Add a Little Structure to Your Decision-Making

Mark C. Otto and MinJae Lee

Decision-making is both a process and a matter of luck. In our previous articles, we argued that good decisions are judged by the quality of the process. Structured Decision Making, here referred to as SDM, provides a practical framework for putting that philosophy into action.

Think of Atul Gawande’s Checklist Manifesto. Running down a list keeps us on track. The list tells us what to look for in the little corners and frees us to be more creative because the routine stuff is laid out for us. The list is the acronym PrOACT: Problem framing, listing Objectives, creating Alternatives, estimating Consequences, and working through the Tradeoffs.

The Alliance for Decision Education provides an accessible introduction to these steps and practical examples of how they can be applied. If you can explain and reflect on your process, it will be clearer to you, and you will be able to improve your decision-making next time. Finally, you need to act on the decision. Duh, but how many times have we sweated over what to do until we feel we have already done it?

Consider a man buying a car. When it is just him and the dealer, framing is easy: He wants something stylish, cool, fast, red, and affordable-ish. On impulse, an expensive, flashy red Miata wins.

Now the man is reminded he is a husband, and his wife adds safety, reliability, and room for a car seat, while objecting to the flashy objective. New alternatives emerge. The more expensive Miata loses out to the Subaru’s efficiency, reliability, and safety. The objectives expand, and the consequences of each remaining alternative are weighed against that longer list. Groups fight to the death over alternatives but can talk civilly about values or objectives. They can do this in marriage, too.

Ralph Keeney suggests you can almost always find more objectives than you started with, and the ones you identify later often decide the choice. More objectives and alternatives emerge when we give up the status quo. It is an acquired skill that requires creativity. Keeney’s favorite example is choosing a restaurant to meet a client. You think about cost, cuisine, and location. Halfway through the meal, you realize the thing that mattered was whether you could hear each other over the noise. That objective was not on your list when you started; now it is the one that decides where you meet next time.

With more objectives and alternatives, we need the SDM structure. Consequences are the predictions of how well each alternative fulfills each objective, and prediction is entirely within the statistician’s domain, as is weighting. We weight the objectives, multiply them by the consequence predictions, and sum the products for each alternative to calculate utilities. Adjusting the consequences and weights until they make sense is how we work through the tradeoffs.

With kids, objectives multiply again. Safety was on the list; now it is at the top. Room for a car seat becomes room for two car seats, then for a stroller, then for the whole under-eight soccer team when it is your carpool week. Sliding doors, once a joke, are now their own objective. The alternative set has gone through three revisions and is closing in on a minivan.

Next, the wife realizes, “We live in DC. Do we need a car at all?” This is not a small question. The Metro runs. Zipcar is a block away. Uber is everywhere. Rental cars are available for beach week. A stroller-friendly neighborhood beats a stroller-friendly SUV for at least half of their trips—status quo be damned.

Now we are back at framing—the Pr of PrOACT—but with a much longer list of objectives and a much richer sense of alternatives. This is where the structure earns its keep. Without it, this conversation goes nowhere or, worse, devolves into an argument. With it, they can list their objectives together (get to work, get the kid to daycare, get to soccer, get to grandma’s four hours away, keep the monthly costs sane, avoid the parking headaches) and score the alternatives (own a minivan, own something smaller and rent for soccer weekends, own nothing and rely on transit plus Zipcar plus the occasional rental) against those objectives honestly. They may still buy a car. They may not. The point is that the decision they make will be a decision, not a default.

While car buying is an informative example, the same structured thinking applies to scientific research, clinical care, and the many collaborative decisions statisticians guide clients and colleagues through every day. Statisticians already practice much of SDM, often without realizing it. Framing is study design. Objectives are reflected in the research questions and outcome variables. Alternatives are the hypothesis space. Consequences are predictions and their uncertainties. Tradeoffs are familiar to anyone who has balanced multiple outcomes, optimized competing objectives, or developed a composite measure. The value is not that statisticians learn a completely new process, but that SDM makes an intuitive process explicit, transparent, and repeatable. What SDM adds is the discipline of doing each of these steps deliberately, on paper, with everyone in the room—instead of privately, in your head, or in front of the salesman at the dealership.

In collaborative research, we often follow these same steps without explicitly calling them SDM. We begin by clarifying the scientific question, identifying the outcomes that matter most, considering alternative study designs, anticipating the consequences of each design, and weighing tradeoffs such as feasibility, cost, precision, and participant burden. SDM provides a common language and framework that make this process explicit, transparent, and collaborative, helping multidisciplinary teams make better-informed decisions together. This is Gawande’s list for decision-making. Use it well.

Next time, we move beyond framing the problem to the difficult step of identifying the values and objectives that guide our decisions.

Mark Otto

Mark C. Otto is a retired, but not retiring, US Census Bureau and US Fish and Wildlife Service statistical scientist. He is chair of the American Statistical Association Committee on Applied Statisticians and has served as the president of the Washington Statistical Society and as the Membership Council vice chair. (Consider volunteering.) He is an ASA Fellow. In these interesting times, he works toward long-term cultural change with the Alliance for Decision Education and Braver Angels.

MinJae Lee

MinJae Lee is a biostatistician and professor at McGovern Medical School, UT Health Houston, where she leads the division of clinical and translational sciences. Her research focuses on developing and applying innovative statistical methods to address real-world research challenges. She is a member of the ASA Committee on Applied Statisticians and serves on the Nature Medicine Statistical Advisory Panel.

Deciding Factors, a column by Mark Otto, explores decision analysis and how statistical scientists can apply their expertise to make better choices in their careers, communities, and personal lives.