Print version

Naturalistic decision making in aviation environments

• Download this publication

Abstract

The majority of accidents and incidents in aviation can be attributed partially to poor decision making and judgement strategies. By gaining an understanding of aircrew decision strategies used in the operational environment of the cookpit, it may be possible to improve decision processes and outcomes for both expert and novice pilots. Naturalistic decision making (NDM) is proposed as an intuitive decision strategy used by experienced pilots to make operational decisions. The cockpit is considered a naturalistic environment due to characteristics such as experienced operators, multiple players and teams, dynamic conditions, shifting and competing goals, high risks, time pressure, and ambiguous or missing data. NDM strategies focus upon situation assessment and the serial evaluation of decisions through mental simulation. Decision aiding to improve situation assessment and decision training to impart awareness of the limitations and weaknesses (heuristics and biases) of human decision processes are both aimed at improving the overall decision performance of both expert and novice pilots. The overall aim is to improve aviation safety and reduce the number of accidents and incidents due to poor decision making and judgements.

Executive Summary

Judgement and decision making errors are the primary factor in over 50% of general aviation accidents. However, through appropriate training, decision skills may be improved, thereby reducing the number of decision-related accidents. This report examines both the theoretical and practical aspects of operational aircrew decision models, before examining various methods such as decision aiding and decision training to actually improve the decision skills of both expert and novice pilots. It has been discovered through several studies that experienced operators in their operational settings (including pilots) make many decisions using intuitive rather than analytical strategies. Naturalistic Decision Making (NDM) is one such style of intuitive decision making. Naturalistic decision making describes how experienced people make decisions in dynamic, naturalistic environments, under conditions of time pressure, dynamic goals, uncertain cues and high risk. All of these characteristics are found in aviation environments. The initial sections of this paper examine the principles of NDM, detailing theoretical perspectives that explain how the strategy works, and how such skills are acquired. Intuitive and analytical decision making strategies are compared and contrasted, highlighting the appropriateness of intuitive methods to operational settings such as aviation. Analytical styles of decision making are appropriate in certain situations. These situations are usually those where the decision maker has specific goals, minimal time constraints, and a complete, correct set of information. When these conditions are met, it may be possible to make an optimal decision. However, these conditions are rarely met in operational aviation environments. In such naturalistic environments, intuitive styles of decision making (such as NDM) are more appropriate. Indeed, these are the descriptive styles of decision making employed by most experienced operators, such as aircrew. Recognition-primed decision making (RPD) is one model of NDM involving decisions for which alternative courses of action are directly derived from the recognition of critical information and prior knowledge. These alternatives are serially evaluated, without the need for the comparison of options. Decision makers compare current events with previously experienced events and known rules stored in memory, recognising similarities to help them select appropriate reactions, judgments and decisions. Hence, the name, recognition-primed decision making. There are three basic phases to RPD; 1) situation recognition, 2) serial option evaluation, and 3) mental simulation. Because naturalistic environments are dynamic, the decision maker must continuously reassess the situation in order to reassess the appropriateness of situation models and hypotheses. Once the pilot understands a situation an acceptable course of action is often easily identified. Thus, situation assessment and awareness is a crucial and time consuming process for intuitive decision strategies. Situation assessment is also a prerequisite for good decision making, correlating positively with decision accuracy . Mental simulation is the final process carried out before (or during) a decision action is implemented. The decision maker evaluates the course of action by acting out the decision in his/her mind, imaging how a sequence of events might unfold within a given context. The simulation often includes the successive steps to be taken, the potential outcome of these steps, the problems that are likely to be encountered, and how these problems can be handled. As a result of the simulation, the decision maker either rejects, modifies or implements the action. Because NDM emphasises satisfactory rather than optimal decision outcomes, the mental simulation is generally a rapid procedure, acting as a go/no-go check. If time is not available for a complete mental simulation, the decision maker will simply implement the decision action that experience has generated as the most likely to be successful and make subsequent changes as necessary to maintain a satisfactory outcome. The importance of expertise and experience is explored throughout the paper, and the differences between expert and novice decision making are presented. It appears that only experienced pilots are capable of implementing RPD due to the reliance on memory traces and pattern recognition gained through experience, or possibly through training. The final section of this paper addresses the improvement of decision strategies (and outcomes) though decision aiding and decision instruction, and examines how some areas of NDM and RPD may be able to be taught to novices and inexperienced pilots. Although there has been little work done to formulate teaching strategies for NDM and RPD, some ideas are presented. It appears possible to teach novice pilots at least the basics of NDM, which they are then able to develop.

Back to the top