Academic journal article Memory & Cognition

Raising Argument Strength Using Negative Evidence: A Constraint on Models of Induction

Academic journal article Memory & Cognition

Raising Argument Strength Using Negative Evidence: A Constraint on Models of Induction

Article excerpt

Abstract Both intuitively, and according to similarity-based theories of induction, relevant evidence raises argument strength when it is positive and lowers it when it is negative. In three experiments, we tested the hypothesis that argument strength can actually increase when negative evidence is introduced. Two kinds of argument were compared through forced choice or sequential evaluation: single positive arguments (e.g., "Shostakovich's music causes alpha waves in the brain; therefore, Bach's music causes alpha waves in the brain") and double mixed arguments (e.g., "Shostakovich's music causes alpha waves in the brain, X's music DOES NOT; therefore, Bach's music causes alpha waves in the brain"). Negative evidence in the second premise lowered credence when it applied to an item X from the same subcategory (e.g., Haydn) and raised it when it applied to a different subcategory (e.g., AC/DC). The results constitute a new constraint on models of induction.

Keywords Induction . Negative evidence . Categories . Models of induction

Much of everyday reasoning consists of inductive inference, in which induction is taken in its broadest sense as an inference to an uncertain conclusion. One strategy to make such an inference is to use past experiences-for example, "My car has always started, so it's reliable and will start today." Another strategy is to use category membership- for example, "My car is a German car, so it's reliable and will start today." Research on induction in psychology has predominately focused on the latter of the two. In a typical experiment, people are told that one or several categories have a particular property and are asked to extend that property to other categories within a given domain. Participants, for instance, might be asked to judge how likely it is that bobcats use serotonin as neurotransmitter given that both tigers and cougars do (Smith, Shafir, & Osherson 1993). Numerous phenomena relating to category-based property induction have been documented (e.g., Hampton & Cannon, 2004; Heit, Hahn, & Feeney 2004; Rips, 1975; for a review, see Heit, 2000), and various models have been proposed to account for people's judgments of inductive arguments such as these (e.g., Blok, Medin, & Osherson 2007; Heit, 1998; Kemp & Tenenbaum, 2009; Medin, Coley, Storms, & Hayes 2003; Osherson, Smith, Wilkie, Lopez, & Shafir 1990; Sloman, 1993).

In the majority of cases, experimental work and modeling efforts have focused on arguments involving positive evidence-premises that state that some entity possesses the to-be-projected property. In reality, though, we do not only receive positive evidence for our hypotheses. We are often confronted with negative evidence, evidence that states that some entity of the same or of a similar category DOES NOT possess the to-be-projected property. For instance, in evaluating whether bobcats use serotonin as neurotransmitter, we might find out that tigers do, but cougars do not. How do we integrate the negative evidence, and how does it influence our judgment about bobcats?

In the present article, we are interested in the influence of negative evidence on argument strength. More precisely, we are interested in the direction of this influence. Contrary to intuition and to the general predictions of similarity-based theories of induction, we present evidence that, under certain circumstances, negative evidence can actually facilitate the generalization of a property. In what follows, we will identify and discuss in more detail the general assumption that negative evidence has a negative effect on argument strength. Then, we will present three experiments that undermine the universal validity of this assumption. In the General Discussion section, we address in more detail the implications of these findings for well-known models of induction.

Monotonicity in inductive reasoning

In absence of other information, it is generally assumed that generalization relies on similarity (Blok et al. …

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


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.