A local sampling account of random generation

Read full text Summarise with VIRAT
Authors: Lucas Castillo
Cited by: 0

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

Across more than seventy years of Random Generation research it has been shown that, when attempting to produce a series of random numbers or letters, people instead produce sequences that deviate from randomness in predictable ways. On the other hand, however, descriptions of human behaviour in most other tasks involve randomness, given that people are quite unpredictable in how they will act. In this Thesis I propose a new way to understand how people produce items in a Random Generation task, the Local Sampling account, which resolves this apparent paradox. I postulate that when generating items for a Random Generation task, people are tapping into a general cognitive ability to produce samples, which they typically use to make judgments or decisions. These samples are by nature somewhat unpredictable, which explains why human behaviour cannot be predicted with certainty, but they are also autocorrelated, in ways that explain the deviations from randomness found in Random Generation tasks. Across three Chapters and four empirical experiments, I show that this novel account can reproduce patterns previously found in the literature, while making novel predictions on how people will generate items when these are not known prior to the experiment, or not uniformly distributed (Chapter 2). I also show that, because this faculty is a task-general faculty, not asking people to produce random items does not change the patterns found in their sequences (Chapter 3). Finally, I link back to the randomness found in behaviour elsewhere, showing that while not random instantaneously people can become unpredictable if given enough time, which has implications for process models of cognition that include random draws (Chapter 4).

Fields of Study & Subjects

Decision-Making and Behavioral Economics Cognitive Science and Education Research Neural and Behavioral Psychology Studies

Publication details

Journal / source
Warwick Research Archive Portal (University of Warwick)
Publication date
17 September 2026
Language
EN
Document type
Thesis