Self-Directed Learning Favors Local, Rather Than Global, Uncertainty

Douglas B. Markant, Burr Settles, Todd Gureckis

Research output: Contribution to journalArticle

Abstract

Collecting (or "sampling") information that one expects to be useful is a powerful way to facilitate learning. However, relatively little is known about how people decide which information is worth sampling over the course of learning. We describe several alternative models of how people might decide to collect a piece of information inspired by "active learning" research in machine learning. We additionally provide a theoretical analysis demonstrating the situations under which these models are empirically distinguishable, and we report a novel empirical study that exploits these insights. Our model-based analysis of participants' information gathering decisions reveals that people prefer to select items which resolve uncertainty between two possibilities at a time rather than items that have high uncertainty across all relevant possibilities simultaneously. Rather than adhering to strictly normative or confirmatory conceptions of information search, people appear to prefer a "local" sampling strategy, which may reflect cognitive constraints on the process of information gathering.

Original languageEnglish (US)
Pages (from-to)100-120
Number of pages21
JournalCognitive Science
Volume40
Issue number1
DOIs
StatePublished - Jan 1 2016

Fingerprint

Uncertainty
Learning
Sampling
Problem-Based Learning
Learning systems
Research
Self-directed Learning
Machine Learning

Keywords

  • Active learning
  • Information sampling
  • Machine learning
  • Self-directed learning

ASJC Scopus subject areas

  • Language and Linguistics
  • Artificial Intelligence
  • Cognitive Neuroscience
  • Experimental and Cognitive Psychology

Cite this

Self-Directed Learning Favors Local, Rather Than Global, Uncertainty. / Markant, Douglas B.; Settles, Burr; Gureckis, Todd.

In: Cognitive Science, Vol. 40, No. 1, 01.01.2016, p. 100-120.

Research output: Contribution to journalArticle

Markant, Douglas B. ; Settles, Burr ; Gureckis, Todd. / Self-Directed Learning Favors Local, Rather Than Global, Uncertainty. In: Cognitive Science. 2016 ; Vol. 40, No. 1. pp. 100-120.
@article{620dcfa223494a7b8270b034ecd43d42,
title = "Self-Directed Learning Favors Local, Rather Than Global, Uncertainty",
abstract = "Collecting (or {"}sampling{"}) information that one expects to be useful is a powerful way to facilitate learning. However, relatively little is known about how people decide which information is worth sampling over the course of learning. We describe several alternative models of how people might decide to collect a piece of information inspired by {"}active learning{"} research in machine learning. We additionally provide a theoretical analysis demonstrating the situations under which these models are empirically distinguishable, and we report a novel empirical study that exploits these insights. Our model-based analysis of participants' information gathering decisions reveals that people prefer to select items which resolve uncertainty between two possibilities at a time rather than items that have high uncertainty across all relevant possibilities simultaneously. Rather than adhering to strictly normative or confirmatory conceptions of information search, people appear to prefer a {"}local{"} sampling strategy, which may reflect cognitive constraints on the process of information gathering.",
keywords = "Active learning, Information sampling, Machine learning, Self-directed learning",
author = "Markant, {Douglas B.} and Burr Settles and Todd Gureckis",
year = "2016",
month = "1",
day = "1",
doi = "10.1111/cogs.12220",
language = "English (US)",
volume = "40",
pages = "100--120",
journal = "Cognitive Science",
issn = "0364-0213",
publisher = "Wiley-Blackwell",
number = "1",

}

TY - JOUR

T1 - Self-Directed Learning Favors Local, Rather Than Global, Uncertainty

AU - Markant, Douglas B.

AU - Settles, Burr

AU - Gureckis, Todd

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Collecting (or "sampling") information that one expects to be useful is a powerful way to facilitate learning. However, relatively little is known about how people decide which information is worth sampling over the course of learning. We describe several alternative models of how people might decide to collect a piece of information inspired by "active learning" research in machine learning. We additionally provide a theoretical analysis demonstrating the situations under which these models are empirically distinguishable, and we report a novel empirical study that exploits these insights. Our model-based analysis of participants' information gathering decisions reveals that people prefer to select items which resolve uncertainty between two possibilities at a time rather than items that have high uncertainty across all relevant possibilities simultaneously. Rather than adhering to strictly normative or confirmatory conceptions of information search, people appear to prefer a "local" sampling strategy, which may reflect cognitive constraints on the process of information gathering.

AB - Collecting (or "sampling") information that one expects to be useful is a powerful way to facilitate learning. However, relatively little is known about how people decide which information is worth sampling over the course of learning. We describe several alternative models of how people might decide to collect a piece of information inspired by "active learning" research in machine learning. We additionally provide a theoretical analysis demonstrating the situations under which these models are empirically distinguishable, and we report a novel empirical study that exploits these insights. Our model-based analysis of participants' information gathering decisions reveals that people prefer to select items which resolve uncertainty between two possibilities at a time rather than items that have high uncertainty across all relevant possibilities simultaneously. Rather than adhering to strictly normative or confirmatory conceptions of information search, people appear to prefer a "local" sampling strategy, which may reflect cognitive constraints on the process of information gathering.

KW - Active learning

KW - Information sampling

KW - Machine learning

KW - Self-directed learning

UR - http://www.scopus.com/inward/record.url?scp=84956799461&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84956799461&partnerID=8YFLogxK

U2 - 10.1111/cogs.12220

DO - 10.1111/cogs.12220

M3 - Article

VL - 40

SP - 100

EP - 120

JO - Cognitive Science

JF - Cognitive Science

SN - 0364-0213

IS - 1

ER -