scholarly journals Towards a theory of individual differences in statistical learning

2017 ◽  
Vol 372 (1711) ◽  
pp. 20160059 ◽  
Author(s):  
Noam Siegelman ◽  
Louisa Bogaerts ◽  
Morten H. Christiansen ◽  
Ram Frost
Keyword(s):  
Individual Differences ◽  
Statistical Learning ◽  
Experimental Approach ◽  
Methodological Approach ◽  
Individual Performance ◽  
Forced Choice ◽  
Choice Test ◽  
Cognitive Sciences ◽  
Choice Tests ◽  
Good Proxy

In recent years, statistical learning (SL) research has seen a growing interest in tracking individual performance in SL tasks, mainly as a predictor of linguistic abilities. We review studies from this line of research and outline three presuppositions underlying the experimental approach they employ: (i) that SL is a unified theoretical construct; (ii) that current SL tasks are interchangeable, and equally valid for assessing SL ability; and (iii) that performance in the standard forced-choice test in the task is a good proxy of SL ability. We argue that these three critical presuppositions are subject to a number of theoretical and empirical issues. First, SL shows patterns of modality- and informational-specificity, suggesting that SL cannot be treated as a unified construct. Second, different SL tasks may tap into separate sub-components of SL that are not necessarily interchangeable. Third, the commonly used forced-choice tests in most SL tasks are subject to inherent limitations and confounds. As a first step, we offer a methodological approach that explicitly spells out a potential set of different SL dimensions, allowing for better transparency in choosing a specific SL task as a predictor of a given linguistic outcome. We then offer possible methodological solutions for better tracking and measuring SL ability. Taken together, these discussions provide a novel theoretical and methodological approach for assessing individual differences in SL, with clear testable predictions. This article is part of the themed issue ‘New frontiers for statistical learning in the cognitive sciences’.

scholarly journals Yes/No Recognition, Forced-choice Recognition, and the Human Hippocampus

2008 ◽  
Vol 20 (3) ◽  
pp. 505-512 ◽  
Author(s):  
P. J. Bayley ◽  
J. T. Wixted ◽  
R. O. Hopkins ◽  
L. R. Squire
Keyword(s):  
Forced Choice ◽  
Fundamental Difference ◽  
Choice Test ◽  
Test Items ◽  
Human Hippocampus ◽  
Choice Tests ◽  
Hippocampal Lesions ◽  
Study Test ◽  
Stimulus Materials ◽  
Test Delay

Two recent studies reported that yes/no recognition can be more impaired by hippocampal lesions than forced-choice recognition when the targets and foils are highly similar. This finding has been taken in support of two fundamental proposals: (1) yes/no recognition tests depend more on recollection than do forced-choice tests; and (2) the hippocampus selectively supports the recollection process. Using the same stimulus materials as in the earlier studies, we tested five memory-impaired patients with circumscribed hippocampal lesions and 15 controls. As in the earlier studies, participants studied 12 pictures of objects and then took either a 12-item forced-choice test with four alternatives or a 60-item yes/no test. Patients were impaired on both tests but did more poorly on the yes/no test. However, a yes/no test based on 12 study items would conventionally involve only 24 test items (i.e., 12 study items and 12 foil items). When we scored only the first 24 test items, the patients performed identically on the yes/no and forced-choice tests. Examination of the data in blocks of 12 trials indicated that the scores of the patients declined as testing continued. We suggest that a yes/no test of 60 items is difficult relative to a 12-item forced-choice test due to the increased study-test delay and due to increased interference, not because of any fundamental difference between the yes/no and forced-choice formats. We conclude that hippocampal lesions impair yes/no and forced-choice recognition to the same extent.

Self-Reported Sense of Smell Predicts Disgust Sensitivity and Disgust Reactivity

2018 ◽  
Vol 39 (4) ◽  
pp. 191-195
Author(s):  
Nicholas J. Kelley ◽  
Adrienne L. Crowell
Keyword(s):  
Individual Differences ◽  
Individual Difference ◽  
Olfactory Stimulus ◽  
Disgust Sensitivity ◽  
Sense Of Smell ◽  
Olfactory Ability ◽  
Good Proxy ◽  
Insight Into

Abstract. Two studies tested the hypothesis that self-reported sense of smell (i.e., metacognitive insight into one’s olfactory ability) predicts disgust sensitivity and disgust reactivity. Consistent with our predictions two studies demonstrated that disgust correlates with self-reported sense of smell. Studies 1 and 2 demonstrated, from an individual difference perspective, that trait-like differences in disgust relate to self-reported sense of smell. Physical forms of disgust (i.e., sexual and pathogen disgust) drove this association. However, the association between self-reported sense of smell and disgust sensitivity is small, suggesting that it is likely not a good proxy for disgust sensitivity. The results of Study 2 extended this finding by demonstrating that individual differences in self-reported sense of smell influence how individuals react to a disgusting olfactory stimulus. Those who reported having a better sense of smell (or better insight into their olfactory ability) found a disgusting smell significantly more noxious as compared to participants reporting having a poor sense of smell (or poor insight into their olfactory ability). The current findings suggest that a one-item measure of self-reported sense of smell may be an effective tool in disgust research.

An Exploration of Individual Differences in Statistical Learning

2014 ◽  
Author(s):  
Lucy C. Erickson ◽  
Michael Kaschak ◽  
Erik D. Thiessen ◽  
Cassie Berry
Keyword(s):  
Individual Differences ◽  
Statistical Learning

The Science of Wisdom in a Polarized World: Knowns and Unknowns

2020 ◽  
Author(s):  
Igor Grossmann ◽  
Nic M. Weststrate ◽  
Monika Ardelt ◽  
Justin Peter Brienza ◽  
Mengxi Dong ◽  
...  
Keyword(s):  
Individual Differences ◽  
Problem Solving ◽  
Best Practices ◽  
Cognitive Sciences ◽  
Methodological Challenges ◽  
The Common ◽  
Empirical Approaches ◽  
And Training ◽  
Contextual Differences

Interest in wisdom in the cognitive sciences, psychology, and education has been paralleled by conceptual confusions about its nature and assessment. To clarify these issues and promote consensus in the field, wisdom researchers met in Toronto in July of 2019, resolving disputes through discussion. Guided by a survey of scientists who study wisdom-related constructs, we established a common wisdom model, observing that empirical approaches to wisdom converge on the morally-grounded application of metacognition to reasoning and problem-solving. After outlining the function of relevant metacognitive and moral processes, we critically evaluate existing empirical approaches to measurement and offer recommendations for best practices. In the subsequent sections, we use the common wisdom model to selectively review evidence about the role of individual differences for development and manifestation of wisdom, approaches to wisdom development and training, as well as cultural, subcultural, and social-contextual differences. We conclude by discussing wisdom’s conceptual overlap with a host of other constructs and outline unresolved conceptual and methodological challenges.

scholarly journals Item desirability matching in forced-choice test construction

2021 ◽  
Vol 183 ◽  
pp. 111114
Author(s):  
Goran Pavlov ◽  
Dexin Shi ◽  
Alberto Maydeu-Olivares ◽  
Amanda Fairchild
Keyword(s):  
Test Construction ◽  
Forced Choice ◽  
Choice Test

scholarly journals Statistical Learning with a Nuisance Component (Extended Abstract)

2020 ◽  
Author(s):  
Dylan J. Foster ◽  
Vasilis Syrgkanis
Keyword(s):  
Statistical Learning ◽  
Domain Adaptation ◽  
Estimation Error ◽  
Nuisance Parameter ◽  
Individual Performance ◽  
Excess Risk ◽  
Risk Minimization ◽  
Target Parameter ◽  
Estimation Algorithms ◽  
The Impact

We provide excess risk guarantees for statistical learning in a setting where the population risk with respect to which we evaluate a target parameter depends on an unknown parameter that must be estimated from data (a "nuisance parameter"). We analyze a two-stage sample splitting meta-algorithm that takes as input two arbitrary estimation algorithms: one for the target parameter and one for the nuisance parameter. We show that if the population risk satisfies a condition called Neyman orthogonality, the impact of the nuisance estimation error on the excess risk bound achieved by the meta-algorithm is of second order. Our theorem is agnostic to the particular algorithms used for the target and nuisance and only makes an assumption on their individual performance. This enables the use of a plethora of existing results from statistical learning and machine learning literature to give new guarantees for learning with a nuisance component. Moreover, by focusing on excess risk rather than parameter estimation, we can give guarantees under weaker assumptions than in previous works and accommodate the case where the target parameter belongs to a complex nonparametric class. We characterize conditions on the metric entropy such that oracle rates---rates of the same order as if we knew the nuisance parameter---are achieved. We also analyze the rates achieved by specific estimation algorithms such as variance-penalized empirical risk minimization, neural network estimation and sparse high-dimensional linear model estimation. We highlight the applicability of our results in four settings of central importance in the literature: 1) heterogeneous treatment effect estimation, 2) offline policy optimization, 3) domain adaptation, and 4) learning with missing data.

Sign in / Sign up

Export Citation Format

Share Document