Effect of theory-based feature correlations on typicality judgments

Abstract

It is generally agreed that our concepts are formed around clusters of correlated features (Rosch, 1978). For instance, having wings is more likely to occur with being able to fly and having feathers than with being able to swim and having gills. Thus, we give the name bird to creatures that have wings and feathers and are able to fly. Within these correlation clusters, there are specific feature correlations that are more important and explicit in our conceptual representations.In our bird concept, having wings and being able to fly would be more closely tied to each other than having feathers and chirping. The goal of the present study is to examine which feature correlations, among all possible pairs in a cluster, are explicitly represented in concepts and affect typicality ratings of exemplars beyond the contribution of their individual features.The work on people's naive theories of natural categories (Carey, 1985;Keil, 1989;Murphy & Medin, 1985) sheds some light on our goals. According to this approach, people do not simply notice and record statistically correlated propertieswhen learning categories. Instead, people's knowledge of a domain emphasizes only a subset of the statistical correlations that occur. We believe that this possibility has yet to be tested, especially with natural categories, as will be elaborated upon below. In our studies we will strive to (1) demonstrate that explicitly represented feature correlations in concepts are theory based, (2) examine the content of these naive theories, and (3) demonstrate that these theory-based, explicitly represented feature correlations are more determinativeof category members' typicalitythan are theory-neutral feature correlations. To achieve these goals, we must first review previous studies on the effect of correlated features, particularly noting studies that involved correlated features in real-life categories. Effect of Correlated Features Found in Previous StudiesAre people sensitive to all correlations among features when learning novel categories? In an attempt to answer this, Medin, Altom, Edelson, and Freko (1982) had participants learn artificial categories. When asked to judge category membership for new items, those that preserved the study phase correlation were more likely to be selected as members, even when the item with the preserved correlation contained fewer typical features (see also Wattenmaker, 1991Wattenmaker, , 1993. However, these results might not generalize to realworld cases, because the correlation between features was perfect, unlike in natural concepts (e.g., having wings and being able to fly). Furthermore, in these studies, a rule involving the correlated features was more diagnostic and 107Copyright 2002 Psychonomic Society, Inc.This project was supported by National Institute of Mental Health Grant RO1 MH57737, awarded to the first author. We thank Kathy Johnson, Barbara Malt, Thomas Palmeri, and Steven Sloman for their extremely helpful comments on earlier drafts of this paper. We also thank Barbara Malt for pro...

Feature-feature causal relations and statistical co-occurrences in object concepts

“…In a subsequent experiment, participants indicated the nature and direction of the relationship with a labeled arrow and numerically rated its strength. Because a greater number of participants provided labeled relations for explicitly related pairs and these relationships were judged to be stronger, Ahn et al (2002) concluded that explicitly related features were theory based. Finally, in a typicality rating experiment using artificial category exemplars, broken correlations influenced typicality ratings only when they involved explicit relations.…”

“…In this article, the term causal relations is used to refer to explicit beliefs about causal connections between features, in concert with definitions used previously (Ahn, Marsh, Luhmann, & Lee, 2002;Murphy & Wisniewski, 1989; although causal theories often refers to a wider range of causally based knowledge). A causal theory between two features explains how one feature, such as has a blade, is causally responsible for a second, such as used for cutting (Murphy & Medin, 1985).…”

“…Finally, in a typicality rating experiment using artificial category exemplars, broken correlations influenced typicality ratings only when they involved explicit relations. Ahn et al (2002) concluded that "the correlations that are present in our representations of realworld concepts are the ones with explanatory relations" (p. 115).…”

“…First, the vast majority of previous research on feature correlations in natural concepts has focused on the basic level, and feature cooccurrence at this level provides insights into conceptual representations and processing (Ahn et al, 2002;Malt & Smith, 1984). More pragmatically, we possess feature production norms for 541 living and nonliving things, which allows the computation of statistical co-occurrences for a large number of feature pairs across a sizeable sample of concepts.…”

“…Not all correlated feature pairs had the same effect on typicality ratings, leading Malt and Smith to speculate that knowledge of statistical co-occurrences may be limited to only certain salient relations. Ahn et al (2002) hypothesized that the feature pairs in Malt and Smith (1984) that influenced typicality ratings were those for which participants possess a theory-based relation. Participants rated the correlational magnitude of Malt and Smith's feature pairs, and these ratings were used to distinguish between explicit (high magnitude) and implicit (low magnitude) pairs.…”