Two extensions of the anchor-range effect

Abstract

A stimulus from one category increased the rated similarity oftwo stimuli from another category. In addition, "different" responses, but not "same" responses, to same-category stimuli became slower when they occurred in blocks of trials with a stimulus from a second category. Another experiment employed figural stimuli in which two small disks and a large or very large (extreme) disk were simultaneously present. The extreme disk led to slower "different" and faster "same" responses to the two small disks. Slow "different" and fast "same" responses may indicate high perceived similarity. Consequently, both the second-category stimulus and the extreme disk, although logically irrelevant, may have increased the perceived similarity of more related stimuli. It is concluded that the second-category stimuli and the extreme disk functioned as conventional anchors. The results also call into question memorial and criterial variance theories and support category-superordinate and distance-density theories of the anchor-range effect.

Bridging the Gap between Similarity and Causality: An Integrated Approach to Concepts

“…26 As will become clear in Section 4.2, the difficulty is inherited by other theories that take up a similarly flat view of the application of causal knowledge in categorization. 27 It does not, however, serve as a direct argument against hybridism, pluralism, or eliminativism. 28 For recent discussions see (Casasanto and Lupyan [2015]; Machery [2015]).…”

“…It has been shown that the similarity rating for two targets, even when confined to a single dimension, is influenced by the presentation of a third target that serves as a foil. For example, two small black disks, presented along with a large black disk, were judged to be more similar as the size of large disk increased (King and Atef-Vahid [1986]). This means that what one takes as the range of acceptable values for a dimension might change as one encounters new targets.…”

“…I suggest that the model undercuts the motivation for the hybridism, pluralism, or eliminativism formulated by the new consensus. 27 I should emphasize that the suggested model should not be taken to be exact or comprehensive. Rather, it provides a framework within which the above discussion can be understood and integrated.…”

“…Given the effect of context on the selection of relevant respects for similarity (see discussion in Section 2.2), a presentation of a new target may, for example, affect the selection of relevant dimensions for similarity calculation. Additionally, in accordance with the anchor-range effect (King and Atef-Vahid [1986]), a new target may alter the ranges of the appropriate values with respect to the dimensions deemed relevant for similarity calculations, thus changing the resolution of the similarity judgements.…”

A large rectangle delays the perception of a separate small rectangle

“…Furthermore, this principle holds across areas as diverse as the anchor-range effect (Gravetter & Lockhead, 1973;Pollack, 1952), the resolution of perceptual differences (Welch, 1978), and the aftereffect obtained by presenting one object and then a second similar object (Gibson, 1933). With figural stimuli, an extreme (anchor) object strengthens the gestalt established by two other objects and also increases the perceived similarity of the other objects (King & Atef-vahid, 1986;Lockhead, 1988). For example, in the research of Gaylord (1984) that Lockhead describes, a spatially extreme third dot made two other dots group and also made them appear closer together (i.e., in more similar spatial positions).…”