The form change paradigm was used to explore the basis for the picture supe
riority effect. Recognition memory for studied pictures and words was teste
d in their study form or the alternate form. Form change cost was defined a
s the difference between recognition performance for same and different for
m items. Based on the results of Experiment 1 and previous studies, it was
difficult to determine the relative cost for studied pictures and words due
to a reversal of the mirror effect. Wie hypothesized that the. reversed mi
rror effect results from subjects' basing their recognition decisions on th
eir assumptions about the study form. Experiments 2 and 3 confirmed this hy
pothesis and generated a method for evaluating the relative cost for pictur
es and words despite the reversed mirror effect. More cost was observed for
pictures than words, supporting the distinctiveness model of the picture s
uperiority effect.