In one experiment, people who were walking across a college campus were asked by a stranger for directions. During the resulting chat, two men carrying a wooden door passed between the stranger and the subjects. After the door went by, the subjects were asked if they had noticed anything change.

Half of those tested failed to notice that, as the door passed by, the stranger had been substituted with a man who was of different height, of different build and who sounded different. He was also wearing different clothes.

Despite the fact that the subjects had talked to the stranger for 10-15 seconds before the swap, half of them did not detect that, after the passing of the door, they had ended up speaking to a different person. This phenomenon, called change blindness, highlights how we see much less than we think we do.

Working with Christopher Chabris at Harvard University, Simons came up with another demonstration that has now become a classic, based on a videotape of a handful of people playing basketball. They played the tape to subjects and asked them to count the passes made by one of the teams.

Around half failed to spot a woman dressed in a gorilla suit who walked slowly across the scene for nine seconds, even though this hairy interloper had passed between the players and stopped to face the camera and thump her chest.

However, if people were simply asked to view the tape, they noticed the gorilla easily. The effect is so striking that some of them refused to accept they were looking at the same tape and thought that it was a different version of the video, one edited to include the ape.

Prof Richard Wiseman of the University of Hertfordshire recently repeated this experiment before a live audience in London (as part of his Theatre of Science, performed with the author Dr Simon Singh) and found that only 10 per cent of the 400 or so people who saw the show managed to spot the gorilla.