Tversky and Kahneman’s Cognitive Illusions: Who Can Solve Them, and Why?

Abstract

In the present paper we empirically investigate the psychometric properties of some of the most famous statistical and logical cognitive illusions from the “heuristics and biases” research program by Daniel Kahneman and Amos Tversky, who nearly 50 years ago introduced fascinating brain teasers such as the famous Linda problem, the Wason card selection task, and so-called Bayesian reasoning problems (e.g., the mammography task). In the meantime, a great number of articles has been published that empirically examine single cognitive illusions, theoretically explaining people’s faulty thinking, or proposing and experimentally implementing measures to foster insight and to make these problems accessible to the human mind. Yet these problems have thus far usually been empirically analyzed on an individual-item level only (e.g., by experimentally comparing participants’ performance on various versions of one of these problems). In this paper, by contrast, we examine these illusions as a group and look at the ability to solve them as a psychological construct. Based on an sample of N = 2,643 Luxembourgian school students of age 16–18 we investigate the internal psychometric structure of these illusions (i.e., Are they substantially correlated? Do they form a reflexive or a formative construct?), their connection to related constructs (e.g., Are they distinguishable from intelligence or mathematical competence in a confirmatory factor analysis?), and the question of which of a person’s abilities can predict the correct solution of these brain teasers (by means of a regression analysis).

(Ir)rationality and cognitive biases in large language models

“…The tasks used to evaluate these models are taken primarily from Kahneman and Tversky's work [5][6][7]13], who designed a series of tasks to highlight biases and heuristics in human reasoning. Additional tasks [8][9][10] and facilitated versions [11] are also included. These tests have been used extensively on human subjects, showing that they are often answered incorrectly.…”

“…In total, we study the performance of seven language models on 12 cognitive tasks, listed in table 1 (full task details are included in electronic supplementary material, appendix 2). Nine of them are from the set of tasks originally designed by Kahneman and Tversky [5][6][7], Wason [8], Eddy [9] and Friedman [10], and three which are facilitated versions of these tasks [11]. For the birth sequence problem [5], two versions are included: one with an ordered sequence and one with a random sequence.…”

“…In this paper, we evaluate seven LLMs using cognitive tests proposed by Kahneman & Tversky [5][6][7] and others [8][9][10], as well as some facilitated versions formulated by Bruckmaier et al [11], and evaluate the responses across two dimensions: correct and human-like [12]. These tasks were initially designed to illustrate cognitive biases and heuristics in human reasoning, showing that humans often do not reason rationally [13]; in this case, we use them to evaluate the rationality of LLMs.…”

The theoretical framework on humanist ethno-metaphorical mathematics learning model: An impactful insight in learning mathematics

“…Mathematical inductive reasoning is based on some conclusion-drawing characteristics: transudative reasoning, or the ability to draw a conclusion from one case with another case; analogical reasoning, or the ability to draw a conclusion based on similarities of processes or data; generalization reasoning, or the ability to draw a general conclusion based on some limited data under scrutiny; the ability to estimate answers of solution and trend, interpolation, and extrapolation; the ability to provide explanation on an existing model, fact, property, relationship, or pattern; and the ability to use a relationship pattern to analyze a situation and establish a conjecture (Hendriana, 2017a). Indicators of mathematical deductive reasoning, meanwhile, include the following: the ability to perform calculation based on a certain rule or formula; the ability to draw a logical conclusion based on rules of inference, appropriate proportion, probability, correlation between two variables, and combination of multiple variables; and the ability to formulate direct evidence, indirect evidence, and evidence with mathematical induction (Hendriana et al, 2014(Hendriana et al, , 2017bBruckmaier et al, 2021;Shodikin et al, 2021).…”

The Influence of Visual Provenance Representations on Strategies in a Collaborative Hand-off Data Analysis Scenario

“…From the early work of decision theory [10,63], the goal was to understand how users arrived at a clear conclusion when working with ill-structured data. Researchers continued this work with the study of the sensemaking process, which covers the human tendency to oscillate between foraging for new information and schematizing how it fits with what one already knows [48].…”