Convince Me: Modeling Naturalistic Decision Making

McAndrew, Claire; Gore, Julie; Banks, Adrian P.

doi:10.1518/155534309x441862

Cited by 5 publications

(5 citation statements)

References 19 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Cropp, Banks, and Elghali (2011) compared three models on their ability to capture experts’ decision policies when performing contaminated-land risk assessments, namely, the lens model (i.e., a linear equation typically derived using multiple regression; see Brunswik, 1952), the matching heuristic (a fast-and-frugal heuristic; see Dhami, 2003), and a coherence model (based on the theory of explanatory coherence; see McAndrew, Gore, & Banks, 2009). Results using a cross-validation technique showed that the lens model (i.e., linear regression) provided the best account of experts’ decisions in that task context.…”

Section: Discussionmentioning

confidence: 99%

Judgment Analysis in a Dynamic Multitask Environment: Capturing Nonlinear Policies Using Decision Trees

Lafond

Roberge-Vallières²,

Vachon³

et al. 2016

Journal of Cognitive Engineering and Decision Making

View full text Add to dashboard Cite

Policy capturing is a judgment analysis method that typically uses linear statistical modeling to estimate expert judgments. A variant to this technique is to capture decision policies using data-mining algorithms designed to handle nonlinear decision rules, missing attributes, and noisy data. In the current study, we tested the effectiveness of a decision-tree induction algorithm and an instance-based classification method for policy capturing in comparison to the standard linear approach. Decision trees are relevant in naturalistic decision-making contexts since they can be used to represent "fast-and-frugal" judgment heuristics, which are well suited to describe human cognition under time pressure. We examined human classification behavior using a simulated naval air defense task in order to empirically compare the C4.5 decision-tree algorithm, the k-nearest neighbors algorithm, and linear regression on their ability to capture individual decision policies. Results show that C4.5 outperformed the other methods in terms of goodness of fit and cross-validation accuracy. Decisiontree models of individuals' judgment policies actually classified contacts more accurately than their human counterparts, resulting in a threefold reduction in error rates. We conclude that a decision-tree induction algorithm can yield useful models for training and decision support applications, and we discuss the application of judgmental bootstrapping in real time in dynamic environments.

show abstract

Section: Discussionmentioning

confidence: 99%

Judgment Analysis in a Dynamic Multitask Environment: Capturing Nonlinear Policies Using Decision Trees

Lafond

Roberge-Vallières²,

Vachon³

et al. 2016

Journal of Cognitive Engineering and Decision Making

View full text Add to dashboard Cite

show abstract

“…For example, in a coherent account, all the facts are explained or all the contradictory arguments are resolved. Coherence has been proposed as a basis for NDM (Cropp, Banks, & Elghali, ; McAndrew et al ., ) based on the work of Thagard (), but these models do not distinguish between different forms of coherent representation. We suggest that both story building and argumentation‐based decision‐making are based on different forms of representation, but the aim of both is to be coherent.…”

Section: Discussionmentioning

confidence: 99%

“…Whilst there are some NDM studies in domains that have greater opportunity for analytic decision‐making such as day traders (e.g., McAndrew & Gore, ; McAndrew, Gore, & Banks, ), these studies are less common and there is less understanding of how analytic naturalistic decisions are made. The aim of this study is to develop this understanding by comparing current theoretical explanations of analytic NDM with an alternative account that we propose based on decision‐making through argumentation.…”

Section: Introductionmentioning

confidence: 99%

Naturalistic decision making through argumentation: Resolving labour disputes

Ramiah

Banks

2015

J Occupat & Organ Psyc

View full text Add to dashboard Cite

Naturalistic decision making (NDM) research has been successful in describing fast, intuitive decision‐making by experts in complex, time‐pressured situations. However, less is known about how experts make naturalistic decisions when they think analytically. Previous research suggests that this process involves constructing a story of the situation to explain and simulate the consequences of decisions, but we hypothesize that explicit argumentation is valuable when it is necessary to demonstrate the accuracy of a decision and to explicitly resolve inconsistent and uncertain information. This study investigates analytic NDM by contrasting experts and novices resolving labour disputes. Fifty expert and novice labour officers resolved labour cases whilst thinking aloud. The verbal protocols were analysed using Toulmin's argument model. Expert decision‐makers made more accurate decisions, used more warrants (in other words, reasoned justifications) during the decision process, and increased use of warrants was associated with accurate decision‐making. In addition, a decision aid was developed in the form of a checklist that encouraged decision‐makers to explicitly consider components of the argument. This increased the frequency of warrants and the decision accuracy of both novice and expert decision‐makers. Together, these findings imply that accurate and expert analytic decision‐making in this domain involves a process of argumentation and in particular the use of explicit warrants linking evidence to conclusions. Practitioner points Analytic naturalistic decision making in this domain involves argumentation Superior decision accuracy arises from exploring more reasoned justifications A checklist encouraging all components of the argument to be made explicit improves decision accuracy in both novice and expert decision‐makers.

show abstract

“…When the model is run, activation spreads throughout the network, and it settles into a state in which the activation of the nodes is stable, given the excitatory and inhibitory links. This theory has been used to model decision making by jurors (Byrne, 1995) and investment fund managers (McAndrew, Gore, & Banks, 2009), providing evidence for the sufficiency of this account of decision making. Defra, 2004).…”

Section: Application Of Decision-making Modelsmentioning

confidence: 99%

Expert Decision Making in a Complex Engineering Environment

Cropp

Banks

Elghali

2011

Journal of Cognitive Engineering and Decision Making

Self Cite

View full text Add to dashboard Cite

This study investigated the complex decisions made by engineers when conducting contaminated-land risk assessments. Experienced assessors studied summaries of site reports, which were composed of different combinations of relevant cues, and decided on the risk level of each site. Models from three theories of decision making were compared. Applying judgment analysis to develop a lens model provided the best account of the data, lending support to social judgment theory. A model based on a fast-and-frugal heuristic, the matching heuristic, did not fit the data as well; nor did a coherence model based on the theory of explanatory coherence. Comparison with decisions generated with the use of industry guidance showed only a moderate fit, suggesting that the standard procedure does not accurately represent how highly proficient domain practitioners make assessments in this context. Qualitative analyses of comments made by participants suggested that they used a combined approach that applied key cues as predicted by social judgment theory, integrated into a meaningful, coherent account, as predicted by the theory of explanatory coherence. Overall, these findings suggest a novel process in which a range of information is combined to form a coherent explanation of the data but in which key cues are more influential than others.

show abstract

Convince Me: Modeling Naturalistic Decision Making

Cited by 5 publications

References 19 publications

Judgment Analysis in a Dynamic Multitask Environment: Capturing Nonlinear Policies Using Decision Trees

Judgment Analysis in a Dynamic Multitask Environment: Capturing Nonlinear Policies Using Decision Trees

Naturalistic decision making through argumentation: Resolving labour disputes

Expert Decision Making in a Complex Engineering Environment

Contact Info

Product

Resources

About