Advisor:  Alan Searleman

SLU Festival of Science 2001 Poster Presentation

 

The Mueller-Lyer Illusion (ML) is one of the oldest and best known of the visual illusions.  In its standard form, the illusion consists of two horizontal lines, one having arrowheads pointing outwards and one having arrowheads pointing inwards.  Even though the horizontal lines are of equal length, most people perceive the one with the outward arrowheads as being from 20% to 30% longer.  Although people have been trying to explain this illusion (and its many variants) since it was first introduced in 1889, so far no one has been able to do so.

Coren (1986) has suggested that as people shift their eyes from one target to another, they do so in such a way as to maximize the amount of information that can be seen in a single fixation.  These more encompassing eye movements have been referred to as the Functional Fovea, which is a 2-4 degree circle within which visual acuity is optimal.  By adding an extraneous stimulus close to the intended target, and thereby manipulating the placement of the functional fovea, Coren has been able to show that this can alter a person’s estimate of length.

We decided to investigate how manipulating the functional fovea can influence the ML illusion.  In Experiment 1, 17 participants were shown the normal arrowhead version of the illusion, and also two additional versions in which extraneous dots were added to change the functional fovea.  The extraneous dots were placed in both the arrowhead outward and inward components of the illusion so as to either maximize the illusion's effect or to minimize it.  The task for the participants was to indicate which component appeared longer and then to indicate how much longer by making a slash on a line that represented the difference in perceived line lengths.  For all stimuli, the horizontal lines were always 83mm in length.  The results indicated that adding the extraneous dots significantly enhanced the normal illusion effect in the MAXIMIZE condition and completely eliminated the illusion effect in the MINIMIZE condition.  In a second experiment, again with 17 participants, we were able to replicate the first set of results and to include appropriate control conditions.  In, addition, in the second experiment we also examined a different variant of the ML illusion -- known as the dumbbell version.  Manipulating the functional fovea for this variant of the illusion was not as powerful as was found with the arrowhead version.  We believe that this may have occurred because the dumbbell version already contains a very visible extraneous target to draw the functional fovea away from the ends of the lines.  Thus, adding the dots would not have the same dramatic effect as it did for the arrowhead version.