identical gestures and patterns across participants show which pairings are intuitive. In the formalised academic version, agreement per referent is quantified with an agreement score, so that high-agreement gestures become the default set and conflicts are resolved explicitly. Saffer then filters candidates against practical criteria: sensor capability, task steps, limits of the body, and matching gesture complexity to task complexity.

Attributes

Loss type
Active
Stage
Design
Data
Mixed
Effort
Days

Steps

  1. List the commands or features (referents) the gestural system must support, typically 10 to 30.
  2. Show each referent's effect to participants (often 12 to 20) in randomised order and ask them to perform the gesture they would use, ideally thinking aloud.
  3. Optionally run the reverse task: demonstrate candidate gestures and ask what each should trigger.
  4. Group identical proposals per referent and calculate an agreement score per referent.
  5. Assign the highest-agreement gesture to each referent, resolving conflicts where one gesture is proposed for several commands.
  6. Filter the set against sensor capability, task steps, body limits and complexity matching, then validate it in a usability test.

Sources

  • Designing Gestural Interfaces (Saffer, 2008)

Also known as

  • User-Defined Gesture Elicitation
  • Using people to determine the appropriate gesture
  • Participant-chosen gestures
  • Gesture-to-feature matching
  • Guessability study

Categories

  • Design
  • Participatory, Design & Futures