Academics · Middle School · Grades 6 to 8

Measure how far
your eyes are off.

Set two lines equal by eye, lock in, and the page shows the truth and your error. Do it with the trick and without, and the difference is the size of the illusion, measured on you.

Open the illusion labWhat you get
5figures to measure, each with a plain twin
3figures with a switch that removes the trick
31matches in the example notebook

The illusion lab

Pick a figure, move the slider until the two parts look equal, then lock in. Nothing is marked right or wrong: the numbers are data about how you see.

Loading the illusion lab. If nothing appears, turn on JavaScript.

Five figures that have fooled people for more than a century

Each one is in the Measure tab with its plain twin. The dates are when each was first described in print.

Franz Müller-Lyer, 1889

Two lines with fins

The line with fins pointing out looks longer than the same line with fins pointing in.

Adolf Fick, 1851

An upright line on a flat one

The upright line looks longer, though both are the same length.

Mario Ponzo, 1911

Two bars between closing rails

The bar nearer the narrow end looks longer.

Hermann Ebbinghaus, 1890s

Two discs, two rings

A disc ringed by small discs looks bigger than its twin ringed by large ones.

Johann Poggendorff, 1860

A line passing behind a bar

The two visible pieces do not seem to line up, though they are one straight line.

Knowing the answer does not switch an illusion off. Measure a figure, read the result, then measure it again: the pull is usually still there. That is the reason researchers treat illusions as clues to how seeing works, and not as mistakes to be corrected.

What you get

Academics · Middle SchoolFree

What you finish with
Your own measurements for up to five illusions, each with a control: the average error with the trick, the average without it and the difference, a chart, the evidence written out in sentences and your explanation. A printed lab report and a spreadsheet of every trial.
Who it is for
Middle school students studying the senses, the brain or how to run a fair test; science teachers who want a lab that needs only a screen; homeschool families; anyone curious why a ruler and the eye disagree.
Format
Browser lab with three parts: Measure, Look closer and Lab notebook. Works by touch, mouse or keyboard on a phone, tablet or computer. Prints a lab report; saves every trial as a .csv.
Access
Free. No account and no sign-in. Trials and notes are saved only in this browser; no names are asked for and nothing is sent to Caveat.
Where it lives
The Middle School space in Caveat Academics.
Support
Questions, a correction or a request: team@caveatacademic.com.
Ringed by small discs, the right-hand disc looks bigger. The dashed lines are what the lab draws after you lock in a match.

How to use it

  1. Pick a figure and read the task above it. Move the slider until the two parts look equal to you.
  2. Press Lock in my match. The dashed lines show the true answer and the dark card gives your error.
  3. Take three trials, then switch to the plain version and take three more.
  4. Open Lab notebook for the averages, the chart and the evidence sentences. Write what you think is going on and print the report.

Open the illusion lab

How the lab measures seeing

A match set by eye, a hidden answer, a control and an average.

How a match becomes a measurement

  • You adjust, the page hides the answer. This is the method of adjustment from early psychophysics: the setting where two things look equal to you is called your point of subjective equality.
  • Error is in percent of the true size, with a sign: minus means too short, too small or too low. Percent does not change with the size of the screen. For the Poggendorff bar the error is the distance off line as a percent of the bar’s width.
  • Each figure has a plain twin with the misleading part taken out. The difference between your two averages, not either one alone, is the size of the illusion for you.
  • Every trial starts from a random setting, sometimes too big and sometimes too small, because where you start nudges where you stop.
  • One trial is noisy. The notebook asks for three or more of each version and says so when an average rests on fewer.

For a teacher or parent

This works as a 30-minute lab for one student or a class set of devices. Have students predict first: will the plain version be closer to zero? Then collect three trials of each version for one figure before moving on.

Good follow-up questions: Did everyone’s error go the same way? Whose effect was largest? Does the effect shrink with practice? A class can pool the CSV files and compare averages.

The explanations on the page are worded as ideas that researchers still test. That is deliberate: none of these five figures has one settled explanation.

Sources

  1. Gustav Fechner, Elemente der Psychophysik (1860): the method of adjustment and the idea of measuring a sensation by the setting at which two things appear equal.
  2. Franz Carl Müller-Lyer, “Optische Urteilstäuschungen,” Archiv für Anatomie und Physiologie, Physiologische Abteilung, Supplement-Band (1889), 263–270. Adolf Fick’s 1851 dissertation on the upright-and-flat illusion. Johann Zöllner’s 1860 report of the figure he credited to J. C. Poggendorff. Mario Ponzo (1911). Edward Titchener, Experimental Psychology: A Manual of Laboratory Practice (1901), which brought the figure Hermann Ebbinghaus had described to English-speaking laboratories.
  3. Richard L. Gregory, “Distortion of visual space as inappropriate constancy scaling,” Nature 199 (1963), 678–680, for the depth-scaling account of the Müller-Lyer and Ponzo figures; Marshall Segall, Donald Campbell and Melville Herskovits, The Influence of Culture on Visual Perception (1966), on how much the size of these effects differs between groups of people.
  4. Richard L. Gregory and Priscilla Heard, “Border locking and the Café Wall illusion,” Perception 8 (1979), 365–380. Gaetano Kanizsa, “Margini quasi-percettivi in campi con stimolazione omogenea,” Rivista di Psicologia 49 (1955), 7–30.
  5. Next Generation Science Standards, MS-LS1-8 (sensory receptors respond to stimuli by sending messages to the brain) and the science practices of planning an investigation and analyzing data.
  6. The true sizes, the error rule and the example averages are computed by a reference written for this page (guide/suite/oip_ref.py) and compared with the page’s own arithmetic by its tests.

This is a perception lab, not a vision screening: it says nothing about eyesight or health. Screens differ, and viewing distance, tilt and glare all add noise, so compare your own trials with each other before comparing with anyone else. Practice with the ideas, not a full course on perception. Support: team@caveatacademic.com.