A Bayesian View of Covariation Assessment

McKenzie, Craig R. M.; Mikkelsen, Laurie A.

doi:10.1037/e501882009-397

Cited by 45 publications

(72 citation statements)

References 39 publications

(68 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…However, when the materials concerned features for which they possessed no relevant background beliefs (genotype and personality type), people were relatively insensitive to rarity information presented in the scenario. McKenzie and Mikkelsen (2007) have extended these results to a covariation detection task where they observed that people's perception of the information carried by the joint presence or joint absence of two possibly correlated variables was affected by whether absence or presence of the possibly correlated variables was rare. However, such rarity effects were only observed for concrete materials for which people were likely to have had pre-existing beliefs about rarity.…”

Section: Discussionmentioning

confidence: 72%

Rarity, pseudodiagnosticity and Bayesian reasoning

Feeney

Evans

Venn

2008

Thinking & Reasoning

View full text Add to dashboard Cite

Three experiments investigated the effect of rarity on people's selection and interpretation of data in a variant of the pseudodiagnosticity task. For familiar (Experiment 1) but not for arbitrary (Experiment 3) materials, participants were more likely to select evidence so as to complete a likelihood ratio when the initial evidence they received was a single likelihood concerning a rare feature. This rarity effect with familiar materials was replicated in Experiment 2 where it was shown that participants were relatively insensitive to explicit manipulations of the likely diagnosticity of rare evidence. In contrast to the effects for data selection, there was an effect of rarity on confidence ratings after receipt of a single likelihood for arbitrary but not for familiar materials. It is suggested that selecting diagnostic evidence necessitates explicit consideration of the alternative hypothesis and that consideration of the possible consequences of the evidence for the alternative weakens the rarity effect in confidence ratings. Paradoxically, although rarity effects in evidence selection and confidence ratings are in the spirit of Bayesian reasoning, the effect on confidence ratings appears to rely on participants thinking less about the alternative hypothesis.When interpreting evidence (D) with which to decide between a pair of hypotheses, X and Y, we need to consider both the prior probabilities of the hypotheses (P(Hx) and P(Hy)), and the probability of the evidence under Correspondence should be addressed to

show abstract

Section: Discussionmentioning

confidence: 72%

Rarity, pseudodiagnosticity and Bayesian reasoning

Feeney

Evans

Venn

2008

Thinking & Reasoning

View full text Add to dashboard Cite

show abstract

“…Moreover, common arithmetic word problem errors can be better simulated using deficient text comprehension strategies than deficient logico-mathematical strat egies (Dellarosa, 1986). Similar arguments concerning the neces sity of investigating how participants interpret decision-making queries in experiments have been put forth by Oaksford and Chater (1994) and McKenzie and Mikkelsen (2007).…”

Section: The Confusability Of Causal Power and Predictive Judgment Qumentioning

confidence: 76%

The impact of disablers on predictive inference.

Cummins¹

2014

Journal of Experimental Psychology: Learning, Memory, and Cogni

View full text Add to dashboard Cite

People consider alternative causes when deciding whether a cause is responsible for an effect (diagnostic inference) but appear to neglect them when deciding whether an effect will occur (predictive inference). Five experiments were conducted to test a 2-part explanation of this phenomenon: namely, (a) that people interpret standard predictive queries as requests to estimate the probability of the effect in the presence of the cause alone, which renders alternative causes irrelevant, and (b) that the impact of disablers (inhibitory causes) on predictive judgments is underestimated, and this underestimation is wrongly interpreted as cause neglect. Experiment 1 showed that standard predictive queries are frequently interpreted as requests to estimate the likelihood of E given C alone. In Experiment 2, a causal Bayes network overestimated predictive inference when it was queried in the standard way, but this overestimation diminished when predictive inference was queried using an alternative wording. In Experiment 3, participants judged alternative causes to be relevant to diagnostic inference and both disablers and alternative causes to be relevant to predictive inference. In Experiment 4, disablers greatly overshadowed alternative causes in predictive judgments, but their impact on diagnostic judgments was negligible. In Experiment 5, the order of disabler retrieval influenced causal judgments. Taken together, these results indicate that human causal inference cannot be adequately modeled unless the manner in which knowledge is retrieved and applied is taken into consideration.

show abstract

“…Models of Bayesian inference dictate that low probability events are especially informative (e.g., McKenzie & Mikkelsen, 2007) and, in a similar fashion, social psychological research on person perception suggests that rare behaviors lead to stronger traitbased attributions (Ditto & Jemmott, 1989;Fiske, 1980;Jones & Davis, 1965;Kelley, 1967). Indeed, participants did report that the cat-beater had performed a more uncommon act than the woman-beater, M = 2.00, 95% CI [1.71, 2.90], p b .001.…”

Section: Resultsmentioning

confidence: 96%