Raising argument strength using negative evidence: a constraint on models of induction

Abstract

Both intuitively, and according to similaritybased 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.

How do people learn from negative evidence? Non-monotonic generalizations and sampling assumptions in inductive reasoning

“…This is a clear violation of monotonicity consistent with earlier work (Heussen et al, 2011;Kalish & Lawson, 2007). Unlike Heussen et al (2011) we did not detect non-monotonic generalizations in the CLOSE NEGATIVE condition, but this seems likely to be a ceiling effect. More tellingly, when we examine the hypotheses that people offered, we find evidence for belief revision that can not be merely due to the falsification of alternative inconsistent hypotheses.…”

“…The prediction of monotonicity is robustly supported by experimental evidence (Osherson et al, 1990;Osherson et al, 1991;Smith et al, 1993), but violations do occur (Heussen, Voorspoels, Verheyen, Storms, & Hampton, 2011;Kalish & Lawson, 2007;Medin et al, 2003;Osherson et al, 1990). Our goal in this paper is to investigate when and why these violations arise, in particular when negative evidence is involved.…”

“…One possibility is that negative observations can point the reasoner to a relevant dimension for inference (Heussen et al, 2011;Medin et al, 2003). Under this theory, the negative observation in the ''alpha waves'' argument highlights a commonality between Mozart and Bach that is not shared by Metallica.…”

“…To construct arguments we used 12 topics from Heussen et al (2011): music, painters, public figures, types of ships, types of glass, types of displays, water bodies, wind instruments, fruit, water birds, insects, polar animals. The properties to be generalized were designed to sound realistic but correspond to things that people were likely to have little or no knowledge about (e.g., contain lycopene, create conversion currents, elicit alpha waves in the brain).…”

“…While the general tendency is for negative observations to weaken conclusions (Osherson et al, 1991), in certain problems they can elicit a rise in willingness to generalize (Heussen et al, 2011;Kalish & Lawson, 2007). For example, Heussen et al (2011) presented participants with arguments like the following:…”