What decision-making biases should you avoid in operations research?

1 Confirmation bias

Confirmation bias is the tendency to seek, interpret, and favor information that confirms your existing beliefs or hypotheses, while ignoring or rejecting evidence that contradicts them. This can lead to overconfidence, selective attention, and confirmation traps in OR. For example, you might ignore the results of a sensitivity analysis that show how your optimal solution changes under different scenarios, or you might disregard the feedback of stakeholders who have different perspectives or objectives. To avoid confirmation bias, you should actively seek out and challenge your assumptions, test alternative hypotheses, and consider multiple sources and types of data.

2 Framing effect

Framing effect is the influence of the way a problem or a choice is presented on the decision-maker's preferences and behavior. Different frames can emphasize or downplay certain aspects of the situation, such as risks, benefits, losses, or gains, and trigger different emotional responses. For example, you might be more willing to accept a gamble that has a 90% chance of winning $10 than one that has a 10% chance of losing $10, even though they have the same expected value. To avoid framing effect, you should use consistent and objective criteria to evaluate your options, avoid emotional or vague language, and reframe the problem from different angles.

4 Availability bias

Availability bias is the tendency to judge the likelihood or frequency of an event based on how easily you can recall or imagine examples of it, rather than on objective data or statistics. This can lead to overestimating or underestimating the probability or impact of certain risks, opportunities, or outcomes in OR. For example, you might overestimate the chance of a rare but catastrophic event that has recently occurred or received media attention, or you might underestimate the importance of a common but less salient factor that affects your system. To avoid availability bias, you should use data and evidence to support your judgments, avoid relying on anecdotes or personal experiences, and consider the base rates and sample sizes of your information.

5 Sunk cost fallacy

Sunk cost fallacy is the tendency to continue investing in a project, process, or option that has already incurred significant costs or losses, rather than cutting your losses and switching to a more profitable or efficient alternative. This can lead to wasting resources, time, and effort in OR. For example, you might stick to a suboptimal solution that you have spent a lot of time and money developing, or you might ignore new information or changes in the environment that make your original plan obsolete. To avoid sunk cost fallacy, you should focus on the future benefits and costs of your decisions, rather than the past ones, and be willing to revise or abandon your plans when necessary.

6 Hindsight bias

Hindsight bias is the tendency to overestimate your ability to predict or explain an outcome after it has occurred, and to believe that it was obvious or inevitable. This can lead to false confidence, learning failures, and blame attribution in OR. For example, you might think that you could have easily avoided a mistake or a failure that was actually unforeseeable or unavoidable, or you might judge your decisions or actions based on their results, rather than on their quality or rationality. To avoid hindsight bias, you should document and justify your assumptions, methods, and criteria before making decisions, evaluate your performance and outcomes objectively and constructively, and learn from your successes and failures.