Garner Interference

In the standard paradigm used by Garner and others, Ss are presented with stimuli that vary in two dimensions or elements. E.g., they see circles and squares that are red or green; or they see rectangles that are short or tall and are narrow or wide; or they see letter pairs where the letter on the left can be A or B, and likewise the letter on the right can be A or B. In the Control Condition (one of several conditions tested), Ss classify sequentially presented stimuli on the basis of one of these dimensions or elements while the other one remains constant in the presentation sequence. For example, they might see a random sequence of AA and AB stimuli, where the letter on the left is always an A, and where Ss classify the stimuli as quickly and accurately as possible on the basis of the letter on the right. In the Filtering Condition they do the same, only now the irrelevant dimension or element randomly varies from stimulus to stimulus. For example, Ss would see a random sequence of AA, AB, BA, and BB, and classify them again on the basis of the righthand letter. Garner interference is defined by the difference in performance (RTs, error rates) between these two: Filtering minus Control performance = Garner Interference. Along with Stroop Interference, it is a standard way to measure selective attention. If selective attention is perfect, Garner Interference will be zero. This measure has shown that stimulus elements that appear to group into distinct configurations show Garner Interference whereas elements that do not group show none. (JRP, February 1, 2007)