Memory for ordinal size information

In two experiments, the influence of stimulus modality on accuracy and latency of comparative judgments was explored. Subjects were presented with pictures or sentences enabling the generation of four term linear orders on a size dimension. They were tested with congruent stimulus modalities in Experiment 1, and with incongruent stimulus modalities in Experiment 2. During test, same size pairings of the four elements were presented. Subjects were required to indicate the formerly larger element in each test pair. Error rates and response latencies were considered the dependent variables. The main findings were: (1) The error rare pattern over test pairs differed from the reaction time pattern. (2) The error rate pattern was on the whole independent of study and test modality (3) For reaction times, there was a clear-cut picture-superiority effect dependent on test modality. (4) There were distance and anchor effects for reaction times that were little influenced by modality The obtained findings do not support the assumption that the task relevant information is represented and processed in a modality-specific visual system. The findings are, however; in line with Banks' assumption that relative size as other ordinal information is represented in some amodal code that is faster accessed by pictorial than by verbal stimuli.