Does Haptic Feedback Improve Learning?

Haptic training is widely assumed to be a better way to acquire a skill that involves force feedback, and haptic virtual reality training systems lean on that assumption. The question I wanted answered was narrower: does touching the simulation change what you know about it afterwards?

The experiment used a virtual reality nasendoscopy simulation — a sinus model with a Novint Falcon haptic device. Two groups of participants explored it: one with vision only, one with vision plus force feedback. We then tested their spatial knowledge of the anatomy. The comparison found no significant difference between the two groups.

Coloured dots showing the relative positions of markers inside the sinus simulation — Marker positions inside the simulation, as the participants saw them.

Screenshot of the brown marker inside the sinus simulation — One of the markers in place, inside the sinus.

Frontal CT view of the sinuses with the marker positions overlaid — The same marker positions on a frontal CT view of the sinuses.

A comparison of two groups of participants (visual versus visuo-haptic) revealed no significant differences in their spatial knowledge of the simulation. The findings are discussed with reference to potential variables which may affect spatial learning such as cognitive load.

I think the negative result is the useful part. Force feedback demonstrably helps with tasks where the force is the information — cutting, probing, palpating. It does not automatically help you build a mental map of where things are, and the extra cognitive load of operating the device may well cancel out whatever the haptics add. That is a design constraint for anyone building a training simulator, not a reason to drop haptics.

The paper — Does Haptic Feedback Improve Learning and Recall of Spatial Information? A Study Using a Virtual Reality Nasendoscopy Simulation, by G.S. Ruthenbeck, M. Tlauka and A. Tan — was presented at Eurohaptics in London, 2016. The article is on ResearchGate.

A follow-on study with Michael Tlauka compared performance with and without force feedback over repeated attempts rather than a single session. Learning happened in fewer repetitions when force feedback was present. I will link it here once it is published.