Adding Types, But Not Tokens, Affects Property Induction

Xie, Belinda; Navarro, Daniel J.; Hayes, Brett K.

doi:10.1111/cogs.12895

Cited by 9 publications

(9 citation statements)

References 46 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As with strong sampling, we therefore would expect tighter generalization with increasing data under conditions of property sampling. The predicted differences in the tightening of generalization under strong and weak sampling assumptions have been observed across of range of single-category learning and property inference tasks (e.g., Hayes, Banner, Forrester, et al, 2019;Hendrickson et al, 2019;Ransom et al, 2018;Ransom et al, 2016;Xie et al, 2020). Likewise, the predicted differences between property generalization under property and category sampling have also been demonstrated in property inference (Hayes, Banner, Forrester, et al, 2019).…”

Section: Formalizing Sampling Assumptionsmentioning

confidence: 83%

What do our sampling assumptions affect: how we encode data or how we reason from it?

Ransom¹,

Perfors²,

Hayes³

et al. 2021

Preprint

View full text Add to dashboard Cite

In describing how people generalize from observed samples of data to novel cases, theories of inductive inference have emphasized the learner's reliance on the contents of the sample. More recently, a growing body of literature suggests that different assumptions about how a data sample was generated can lead the learner to draw qualitatively different inferences on the basis of the same evidence. Yet relatively little is known about how and when these two sources of evidence are combined. For instance, do sampling assumptions affect how the sample contents are encoded, or is any influence exerted only at the point of retrieval when a decision is to be made? We report two experiments aimed at exploring this issue. By systematically varying both the sampling cover story and whether it is given before or after the training stimuli we are able to determine whether encoding or retrieval issues drive the impact of sampling assumptions. Across two experiments we find that the sampling cover story affects generalization when it is presented before the training stimuli, but not after, which we interpret in favor of an encoding account.

show abstract

Section: Formalizing Sampling Assumptionsmentioning

confidence: 83%

What do our sampling assumptions affect: how we encode data or how we reason from it?

Ransom¹,

Perfors²,

Hayes³

et al. 2021

Preprint

View full text Add to dashboard Cite

show abstract

“…However, we note that other studies have shown that repetition can increase the memorability of a claim without necessarily increasing confidence in its accuracy (Xie et al, 2020).…”

Section: False Consensus and Source Reliability 30mentioning

confidence: 55%

Getting to the source of the illusion of consensus

Desai¹,

Xie²,

Hayes³

2021

Preprint

Self Cite

View full text Add to dashboard Cite

Consensus between informants is a valuable cue to a claim’s epistemic value, wheninformants’ beliefs are developed independently of each other. Recent work (Yousif et al., 2019) described an illusion of consensus such that people did not generally discriminate between the epistemic warrant of true consensus, where a majority claim is supported by multiple independent sources, and false consensus arising from repeated claims from the same sources. We re-examined this issue in two ways. First, a Bayesian re-analysis of the Yousif et al. (2019) results found that the evidence for the illusion of consensus was not as strong as originally claimed. Second, three experiments tested the hypothesis that the illusion of consensus only arises when people are unsure about the independence of the primary sources on which informant claims are based. When steps were taken to highlight the independence between data sources in the true consensus conditions, and confidence ina claim was measured against a no consensus baseline (where there was an equal number of reports supporting and opposing a claim), we eliminated the illusion of consensus. Under these conditions, more weight was given to claims based on true consensus than false consensus. These findings have implications for how people evaluate consensus and reason with dependent information. Our findings are relevant to the contemporary media context, where the (in)dependence of information sources can be obscured and lead to the appropriation of erroneous beliefs.

show abstract

“…This is important because the strength of generalisation should be monotonically related to the amount of positive evidence observed (i.e., the monotonicity principle; Osherson et al, 1990). Although monotonicity might, in principle, be able to account for the results in Experiment 1, it is worth noting that in other induction and category learning tasks, participants sometimes interpret repetitions of the same instance as separate instances (Barsalou, Huttenlocher, & Lamberts, 1998;Xie, Hayes, & Navarro, 2018). Nonetheless, it was important to ensure that any group differences found in our predictive learning task could be attributed to diversity and not monotonicity.…”

Section: Methodsmentioning

confidence: 99%

Evidential diversity increases generalisation in predictive learning

Lee

Lovibond

Hayes

2019

Quarterly Journal of Experimental Psychology

View full text Add to dashboard Cite

In property induction tasks, encountering a diverse range of instances (e.g., hippos and hamsters) with a given property usually increases our willingness to generalise that property to a novel instance, relative to non-diverse evidence (e.g., hippos and rhinos). Although generalisation in property induction and predictive learning tasks share conceptual similarities, it is unknown whether this diversity principle applies to generalisation of a predictive association. We tested this hypothesis in two predictive learning experiments using differential training where one category of stimuli (e.g., fruits) predicted an outcome and another category (e.g., vegetables) predicted no outcome. We compared generalisation between a Non-Diverse group who were presented with non-diverse evidence in both positive (predicted the outcome) and negative (predicted no outcome) categories, and two groups who received the same training as the Non-Diverse group but with a more diverse range of exemplars in the positive (Diverse+ group) or negative (Diverse– group) category. Diversity effects were found for both positive and negative categories, in that learning about a diverse range of exemplars increased generalisation of a predictive association to novel exemplars from that same category. The results suggest that diversity, a key principle describing how we reason inductively, also applies to generalisation in associative learning tasks.

show abstract

Adding Types, But Not Tokens, Affects Property Induction

Cited by 9 publications

References 46 publications

What do our sampling assumptions affect: how we encode data or how we reason from it?

What do our sampling assumptions affect: how we encode data or how we reason from it?

Getting to the source of the illusion of consensus

Evidential diversity increases generalisation in predictive learning

Contact Info

Product

Resources

About