Cognitive Psychology

Inattentional Blindness

A dense display of small shapes, and one kind to count. It takes more effort than it sounds like it will.

A classroom demonstration — not a test of anybody's attention

Learning objective

By the end you should be able to say what pushes noticing rates up and down, and why one demonstration cannot measure a person.

About 20 minutes. Nothing you do here is saved or sent anywhere.

This demonstration works once

Do the task before reading on. Once you know what to look for, the surprise is gone — permanently, not just today.

  1. Answer the short question below. It unlocks the display.
  2. Read the counting instruction, then start. Count as accurately as you can.
  3. Say when you have finished, enter your count, and answer the question that follows.
  4. The same display is then shown again, in full, with everything labelled.
  5. Only then read the debrief.

First, a judgement about the task

You are about to be shown a display of several dozen small shapes — triangles, squares and circles, some solid and some outlined only. You will be asked to count one particular kind, and there will be somewhere between six and ten of them.

How accurate do you expect your count to be?

The counting task

Count carefully

A static display. Nothing moves and nothing flashes.

Your task Choose a variant and press Start.

Set up the run

Each variant is a fixed display, identical every time it is drawn, so a room can be given the same one.

Closer to the classic version. Switch this off for a self-paced run.

Run

Reveals everything without spending a variant. Do not press this if you intend to be a naive observer.

Answer the question above to unlock the display.

Challenge — which change does what?

Four changes somebody might make to this demonstration. For each, decide whether the published literature would lead you to expect more people to notice the unexpected object, fewer, or roughly the same.

Change Expected effect on noticing

What this demonstrates

Attention and awareness are related, not identical

An object can fall on your retina, be fixated, and still fail to be reported. What you are set to look for shapes what gets through, and a task demanding close attention to one property narrows that set. The question is never whether the light reached your eye. It did.

Not reporting is not the same as not seeing

A person who says "no, there was nothing" may have processed the object without being able to report it, may have seen it and forgotten it by the time the question was asked, or may genuinely have had no experience of it at all. The three are very hard to tell apart, and the disagreement about which one inattentional blindness experiments demonstrate is a live one. Calling the phenomenon "blindness" builds in an answer that the method does not supply.

Noticing rates are a property of the design

Four things move them reliably: how demanding the primary task is; how similar the unexpected object is to what you are attending to; how salient it is; and whether you had any expectation that something might appear. Change any of those and the rate changes, which is why a rate from one demonstration describes that demonstration and nothing else.

Why this cannot measure you

A single trial gives a single binary response. There is no reliability to speak of, no norm, and no evidence that noticing on one occasion predicts noticing on another. Whether individual differences in noticing exist at all is contested, and it is certainly not something one display can settle. If you noticed the object, that is not a fact about your attention; if you missed it, neither is that.