Classification in the Wild

Abstract

Rules for building formal models that use fast-and-frugal heuristics, extending the psychological study of classification to the real world of uncertainty. This book focuses on classification—allocating objects into categories—“in the wild,” in real-world situations and far from the certainty of the lab. In the wild, unlike in typical psychological experiments, the future is not knowable and uncertainty cannot be meaningfully reduced to probability. Connecting the science of heuristics with machine learning, the book shows how to create formal models using classification rules that are simple, fast, and transparent and that can be as accurate as mathematically sophisticated algorithms developed for machine learning. The authors—whose individual expertise ranges from empirical psychology to mathematical modeling to artificial intelligence and data science—offer real-world examples, including voting, HIV screening, and magistrate decision making; present an accessible guide to inducing the models statistically; compare the performance of such models to machine learning algorithms when applied to problems that include predicting diabetes or bank failure; and discuss conceptual and historical connections to cognitive psychology. Finally, they analyze such challenging safety-related applications as decreasing civilian casualties in checkpoints and regulating investment banks.

Effect of Having, but Not Consulting, a Computerized Diagnostic Aid

“…50 Even though a “black box” algorithm may have been demonstrated to be accurate in the past, many humans still find it difficult to trust something that they cannot explain. 51 Eliminating algorithm aversion, then, will likely depend on our ability to develop “explainable” AI. 52,53…”

An exploratory study of heuristics for anticipating prices

“…Brown and Eisenhardt (1997) and Davis et al (2009) argue that simple rules work well because they provide not too little, not too much structure , thus giving direction while also leaving freedom to improvise. The fast-and-frugal-heuristics program has studied the general psychological capacities that enable people—experts as well as laypeople—to make accurate and also transparent decisions, forecasts, and classifications in management and banking, and other domains (Aikman et al , 2021; Gigerenzer et al , 2011; Katsikopoulos et al , 2020, 2018). The theory of ecological rationality delineates formal conditions under which simple, sometimes parameter-free fast and frugal heuristics can outperform complex parameterized models, such as low variance in the predictions of heuristics (Gigerenzer and Brighton, 2009) or the existence of dominating options or cues (Katsikopoulos, 2011).…”

“…Boost (Hertwig and Grüne-Yanoff, 2017) is an alternative to nudge that aims at empowering decisionmakers by providing knowledge and enhancing competencies. In the context of financial markets, boost interventions should go beyond the standard education in economics and finance that focuses on coping with risk, and empower policymakers and bankers to cope with uncertainty (Katsikopoulos et al , 2020). There are, however, gaps in the empirical evidence on the comparative efficacy of nudge and boost (van Roekel et al , 2021), including the anticipation of prices.…”

“…Market participants may realize that they are mutually boundedly rational , that is, that they have access to limited information and time, have to deal with uncertainty as opposed to risk, and that others are on the same boat; and use this realization to achieve their goals. In general, simple rules, also called heuristics , have been proposed as effective means for coping with uncertainty (Bettis, 2017; Bingham and Eisenhardt, 2014; Day and Lord, 1992; Gigerenzer et al , 2011; Guercini and Milanesi, 2020; Katsikopoulos et al , 2020; Maitland and Sammartino, 2015). To the best of our knowledge, however, heuristics for anticipating prices have not been studied empirically outside the lab.…”

“…The identification of heuristic processes is a key part of the endeavor of understanding and improving real-world decision-making (Ehrig and Schmidt, 2019; Gigerenzer and Gaissmaier, 2011; Katsikopoulos, 2022). We discuss how the findings may help to integrate conceptualizations of heuristics in the research programs of simple rules (Davis et al , 2009; Bingham and Eisenhardt, 2014), fast-and-frugal heuristics (Gigerenzer et al , 2011; Katsikopoulos et al , 2020) and to improve the efficiency of the financial market.…”