We also explored the extent towhich the expressions of an emotion
conveyed by faces were incorrectly labeled. For each cell of the
experimental design, misidentification percentages were computed
for each incorrect (non-target) emotion. This was accomplished by
dividing the number of times that each incorrect emotion was selected
by the number of times the target emotion was presented. To illustrate,
we provide an example. In the 100ms presentation condition for happy
faces presented at the profile view, we divided the number of fearful