Pilot see, Pilot do: Examining the predictors of pilots' risk management behaviour.


Drinkwater and Molesworth had carried out a research on examining the predictors of pilots' risk management behaviour in 2010. It has primarily studied how different components from 3 parameters (attitudes towards safe flying, risk perception, demographic variables) affect the pilots' risky flight behaviour respectively, and there are 2 ways in determining the risky flight behaviour was to study. They are (1) the time spent by pilots in flight and (2) the minimum altitude they descended the aircraft to during the flight. This article provides an analysis done on the original results.

Table one shows that there is no significant differences in the parameters to determine a pilot to continue or terminate his or her flight. From the result, only the first two components from the parameters, one from risk perception and the other from demographic variables, were shown to have a slight difference between the Go-Pilots and No-Go Pilots, for which the difference was still not significant enough to determine they can be predictors to pilots' going decision.

As a matter of fact, the study did not clearly examine and explain which parameter can effectively be used to predict whether the pilots will decide to go or not to go, and this had left a question that which parameter, or even if there are any new ones that have not been considered yet, is actually the one to drive the "GO/NO GO" decisions by pilots.

Parameter Component Go-Pilots Mean Go-Pilots s.d. No-Go Pilots Mean No-Go Pilots s.d. Is it significantly different?
Risk Perception Immediate Risk 75.55 8.89 81.41 10.53 NO
Demographic Variables Total Flight Hours (Flight Experience) 186.99 332.62 125.20 104.38 NO
Attitudes towards Safe Flying Self Confidence 2.30 0.41 2.25 0.42 NO
Table 1

Table two shows the correlation between each component of the parameters by focusing on the group of Go-Pilots only. It was found that only two relationships between the components of the parameters and pilots' risky flight behaviour are both correlated and meaningful to this study. It was calculated by the Pearson Product-Moment Correlation method.

The last row of table 2 shows that although they are negatively correlated, but neither of them is one of the parameter in determining the pilots' risky flight behaviour (Total Time in Flight or Minimum Altitude in Flight). Therefore only the relationship between the parameter's component and the determinant of risky flight behaviour is worth to consider in this study.

Relationship Correlative No. Type of correlation Is it Meaningful?
Self Confidence & Total Time in Flight -0.382 Moderately Negative Yes
Age & Minimum Altitude in Flight -0.330 Moderately Negative Yes
Safety Orientation & Nominal Risk -0.522 Moderately Negative NO
Table 2


Research approach

The study was an exploratory research to examine if each parameter is a predictor of pilots' risk management behaviour.


A sample of 56 student pilots from University of New South Wales and other flying schools at Bankstown aerodrome were recruited. Their average age was 20.02 years old and average total flight experience was about 165 hours, while the average flight experience in the past 90 days was about 21.8 hours.

Design & Variables

The study was to study the relationship between the three parameters which are (1) Attitudes towards Safe Flying, (2) Risk Perception, (3) Demographic variables and the risky flight behaviour which is measured by (1) Total flight time to complete the task and (2) Minimum altitude descended during the flight. The three parameters were independent variables in the study, while the risky flight behaviour was the dependent variable.

In addition to the first two parameters, the Attitudes towards Safe Flying (ASAS) was comprised of three components, and each component was rated by the participant on a five point Likert Scale (i.e. 1 = Strongly Disagree; 2 = Disagree; 3 = Neutral; 4 = Agree; 5 = Strongly Agree). Thus in the study, participants were asked to rate a point on each statement from the respective component. On the other hand, "Risk Perception" was measured by two scales, for which the participants had to rate on a scale of 1-100 ( 100 = high risk ) on perceiving the risk from the (1) first person and (2) third person of view.


The study was conducted in a flight simulator.

First of all, pre-test experimental design was carried out that the participants were asked to sign the consent forms and finished rating on the scales of the parameters.

After that, they were asked to fly circuits at Moruya aerodrome until a moment that they were on the downwind leg of the last circuit with low fuel on board. At this time, the operator would ask them to fly northwards to a place to conduct a search.

This study was to see if the pilots elected to go and search. If they decided to go, it was to examine how much time they would spent more in the remaining flight and the minimum altitude they would descend to in the situation of low fuel on board and low flying has to be necessarily carried out.

Data analysis

The original research article had provided the mean with standard deviation, and the correlation numbers in the data.

Generalization potential

The sample used by the study was comprised of solely student pilots and they are mainly from University of New South Wales, hence the findings were potentially valid to this population type only.

This study had given people a brief idea that how self confidence (Attitude towards safe flying) and age (Demographic variables) of pilots predict how risky their behaviours are in flights. However, this study should be further carried out to examine the fundamental question: What is the predictor for pilots deciding to continue or terminate their flights in risky situations.

1. Drinkwater, J. L., & Molesworth, B. R. (2010). Pilot see, pilot do: Examining the predictors of pilots’ risk management behaviour. Safety Science, 48(10), 1445-1451.

Contributors to this page

Alvin Fong

Unless otherwise stated, the content of this page is licensed under Creative Commons Attribution-ShareAlike 3.0 License