Academic journal article Memory & Cognition

The Influence of Causal Information on Judgments of Treatment Efficacy

Academic journal article Memory & Cognition

The Influence of Causal Information on Judgments of Treatment Efficacy

Article excerpt

How does the causal structure of a problem concept influence judgments of treatment efficacy? We argue that the task of evaluating treatment efficacy involves a combination of causal reasoning and categorization. After an exemplar has been categorized, a treatment task involves judging where to intervene in the causal structure to eradicate the problem, removing the exemplar from category membership. We hypothesized that the processes underlying such category membership removal tasks are not identical to those underlying categorization. Whereas previous experiments have shown that both the root cause (as the most generative feature) and the coherence of the exemplar heavily influence categorization, Experiments 1 and 2 showed that people base category membership removal judgments on the root cause. In Experiment 3, people spontaneously chose to remove an exemplar from category membership when asked to treat the terminal effect. We discuss how our findings are compatible with existing models of categorization. A description of pilot studies for Experiment 1 may be downloaded as supplemental materials from mc.psychonomic-journals.org.

In the span of a lifetime, people commonly encounter some form of a problem (e.g., a mechanical problem or a disorder or disease) that requires treatment and often need to assess the efficacy of the different kinds of treatments available. For example, a person diagnosed with depression might consider whether to seek out psychotherapy or a prescription for antidepressants; another person might give an opinion about both general types of treatments to a friend. Yet despite the fact that making judgments about treatment efficacy is a ubiquitous task in the real world and, thereby, has high ecological validity, surprisingly little research in cognition has considered the question of how people make such judgments. Many kinds of information could conceivably be used in judging treatment effectiveness, including but not limited to information or misinformation about the general efficacy of broad types of treatments for a problem (e.g., psychotherapies or drug therapies), beliefs about the problem features' salience (Sloman, Love, & Ahn, 1998), rarity (McKenzie, 2006) or severity, and beliefs about the causal relationship(s) among the features of the problem (Waldmann & Holyoak, 1992).

The driving question of this article is how people's beliefs about the causal relationships between the features of a problem affect their judgments about the effectiveness of treatments, all else held equal. In this project, we used the problem domain of mental disorders to examine this broad question; prior work has already established that both laypeople and domain experts hold beliefs about causal relations between the symptoms, or features, of disorder concepts (Kim & Ahn, 2002a, 2002b). These relations constitute the perceived causal system underlying the concept.

At first glance, judgments about treatment effectiveness may seem to be merely a specific case of judging the likely consequences of interventions on such a causal system. Indeed, there is ample reason to believe, from the causal-reasoning literature, that adults and even children can readily predict the consequences of specific interventions on a causal system (e.g., Gopnik, Sobel, Schulz, & Glymour, 2001; Schulz, Gopnik, & Glymour, 2007; Sloman, 2005; Waldmann & Hagmayer, 2005). For example, Gopnik et al. (2001) showed children that Object A alone, but not Object B alone, activated a "blicket detector" when placed on its top. When the children were then shown an activated blicket detector with both objects on top, they reliably preferred to remove Object A to make the blicket detector stop activating. Findings by Sloman and Lagnado (2005) with adults have further shown that people told of an A [arrow right] B [arrow right] C causal chain can reliably predict what would happen to other, individual nodes in the chain when the intermediate cause (B) is intervened upon. …

Search by... Author
Show... All Results Primary Sources Peer-reviewed

Oops!

An unknown error has occurred. Please click the button below to reload the page. If the problem persists, please try again in a little while.