How Estimation Can Benefit From an Imbalanced World

Hertwig, Ralph; Hoffrage, Ulrich; Sparr, Rüdiger

doi:10.1093/acprof:oso/9780195315448.003.0116

Cited by 5 publications

(4 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Previous research for other tasks, such as paired comparison (Czerlinski, Gigerenzer, & Goldstein, 1999; Martignon & Hoffrage, 2002) or estimation (Hertwig, Hoffrage, & Sparr, 2012; Woike, Hoffrage, & Hertwig, 2012), has shown that simple heuristics, when making inferences out of sample, can outperform models that are optimal when it comes to making inferences within the same sample for which the strategies fitted their parameters. We found the same pattern in our simulations: When making classifications out of sample with multiple cues, the two heuristic approaches, naïve Bayes and some of the fast-and-frugal trees, outperformed the model that was normative for the case of fitting known data, namely classification based on natural frequencies (or, equivalently, profile memorization).…”

Section: Discussionmentioning

confidence: 99%

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Woike¹,

Hoffrage²,

Martignon³

2017

Decision

Self Cite

View full text Add to dashboard Cite

This article relates natural frequency representations of cue-criterion relationships to fast-and-frugal heuristics for inferences based on multiple cues. In the conceptual part of this work, three approaches to classification are compared to one another: The first uses a natural Bayesian classification scheme, based on profile memorization and natural frequencies. The second is based on Naïve Bayes, a heuristic that assumes conditional independence between cues (given the criterion). The third approach is to construct fast-and-frugal classification trees, which can be conceived as pruned versions of diagnostic natural frequency trees. Fastand-frugal trees can be described as lexicographic classifiers but can also be related to another fundamental class of models, namely linear models. Linear classifiers with fixed thresholds and noncompensatory weights coincide with fast-and-frugal trees-not as processes but in their output. Various heuristic principles for tree construction are proposed. In the second, empirical part of this article, the classification performance of the three approaches when making inferences under uncertainty (i.e., out of sample) is evaluated in 11 medical data sets in terms of Receiver Operating Characteristics (ROC) diagrams and predictive accuracy. Results show that the two heuristic approaches, Naïve Bayes and fast-and-frugal trees, generally outperformed the model that is normative when fitting known data, namely classification based on natural frequencies (or, equivalently, profile memorization). The success of fast-and-frugal trees is grounded in their ecological rationality: their construction principles can exploit the structure of information in the data sets. Finally, implications, applications, limitations, and possible extensions of this work are discussed.

show abstract

Section: Discussionmentioning

confidence: 99%

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Woike¹,

Hoffrage²,

Martignon³

2017

Decision

Self Cite

View full text Add to dashboard Cite

show abstract

“…People basing their choices on the choices already made by others (e.g., using imitation or copying heuristics) result in "rich-get-richer" dynamics that create distributions of popularity o f choices with just a few winners and many also-rans. Such emergent "J-shaped" distribu tions, seen in domains from music and book sales to social popularity to website visits, can in turn pre dictably influence the decisions o f others (Salganik, Dodds, & Watts, 2006;Hertwig, Hoffrage, & Sparr, 2012). Similarly, drivers seeking a parking space using a particular search heuristic, as mentioned previously, create a pattern o f available spots that is the environ ment for future drivers, who will search with their own heuristics that may fit that newly created envi ronment structure less well (Hutchinson, Fanselow, & Todd, 2012).…”

Section: Types O F Environm Ent Structurementioning

confidence: 99%

The Ecological Rationality of Situations

Todd

Gigerenzer

2019

The Oxford Handbook of Psychological Situations

View full text Add to dashboard Cite

The study of situations involves asking how people behave in particular environmental settings, often in terms of their individual personality differences. The ecological rationality research program explains people’s behavior in terms of the specific decision-making tools they select and use from their mind’s adaptive toolbox when faced with specific types of environment structure. These two approaches can be integrated to provide a more precise mapping from features of situation structure to decision heuristics used and behavioral outcomes. This chapter presents three examples illustrating research on ecological rationality and its foundations, along with initial directions for incorporating it into an integrated situation theory.

show abstract

“…Table presents details of the data sets' statistical structure. Multiplicative relationships between cues and criterion often lead to highly skewed criteria distributions (often called “J‐shaped”) and linear rules do not work well in such environments (Hertwig, Hoffrage, & Sparr, ; von Helversen & Rieskamp, ).…”

Section: Simulation Study Ii: Performance Of Strategy Selection Versumentioning

confidence: 99%

Strategy Selection Versus Strategy Blending: A Predictive Perspective on Single‐ and Multi‐Strategy Accounts in Multiple‐Cue Estimation

Herzog

Helversen

2016

Behavioral Decision Making

View full text Add to dashboard Cite

The claim that a person can use different strategies or processes to solve the same task is pervasive in decision making, categorization, estimation, reasoning, and other research fields. Yet such multi-strategy approaches differ widely in how they envision that the different strategies are coordinated and therefore do not represent one unitary approach. Toolbox models, for example, assume that people shift from one strategy to another as they adapt to specific task environments based on past experience. Unlike such multi-strategy selection approaches, multi-strategy blending approaches assume that the outputs of different strategies are blended into a joint, hybrid response (i.e., "wisdom of strategies" in one mind). The goal of this article is twofold. First, we discuss strategy blending as a conceptual alternative to strategy selection for modeling human judgment. Second, we investigate the predictive performance of the different approaches in synthetic and real-world environments. Taking a normative perspective, we study the coordination of rule-based and exemplar-based processes in estimation tasks. Our simulations using synthetic and real-world environments indicate that, for medium-sized samples, multi-strategy blending approaches lead to more accurate estimates than relying on a single strategy or selecting a strategy based on past experience-possibly because neither rule-nor exemplar-based processes in isolation are sufficient to capture statistical regularities that enable accurate estimates. This suggests that multi-strategy blending approaches can be advantageous to the degree that they rely on qualitatively different strategies. We thank Susannah Goss for editing the manuscript, and the Swiss National Science Foundation for a grant to the first author (100014_129572/1) and a grant to the second author (100014_146169).Correspondence concerning this article should be addressed to Stefan M. Herzog, Center for Adaptive Rationality, Max Planck Institute for Human Development, Lentzeallee 94, 14195 Berlin, Germany. E-mail: herzog@mpib-berlin.mpg.de This article may be used for non-commercial purposes in accordance with Wiley Terms and Conditions for Use of Self-Archived Versions. STRATEGY SELECTION VS. BLENDING 2 AbstractThe claim that a person can use different strategies or processes to solve the same task is pervasive in decision making, categorization, estimation, reasoning, and other research fields. Yet such multi-strategy approaches differ widely in how they envision that the different strategies are coordinated and therefore do not represent one unitary approach. Toolbox models, for example, assume that people shift from one strategy to another as they adapt to specific task environments based on past experience. Unlike such multi-strategy selection approaches, multi-strategy blending approaches assume that the outputs of different strategies are blended into a joint, hybrid response (i.e., "wisdom of strategies" in one mind).The goal of this article is twofold. First, we introduce strat...

show abstract

How Estimation Can Benefit From an Imbalanced World

Cited by 5 publications

References 0 publications

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

Integrating and testing natural frequencies, naïve Bayes, and fast-and-frugal trees.

The Ecological Rationality of Situations

Strategy Selection Versus Strategy Blending: A Predictive Perspective on Single‐ and Multi‐Strategy Accounts in Multiple‐Cue Estimation

Contact Info

Product

Resources

About