The objective of this research was to see if the evaluation of the effectiv
eness of a given treatment through visual inspection depends on the type or
graphs used. Three subjects were studied with 3000 graphs (1.000 line, 1.0
00 bar and 1.000 bur-plots), each one representing the results of an A-B de
sign. The results indicate that the subjects committed more errors when the
data was presented by lines or bars instead of box-plots.