Once scale is established in a scene, it keeps biasing size judgments of new objects

What it is

Ten minutes in a land of giants, then an ordinary cup looks like toy tea-set china. Size adaptation is scale, once calibrated by the current scene, being used as a ruler on objects that appear later, even when those objects bring their own true size. The scale that was established sticks. Each glance does not estimate from zero.

Without a referent, people estimate badly. Here a referent already exists, and it will not let go.

Why it happens

Vision gathers stable door heights, step risers, and the visual angle of the hand into a working scale, and later objects are read against that ruler. Adaptation takes time: a few minutes of exposure usually locks the ruler; a new object that appears for an instant is swallowed by the old ruler first, then slowly pulled back by its own cues. Cross-scene is sharper — hard-cut from a minified sandbox to a 1:1 room and doors look huge until a few steps and a hand on a table edge recalibrate.

This is kin to contrast illusions, but the object is the whole ruler rather than two lines side by side. Content that switches scale inside a session without giving time to recalibrate will have people grasp new-level props with the old level’s ruler — empty air, or a bursting grip.

Studying it

Use an adapt–test sequence: several minutes in scale A, cut to a neutral test object (or to scale B), and take an immediate size estimate or grasp. Manipulate exposure duration, the scale ratio of adapt to test, and whether a stretch of bodily activity (walking, touching a referent) is inserted. Read the test-phase shift against baseline, and how long the shift takes to fade.

Verbal estimates and grasps can split: the mouth already says “a normal cup” while the hand still shapes a toy-tea envelope. Take both. Aftereffects from minified-to-1:1 and magnified-to-1:1 run opposite directions; do not collapse them into a yes/no “there was an aftereffect.”

Where it stops holding

If the adapting scene itself has no stable ruler (an abstract noise field), later test objects inherit nothing and the aftereffect is weak. A test object with very strong familiar size (the user’s own hand, a standard door) will overwrite the old ruler quickly. Repeated scale cuts inside a session teach people not to trust any ruler, and judgments retreat to conservative relative comparisons. Aftereffects vary; short exposures are almost invisible in some people. Do not mix adaptation with simultaneous contrast — a giant table parked in the corner of the screen for the whole trial is concurrent contrast, not “once established, it keeps biasing.”

Applying it

• After an environment-scale change in a session, give a buffer with bodily referents (visible hands, a door, a few steps) before any size-dependent action.
• Newly appearing grabbable objects should bring their own familiar size. Do not drop an abstract handle with no referent in the moment after a scale cut.
• If levels of one product differ in scale, do not hard-cut with no transition, and do not schedule precise grasps in those seconds.
• How to check: expose people for a few minutes in a minified scene, cut immediately to a 1:1 standard cup, and have them grasp. If the hand shapes a smaller envelope and misses, the old ruler is still in charge. Insert a short buffer with a visible door and hands, then grasp again; empty grasps should fall.

Related

• Same group: N4.08.1 Without a known-size referent, vision alone cannot judge absolute scale · N4.08.2 A world miniature lets people operate a large environment from a god view · N4.08.3 Extreme deviation from real scale weakens presence · N4.08.4 Scaling the self turns far-field operations into near-field ones · N4.08.6 Virtual scale changes perception of distance and of one's own size · N4.08.7 Inconsistent scale breaks action expectations
• Nearby: N1.01 Immersion and Presence · N2.04 Direct Grab
• Search terms: size adaptation · scale aftereffect · size contrast