Embedded Figures Test

Kurt Gottschaldt published the first of these figures in 1926, and he was careful about one thing above all. A shape does not hide because there is clutter near it. It hides because the lines that make it carry on past its corners.

Look at a corner of any shape drawn on its own. Two lines meet there and both of them stop. That full stop is what the eye uses to decide a shape is present.

Every edge of the hidden shape on this page is drawn as a line running the full width of the plate. So each corner of it is a crossing rather than a full stop, and there is nothing there to catch your eye.

The other lines are chords. Each one is drawn between two points of the grid and cut off at the frame, so it has no loose end either. A few areas are shaded in as well, and every shaded triangle is cut out by three lines that were already there.

That last rule matters more than it looks. The eye breaks a picture into areas before it looks for shapes, and none of the areas here is the shape you want. A shaded patch with an edge of its own would put a full stop back on the plate.

Witkin's version of this test hands you a stylus. The simple figure is shown, then taken away, and you trace its outline inside the busy one while somebody times you.

That measurement holds three things at once. There is the search, there is the drawing, and there is keeping a shape in mind that is no longer in front of you.

This page leaves the shape on screen and takes a single tap. What is left in the number is the search, plus the fifth of a second or so that any press costs. Figures for that part are set out in how reaction time changes with age.

One tap per figure is the other half of the design. With unlimited taps a visitor could work across the plate until something turned green, and the number would become a measure of tapping.

A tap that lands outside the shape ends that figure, and the answer is shown before the next one arrives. It is the same bargain the paper version makes when its three minutes run out, brought forward so a run does not take an afternoon.

The headline is the middle time of the figures you found. Half of them took less than that and half took more.

A middle rather than an average, because one figure you nearly gave up on would drag an average around by itself. A stuck figure says something about that figure. It should not rewrite the other nine.

Each figure gets thirty seconds. When they run out the answer is shown and the run moves on, and that figure is not timed at all.

Cutting the run off like that is why the card carries three more numbers rather than one. The share you found tells you how much of the run is standing behind the headline.

The other two split up the figures you did not get. One counts the ones the clock took and the other counts the ones you tapped in the wrong place. Running out of time and being confidently wrong are different results.

Witkin built his test to measure something he called field independence. The idea is that some people lift a shape out of its surroundings easily, and others find the surroundings harder to set aside.

A large literature grew on that idea, and a good deal of it reached well past what a shape in a pattern can hold up. Accounts of personality and of how people get along were hung on a tracing time.

So it is worth saying plainly what this page reports. It reports how long you took to find a shape inside a pattern, on these figures, today.

There is a family likeness to other tasks where one reading of a thing gets in the way of another. The famous one is reading the word blue printed in red ink. How far that result travels is set out in what the Stroop effect actually shows.

Practice helps here, and it helps quickly. The patterns are built by a rule, and once you know the rule you start hunting for crossings instead of corners. That is a better way of doing this task, and it is not a claim about anything else.

Most pages here put your number on a curve and say where it falls. This one shows it on its own, and the reason is worth setting out.

Witkin's manual does carry figures for how long an item takes. Those figures describe a stylus, a trace, a shape held from memory and a three minute limit.

Not one of those four things is on this page. Taking the drawing and the remembering out does not shave a fixed amount off everybody. On the paper version they are much of what separates one person from the next.

The tables also sit inside a manual sold as a kit, so a reader cannot open the primary source and check them. Either problem on its own would be enough to stop us borrowing the curve.

So the page reports your seconds and says there is nothing yet to compare them with. That changes when a thousand runs have been recorded here, and not one run before.