Strategy selection: An introduction to the modeling challenge

Marewski, Julian N.; Link, Daniela

doi:10.1002/wcs.1265

Cited by 37 publications

(41 citation statements)

References 100 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Other studies report conservatism (C. R. Phillips & Edwards, 1966), the finding that people under-react to evidence. 1 Heuristic models respond to this challenge by allowing heuristics to be context-sensitive, an example of strategy selection (Gigerenzer, 2008;Marewski & Link, 2014). Most models of strategy selection assume that people are able to assess the 1 We will mostly avoid the term "conservatism" to denote under-reaction to data, because it is sometimes conflated with a bias to give "conservative" probability judgments (i.e., judgments close to uniform probability).…”

Section: A Theory Of Learning To Infer Introductionmentioning

confidence: 99%

A theory of learning to infer

Dasgupta

Schulz

Tenenbaum

et al. 2019

Preprint

View full text Add to dashboard Cite

Bayesian theories of cognition assume that people can integrate probabilities rationally.However, several empirical findings contradict this proposition: human probabilistic inferences are prone to systematic deviations from optimality. Puzzlingly, these deviations sometimes go in opposite directions. Whereas some studies suggest that people under-react to prior probabilities (base rate neglect), other studies find that people under-react to the likelihood of the data (conservatism). We argue that these deviations arise because the human brain does not rely solely on a general-purpose mechanism for approximating Bayesian inference that is invariant across queries. Instead, the brain is equipped with a recognition model that maps queries to probability distributions. The parameters of this recognition model are optimized to get the output as close as possible, on average, to the true posterior. Because of our limited computational resources, the recognition model will allocate its resources so as to be more accurate for high probability queries than for low probability queries. By adapting to the query distribution, the recognition model "learns to infer." We show that this theory can explain why and when people under-react to the data or the prior, and a new experiment demonstrates that these two forms of under-reaction can be systematically controlled by manipulating the query distribution. The theory also explains a range of related phenomena: memory effects, belief bias, and the structure of response variability in probabilistic reasoning. We also discuss how the theory can be integrated with prior sampling-based accounts of approximate inference.

show abstract

Section: A Theory Of Learning To Infer Introductionmentioning

confidence: 99%

A theory of learning to infer

Dasgupta

Schulz

Tenenbaum

et al. 2019

Preprint

View full text Add to dashboard Cite

show abstract

“…The second one concerns the notion of the adaptive toolbox ): Rather than assuming decision makers' cognitive mechanisms can be best described in terms of just one 'all-purpose', domain-general tool (e.g., one type of strategy) serving to solve every single problem they encounter (see e.g., Marewski and Link 2014), the fast-and-frugal heuristics framework posits that people (and animals) can adaptively select from a large repertoire of different, domain-specific strategies (= including both various heuristics and other more complex methods): each strategy (= each 'tool' in the 'toolbox') is tuned to, and hence suitable for, a given domain, that is, a certain task environment (see e.g., Gigerenzer 2008a; Gigerenzer et al 1999;Gigerenzer and Brighton 2009;Gigerenzer and Gaissmaier 2011;Marewski et al 2010a, d;Todd and Gigerenzer 2000). 10 Being able to smartly choose among the different tools from this toolbox as a function of the task environment at hand makes up the expertise (and art) of clever professional decision making (see e.g., Hafenbrädl et al 2016).…”

Section: Four Points Key To the Study Of Fast-and-frugal Heuristicsmentioning

confidence: 99%

Heuristics as conceptual lens for understanding and studying the usage of bibliometrics in research evaluation

Bornmann

Marewski

2019

Scientometrics

Self Cite

View full text Add to dashboard Cite

While bibliometrics are widely used for research evaluation purposes, a common theoretical framework for conceptually understanding, empirically studying, and effectively teaching its usage is lacking. In this paper, we outline such a framework: the fast-and-frugal heuristics research program, proposed originally in the context of the cognitive and decision sciences, lends itself particularly well for understanding and investigating the usage of bibliometrics in research evaluations. Such evaluations represent judgments under uncertainty in which typically not all possible options, their consequences, and those consequences' probabilities of occurring may be known. In these situations of incomplete information, candidate descriptive and prescriptive models of human behavior are heuristics. Heuristics are simple strategies that, by exploiting the structure of environments, can aid people to make smart decisions. Relying on heuristics does not mean trading off accuracy against effort: while reducing complexity, heuristics can yield better decisions than more information-greedy procedures in many decision environments. The prescriptive power of heuristics is documented in a cross-disciplinary literature, cutting across medicine, crime, business, sports, and other domains. We outline the fast-and-frugal heuristics research program, provide examples of past empirical work on heuristics outside the field of bibliometrics, explain why heuristics may be especially suitable for studying the usage of bibliometrics, and propose a corresponding conceptual framework.

show abstract

“…Aside from that study, the most relevant theories and research on strategy selection are in the judgment and decision literature. This topic has generated rich but somewhat fragmented literature (see Marewski & Link, , for a wide‐ranging review). Early research and theory on human reasoning and decision making adopted the goal of identifying modal or “typical” mental models and processes, with relatively little emphasis on individual differences and, therefore, on questions of strategy selection.…”

Section: Normative Models Of Causal Reasoningmentioning

confidence: 99%

“…While a review of the strategy selection literature is beyond the scope of this paper, we provide here an account of where our paper fits into that literature. Marewski and Link () characterize the study of strategy selection as presenting three challenges. First, there is the descriptive goal (what strategies may comprise the human repertoire?).…”

Section: Normative Models Of Causal Reasoningmentioning

confidence: 99%

Causal Reasoning Under Ambiguity: An Illustration of Modeling Mixture Strategies

Shou

Smithson

2016

Behavioral Decision Making

View full text Add to dashboard Cite

Causal reasoning with ambiguous observations requires subjects to estimate and evaluate the ambiguous observations. Detecting how people process ambiguous observations can be complicated by individual differences in causal reasoning. This paper proposes a hierarchical model that accounts for the uncertainty in both the distribution of the functional form selection and the distribution of the ambiguity treatment selection. The model provides an alternative to self-report measures for identifying subjects' strategic choices in reasoning about causal relationships under ambiguity. The posterior distribution of the causal estimates is determined by both the functional form and the ambiguity processing strategy adopted by the reasoner. Our model is tested in a simulation study where it demonstrates its ability to recover the strategies and functional forms adopted by simulated subjects across a range of hypothetical conditions. In addition, the model is applied to the results of an experimental study.

show abstract

Strategy selection: An introduction to the modeling challenge

Cited by 37 publications

References 100 publications

A theory of learning to infer

A theory of learning to infer

Heuristics as conceptual lens for understanding and studying the usage of bibliometrics in research evaluation

Causal Reasoning Under Ambiguity: An Illustration of Modeling Mixture Strategies

Contact Info

Product

Resources

About