Abstract
Research on children's size estimation across varying distances began in the 1920s and has continued to the present. The most prominent method is the size-matching task, in which the participants are asked, for example, to select from several nearby comparison objects the one that corresponds in size to a distant standard object. In this overview, data on the development of size constancy from 24 studies, which provided 245 mean values for size constancy, were analyzed descriptively. Moreover, 102 out of the 245 means were also statistically analyzed. Overall, children of all ages underestimate the size of distant objects. Moreover, size estimations become more accurate with increasing age and with decreasing distance. However, these trends are modulated by several methodological variants of the size-matching paradigm-that is, the impact of the mode of presentation of the comparison stimuli (single/successive versus series presentation of the comparison objects), of the angular separation between standard and comparison objects (simultaneous versus nonsimultaneous visibility of standard and comparison), of the relative position of standard and comparison objects (comparison nearer than standard versus standard nearer than comparison), of the kind of experimental size instructions (objective size versus apparent size instructions), and of viewing conditions (monocular versus binocular viewing conditions). The existing theories on the development of size constancy include the proximal versus constancy mode theory, the metacognitive theory, and the perceptual learning theory. These theories are discussed against the background of the results of the meta-analysis.