Do we misjudge how far we have turned our bodies?
When turned on a chair with eyes closed, people consistently felt they had rotated further than they really had, and adding sound landmarks did not correct this.
Source
How much I moved: Robust biases in self-rotation perception
Study at a glance
- Design
- Human experiment — Two within-subject psychophysics experiments: blindfolded participants passively rotated in the yaw plane on a motion chair judged whether they ended nearer 0° or 90°, with vestibular-only versus vestibular-plus-auditory-landmark trials (landmarks automatic in Exp 1, self-triggered in Exp 2); adaptive Psi procedure and fitted psychometric functions
- N
- N=32 · 16 participants in each experiment; after exclusions, 15 (bias) and 13 (JND) analysed in Experiment 1, and 11 (bias) and 12 (JND) in Experiment 2
- Population
- Healthy adults (mean ages 27.1 and 25.6 years in the two experiments) with no neurological, hearing or vestibular disorders, in Genoa, Italy
- Outcome
- Bias in perceived rotation amplitude (45° minus PSE) and precision (just noticeable difference) by condition and direction
Structured fields used in claim comparison tables when every cited study has a complete layer.
What they did
Blindfolded adults sat in a motorized chair that rotated them smoothly left or right by different amounts over 3 seconds. After each turn they judged whether they felt closer to their starting direction or to a point 90° to the side, which reveals which rotation feels like exactly halfway (45°). On some trials, virtual office sounds fixed in space were played before and after the turn as landmarks. In Experiment 1 these sounds played automatically; in Experiment 2 a new group of 16 people triggered the sounds themselves with a button press.
What they found
In both experiments and in every condition, the rotation people perceived as 45° was physically smaller than 45°, meaning they overestimated how far they had turned. This bias was the same for clockwise and counterclockwise turns. Adding sound landmarks, whether automatic or self-triggered, did not significantly change the bias or the precision of judgements (in Experiment 2 the bias difference was not significant, p = 0.07).
The limits
What it doesn't show
Samples were small (16 per experiment) and several people were excluded as outliers or for technical problems, leaving as few as 11 in some analyses. The sound landmarks were only played before and after the rotation, never during it, so the study cannot rule out that moving or simultaneous sounds would change the bias, which the authors flag as a design limitation. Only one slow rotation frequency (0.33 Hz) was tested, so the Bayesian idea that faster, more reliable rotations would show less bias remains untested. Proposed explanations, such as a prior favouring straight ahead or temporo-parietal brain involvement, were not measured directly.
Key terms
- Vestibular system
- The inner-ear sense organs (semicircular canals and otoliths) that detect head rotation and linear acceleration.
- Yaw rotation
- Turning around a vertical axis, like spinning on an office chair, as opposed to tilting forward or sideways.
- Point of subjective equality (PSE)
- The physical stimulus value that a person perceives as equal to a reference, here the rotation that feels like exactly 45°.
- Just noticeable difference (JND)
- A measure of how precise or variable judgements are; a smaller JND means finer discrimination.
- Bayesian prior
- Built-in expectations from past experience that are combined with noisy sensory input, pulling perception toward what is usually true.
- Psychometric function
- A curve relating stimulus strength to the proportion of a particular response, used to estimate PSE and JND.
Flashcards
0 of 11 answers reviewed
Research intelligence for this paper
See its role on concept claims, tensions it is part of, placement history, and related discoveries.
Quiz yourself
What systematic error did participants show when judging passive rotations?
Common questions
Why would the brain overestimate rotations on purpose?
The authors suggest it may be functional: exaggerating deviations from straight ahead makes small turns easy to notice and correct, helping people keep a straight path when walking, at the cost of accuracy.
Why did the sounds not help?
The authors propose that vestibular information was more reliable than the sound cues, so it dominated the estimate. Also, sounds were only given before and after the turn, not while moving.
How can a yes/no-style judgement measure a bias in degrees?
By testing many rotation sizes and fitting a psychometric curve, the researchers found the rotation that people judged 'closer to 90°' half the time, which is the rotation that feels like 45°. Comparing that to a true 45° gives the bias.
More on Perception