“Everyone should have realized that won’t work the way they wanted,” is a refrain heard all too often. However, engineers are not machines that convert an input (information) into an output (decision) based on purely objective factors. How we make decisions also counts.
Understanding how bad decisions get made can reduce their frequency. The psychology of decision-making, which already is used in safety analysis, provides insights. So, let’s look at some elements to see how they can influence design choices, maintenance criteria, control system design, and other areas. All these elements involve cognitive biases — factors that can affect decisions without us noticing or controlling them.
Anchoring is the process where the first idea or piece of information presented becomes the reference point for comparing all other ideas. Anchoring gives undue weight to the first information you have available. One example is where the initial proposed design choice for equipment becomes the favored option automatically, even if it was simply a throwaway idea mentioned casually.
Availability is the ease of keeping a particular choice in mind. The adage that every problem has a simple, easy-to-understand, wrong answer underscores the availability problem. If you’re not familiar with a topic, an uncomplicated analysis and solution requires less mental effort than more-complex proposals. The simple proposal is easier to keep in mind, so tends to be favored.
Framing is a response to how a question is asked. People vary from risk takers to risk avoiders. For example, in evaluating a new, but not identical, replacement part, the risk can be stated as “99% chance of working” or “1% chance of not working.” These statements represent the same probability but the risk-averse person is more likely to think a decision that has “99% chance of being right” is better than one that has “1% chance of being wrong,” even though that’s not true.
Polarization is a group decision problem where the group talks itself into a choice that is more extreme than anything a single individual would select. The extreme choice could be in any direction. One case is where participants in a design review intended to cut cost (a value-improvement meeting, for example) talk themselves into cutting project scope to a point well beyond where any individual would agree is realistic.
Habit is a reflexive response to a familiar situation. We’ll use a specific solution like a particular control configuration because we’ve always handled the situation that way. Even when the habit gives a valid technical answer, it can lead us to miss better solutions.
Representative failure is similar to habit. When we have a choice, we often make decisions based on how similar the situation looks to something we’ve already encountered. For example, we may have used a shell-and-tube exchanger in a “very comparable” service before, so let’s opt for the same type of exchanger now. However, perhaps we missed a difference that significantly affects the performance. It’s often difficult to judge the importance of small variations when we’re convinced the situation is similar.
Selective memory can contribute to problems. Over time, we may forget how difficult a particular supplier is to deal with or why we don’t buy those cheaper parts anymore. In other situations, memories can combine or points be forgotten. We think we’re doing the right thing but have mixed together multiple previous experiences and aren’t making as good a decision as we believe.
Deference to authority can undermine making good choices. Are you working on something your boss did when in your job? Does your company’s “expert” on the topic favor a specific option? Both of these can short-circuit a critical analysis of a decision. The authority of some people involved may overwhelm doubts others have.
Conformity is another name for peer pressure. Everyone else thinking something is a good idea can make you believe your doubts are unfounded. Mention any doubts you have and at least get them discussed.
These factors represent only some of the reasons poor choices get made. No one has a magic cure for any of these. The best decisions come when reasonable time is available and all involved are willing to evaluate the possible risks and advantages of specific choices. With effort, you can minimize a future engineer’s hindsight criticism of your own work.