Prototype Fidelity Validation Study

Description

Build prototypes of an existing product at several fidelities, run identical trials on each and on the real product, and see which fidelity reproduces real-product results.

Prototype testing assumes the prototype behaves like the eventual product. This study tests that assumption by using an existing product as a datum. Gill and colleagues took the Equinox phone and built a high-fidelity prototype (real mouldings plus interactive electronics and a Flash interface), a low-fidelity foam 'sketch' prototype, and a touchscreen-only software prototype, then ran identical user trials on all and on the real phone. Tangible prototypes behaved like the real phone while the software-only one did not; a picture-on-block version failed because button edges could not be felt. The low-fidelity tangible prototype took about 80 per cent less build time yet matched the real phone better than software. The result is evidence-based guidance on which fidelity dimensions matter for a product class.

Attributes

Loss type
Latent
Effort
Weeks

Steps

  1. Choose an existing product, similar to the one being designed, as the datum.
  2. Build prototypes that vary physical and visual fidelity independently (for example foam with real buttons, screen-only, full mock-up).
  3. Run identical tasks and measures with comparable participants on each prototype and the real product.
  4. Compare each prototype's results (errors, times, ratings) with the real product's.
  5. Derive fidelity guidance, such as medium visual with high physical fidelity, for future prototyping.

Sources

  • Touch IT (Dix et al., 2022)

Also known as

  • Prototype fidelity validation against a real product (Equinox study)
  • Equinox study
  • Physical fidelity study

Categories

  • Research Operations
  • Research Design & Rigor