Search

Your search keyword '"Chi, Duyi"' showing total 3 results
Start Over Author "Chi, Duyi" Publication Year Range Last 50 years
3 results on '"Chi, Duyi"'

1. Benford's Law: Testing the Effects of Distributions and Anchors on Number Estimation

Author

Chi, Duyi and Burns, Bruce

Subjects
Psychology, Decision making, Computer-based experiment, Statistics
Abstract

Recent research (e.g., Burns & Krygier, 2015; Chi & Burns, 2022) demonstrated that people could exhibit a strong bias towards the smaller first digits, which is consistent with the pattern predicted by Benford's law. However, this psychological phenomenon was predominantly observed when generating meaningful numbers for decision-making. We investigated explanations rooted in the statistical acquisition of distributional information and the impact of anchoring during number estimation. Undergraduate students were asked to estimate the weight, lifespan and group-size of animals after learning different distributions of these variables, supplied with an anchored value, either an average or a starting point, for reference. The Benford bias reasonably emerged regardless of the variable distribution, yet was strongly influenced by the anchored information. Notably, showing average values significantly suppressed Benford bias. These findings offered insights into the cognitive process of number estimation in the presence of statistical evidence and anchored information.

Published
2024

2. Why do People fit to Benford’s Law? – A Test of the Recognition Hypothesis

Author

Chi, Duyi and Burns, Bruce

Subjects
Psychology, Decision making, Statistical learning
Abstract

Burns & Krygier (2015) demonstrated that people could exhibit a strong bias towards the smaller first digits, in a way similar to that described by Benford’s law. This paper sought to expand the scope of this phenomenon and to test a possible explanation, the Recognition Hypothesis that a Benford bias is due to life-long environmental exposure to this statistical relationship. Participants completed three numerical tasks: A Generation Task requiring answering trivia questions; a Selection Task requiring selecting between two numerical responses; and an Estimation task requiring estimating the number of jelly beans in a jar. The results found no evidence of any first digit effect in the Recognition Task, some evidence of Benford bias in the Generation Task and strong evidence in the Estimation Task. Future research should focus on alternatives to the Recognition Hypothesis and investigate the parameters of Benford bias in generation tasks

Published
2022

3. First Digit Phenomenon in Number Generation Under Uncertainty: Through the Lens of Benford’s Law

Author

Chi, Duyi

Subjects
number generation, number estimation, first digit phenomenon, cognitive bias, Benford's law
Abstract

Decision making under uncertainty has been investigated by looking for regularities due to the application of heuristics (Tversky & Kahneman, 1974). Contemporary society demands that we estimate numbers when making decisions, for instance, the value of an item, so regularities in the numbers people generate could help us understand how humans deal with unknown situations. Recent research (e.g., Burns, 2009) suggests that people could spontaneously exhibit a stronger bias towards the smaller leading digits (e.g., 1, 2) that approximates Benford’s law, a well-established phenomenon of the first digits aggregated from the naturally occurring datasets. Hence, it may also represent a potential regularity in how people produce unknown numbers. Therefore, the present study attempted to investigate the conditions under which the first digit phenomenon might occur under uncertainty by examining the degree of fit to Benford’s law with various forms of numerical responses, and more importantly, testing the existing speculations of why people might present such a bias when generating unknown values. The key elements of the designs were the statements of numerical questions and simple visual displays for estimations. As expected, the first digit phenomenon was stronger when generating non-arbitrary numbers, compared to the arbitrary numbers. The critical findings were the extension of Benford’s law to the estimation tasks with a peak of digit-5; the continued failure of the recognition hypothesis as a reliable explanation; and the supporting evidence of the Integration Hypothesis, which emphasises the attribute of processing multiple information for the occurrence of the first digit phenomenon in number generation. Building on and extending the results of the previous research conducted, the outcomes of this project can assist in understanding: 1) how numerical responses to unknown questions inform theories of numerical cognition and decision making, and 2) how the pattern of leading digits generated from humans might offer implications for the practices of Benford’s law in fraud detection.

Published
2020

Catalog

Books, media, physical & digital resources
See catalog results