Although it often suffices, in 
proving a point about a possible 
explanation of an illusion, to 
make up a particular 
modification of the figure and 
simply look at it—–a 
demonstration, as it were—–in 
fact, psychologists generally do 
formal experiments with the 
modified figure. In these 
experiments, the observer is 
asked to match one test line to 
the other until they look equal. 
In experiments with the Ponzo 
illusion, for example, on some 
trials the lower test line is the 
standard and observers match 
the upper one (called the 
comparison stimulus) to it, in 
this case by making it shorter. 
On other trials, the upper test 
line is the standard and 
subjects match the lower one to 
it by making it longer. The 
average amount by which the 
observers must change the 
comparison line in order to 
match the standard line, for all 
trials and for many subjects, is 
a measure of the illusion. In 
the case of the illustrated 
example of the Ponzo illusion 
at the head of the chapter, the 
illusory length effect would be 
of the order of around 10 to 15%. 
(In this chapter, I will 
generally not give such 
quantitative data or describe 
details of experiments. 
However, the statements I make 
about how an illusion figure or 
a variation of it is perceived are 
typically based on such formal 
experiments.)