Dynamic decisions

2. Managing these threats appropriately (which includes avoiding them),

3. And having procedures in place to mitigate the effects of unidentified or mismanaged threats that get through the first two filters
However there have been some recent incidents around the world where relatively minor non routine operations or technical abnormalities, that wouldn’t be considered a safety issue, have degenerated into something far more significant.

This raises the question as to why these threats being missed or mismanaged. Obviously there is no simple answer, but an understanding of the dynamic nature of aviation decision making may help illuminate some areas of concern.

A good starting point for this is to consider some of the basic inputs into our decision making processes. No one gets out of bed in the morning thinking that they are going to make so many decisions that day, rather decision are made in order to achieve our goals. So it is seems commonsense that goals must be considered an input. Next we have to understand the situation that we are in, and how it is likely to develop as we pursue our goals – commonly referred to as “situational awareness”. If it’s raining heavily and our goal is to get to Fred or Julie’s place by lunch time we probably would discount going by push-bike. Finally risk needs to be considered. Heavy rain and a driving a car with bald tyres isn’t a conservative choice – so you would be advised to think of some alternative way of achieving your goals. Perhaps waiting for the storm to pass, or alternatively changing you short term goals slightly by asking them to come over to your house instead, or putting things off for another day which still meets your longer term goal of seeing them.

This is a very simply scenario but it illustrates not only that goals, situational awareness and risk management are basic inputs into our decisions, but also that none of these inputs exist in isolation. We usually have multiple goals which are often competing, hence needing to be prioritised; what we see in a situation is dependent on our plans and risk appetite and dynamically changes; and finally, whilst risk is always present whether we identify it or not, the options to manage it appropriately are also dynamic simply by nature of the dynamic environment.

This can be illustrated by the following simple relationship –

This simple equation is meant to illustrate that our inputs are interdependent; if we change one of the inputs then one (or both) of the others must also change for the same quality of decision. For example if we encounter a technical or operational problem, then for the same quality of decision we need to reduce our level of risk, ie make our goals more conservative. This may mean we will need to abandon (or make subordinate) certain goals eg OTP or passenger comfort and concentrate more on primary safety issues. Alternatively if the terminal WX improves a tighter approach can be flown for the same level of risk. It is probably worth repeating, that since the situation in aviation is always dynamic, the other inputs will need to change appropriately – whether you recognise this or not.

However it is naive to look at this model without considering the human element in each of the inputs.

Firstly, consider the term “appropriate goals”. Several years ago a study of multiyear NTSB records showed that, what are termed “plan continuation errors”, were a causal factor in the majority of aviation accidents over the period. As the term indicates, in these accidents the pilots continued with inappropriate goals (or goal priorities) when the situation demanded a change. Examples include – continuing with unstable approaches, making approaches into thunderstorms, not executing missed approaches, etc. It is well beyond the scope of this short article to address this and other questions fully, but in terms of inappropriate goals selection, several possible reasons can be readily identified.

• Social Proof – in ambiguous situations or when you are unsure of what to do there is a strong tendency to look to others for an answer, especially if you consider them ‘similar’ to yourself (or an ‘authority’) – e.g. everyone else is continuing to make approaches, we should too; the captain seems happy, I’ll keep quiet; etc.

• Our asymmetric risk appetite – because we feel losses twice as much as gains we have a tendency to be risk averse in terms of gains, but risk seeking in terms of losses. This gives rise to what is termed loss aversion, the status quo bias, and the endowment effect – which all apply to ideas as well as goods. What these imply is that since changing a plan is considered a loss we have a strong tendency to stick to our old plans (and related goals) rather than discard them.

• Commitment and Consistency – having publically declared our intention (e.g. briefed the approach and post landing activities; accepted vectors onto final approach, etc) this psychological driver again means we will often attempt to avoid changing plans as this would be seen as being inconsistent (a ‘scatterbrain’).

• These last two can lead to what is termed “escalation of commitment” where we continue to avoid changing our goals against mounting evidence we should, often even finding reasons to ignore or explain away disconfirming evidence. Whilst there is little specific research to confirm this as a significant causal factor, anecdotally it would appear that it is.

Next, consider Adequate Situational Awareness. Some models of situational awareness, or situational assessment to be more precise, are based on a passive ingestion of ‘data elements’ which we process into knowledge. However if a farmer was to look out the same rainy window you did previously, depending on the time of year, he or she might interpret it as a good time to start planting, or a time to buy cattle to put onto a spoiled crop – a totally different interpretation to yours. Anais Nin neatly summed this up in her well known quote – “we don’t see the world as it is, but rather as we are”. In other words what we recognise in a situation is largely a function of our goals and risk assessment. Moreover how you view you situation will also impact on what ‘data elements’ you notice and / or search for, and how they are interpreted.

It is worth mentioning that expertise, which is not the same thing as experience, allows experts to recognise subtle clues that may be invisible to novices. Additionally because of experts greater sense of what should happen they are able to quickly recognise when things aren’t happening that should, meaning their existing assessment of the situational is incorrect or deficient. This then instigates a revised data search, including searching for non obvious or missing data, allowing them to reach a new understanding of the situation. This revised model then guides further information search and interpretation. In short situational assessment is an iterative cycle where your existing situational awareness guides your data search which then allows you to confirm or update your situational awareness, which then guides your data search etc., a process that is very dependent on expertise.

Finally consider the Risk Management component. Whilst the first two steps in Threat and Error management cover some essential risk management objectives they miss two equally important rules of risk management –

• Don’t risk a lot for a little

• Never risk what you can’t afford to lose

Practically in aviation these are usually covered at the front of any operations manual by such terms as ‘safety is our number one priority’. However in line operations these rules sometimes seem to be pushed into the background compared to more pressing operational goals and priorities. Another important factor in our ability to adequately assess and manage risk is termed “frequency gambling”. Essentially this implies that rather than properly assessing operational risks on its own merit, we use the mental shortcut that we’ve been in a similar situation before and ‘gotten away with it’, so the same outcome should be expected. Over time this can produce an inappropriate insensitivity to certain risks – “familiarity breeds contempt”

In summary – The environment we are operating in is dynamic and constantly changing. As a result if we want to maintain the same quality of decisions in a more difficult operational environment (eg non normal operation) we will inevitably need to alter our goals (or the hierarchy of competing goals) if we don’t want to increase our level of risk. The alternative of course is to make the operating environment less challenging by, for example, entering a holding pattern, making a missed approach, asking for a more extended approach etc. In addition we have several powerful psychological propensities hardwired into our thinking that operate at subconscious level, often to the detriment of the quality of our decisions. As such it appears that too often “threat and error management” is seen as something static to be discussed in briefings, whilst in reality the environment is dynamic with each element of our decision making processes, and inbuilt psychological biases, also dynamically adding to or subtracting from threats to our operation.