Asymptotic judgment of cause in a relative validity paradigm

Abstract

We report three experiments in which we tested asymptotic and dynamic predictions of the Rescorla-Wagner(R-W) model and the asymptotic predictions of Cheng's probabilistic contrast model (PCM) concerning judgments of causality when there are two possible causal candidates. We used a paradigm in which the presence of a causal candidate that is highlycorrelated with an effect influences judgments of a second, moderately correlated or uncorrelated cause. In Experiment I, which involved a moderate outcome density, judgments of a moderately positive cause were attenuated when it was paired with either a perfect positive or perfect negative cause. This attenuation was robust over a large set of trials but was greater when the strong predictor was positive. In Experiment 2, in which there was a low overall density of outcomes, judgments of a moderately correlated positive cause were elevated when this cause was paired with a perfect negative causal candidate. This elevation was also quite robust over a large set of trials. In Experiment 3, estimates of the strength of a causal candidate that was uncorrelated with the outcome were reduced when it was paired with a perfect cause. The predictions of three theoretical models of causal judgments are considered. Both the R-W model and Cheng's PCMaccounted for some but not all aspects of the data. Pearce's model of stimulus generalization accounts for a greater proportion of the data.In the last decade, research on human judgment ofcontingency has been motivated by a comparative perspective that stresses an analogous associative learning component in both animal conditioning and human judgments of contingency. When a single predictor variable signals the occurrence of an outcome variable, contingency judgments closely parallel the actual one-way contingency, I1P, between the predictor and the outcome. I1P is defined as the difference between the conditional probabilities p(outcome Ipredictor) and p(outcome Ino predictor) (Allan, 1980).This parallel has been observed in both operant learning tasks (see, e.g., Wasserman, Elek, Chatlosh, & Baker, 1993) and video games that are analogous to classical conditioning tasks (see, e.g., Dickinson, Shanks, & Evenden, 1984). These results suggest that subjects might derive their judgments from an unbiased consideration of I1P. called discounting or blocking (we will use these terms interchangeably throughout). In a series of experiments inspired by Kamin's (1969) blocking preparation, Dickinson et at. (1984) demonstrated that contingency estimates of moderately positive predictor A were reliably attenuated by exposing subjects to a strong positive contingency between predictor B and the same outcome in an initial observation stage. In turn, in a series of experiments inspired by the multiple-cue preparation of Wagner, Logan, Haberlandt, and Price (1968), Baker, Mercier, Vallee-Tourangeau, Frank, andPan (1993) established that estimates of moderately positive or negative contingencies between predictor A and the outcome were influenced ...

Models of covariation-based causal judgment: A review and synthesis

“…It has been repeatedly demonstrated that causal judgments deviate from contingency in a number of experimental preparations (Allan & Jenkins, 1983;Baker, Vallée-Tourangeau, & Murphy, 2000;Lober & Shanks, 2000;Perales & Shanks, 2003). In light of the previous argument, however, this does not imply that judgments deviate from normativity.…”

Locus of Equity and Brand Extension

“…When considered in the context of related research, discounting and more complex causal reasoning theories (e.g., Cheng 1997) lack parsimony (e.g., van Osselaer and Alba 2000). In addition, other authors have pointed out shortcomings of causal reasoning in general as an explanation of cue-interaction phenomena (e.g., Baker, Vallée-Tourangeau, and Murphy 2000;Lober and Shanks 2000).…”

Individual differences are more than a gene × environment interaction: The role of learning.

“…Competition with contextual cues can have powerful effects on learning (Baker et al, 1996;Baker et al, 2000;Murphy et al, 2005). Msetfi and colleagues (2005) showed that individual differences in judgments of causal control, seemingly related to depressive symptoms, were themselves related to context learning.…”