Using SpAM to reveal decision processes and complex new context effects in similarity judgment

Abstract

The “Spatial Arrangement Method” (SpAM) has gained popularity for measuring similarity judgments (Perry, Cook, & Samuelson, 2011; Hout, Goldinger, & Ferguson, 2012; Kriegeskorte, 2012). In SpAM, multiple stimuli are freely arranged in two dimensions such that more similar stimuli are close together. We performed two SpAM experiments to investigate the process by which participants make multiple simultaneous similarity judgments using novel stimuli. The experiments differed across either two or three feature dimensions. Mouse and Eye-tracking measures, as well as the sequence of stimuli placed, provided a rich picture of participants’ decision processes as they made these judgments. Both experiments revealed strong effects of group context. Clustered presentation of stimuli by feature influenced both the order and the timing of placements, and despite equal metric spacing of stimuli along each dimension, participants typically warped placements along dimensions nonlinearly. We discuss implications of these findings for theories and models of similarity.


Back to Table of Contents