Decision Uncertainty
Decision uncertainty concerns what we do not know about a choice and its possible outcomes. Banasiewicz’s idea of a degree of knowability asks how much of that uncertainty useful facts and probability estimates could reduce.
THE IDEA
Decision Uncertainty
Organizational decisions often depend on outcomes we cannot know in advance. Evidence can reduce some of that uncertainty. Banasiewicz calls the extent to which facts and probability estimates can help the degree of knowability. A brand manager considering an extension, for example, might learn from similar launches or use customer data to estimate demand. The question is what can be learned before making the choice.
Knowability depends on the scope of the question and the availability of dependable information. A focused question may be easier to investigate than a broad one. During restructuring, a manager might understand Employee X’s concerns better than the reactions of an entire organization. Even that employee’s response, however, can depend on colleagues, management, and how the changes unfold.
Information availability also requires more than a yes-or-no answer. Has anyone sought the relevant data? Can the organization analyze it? Will decision-makers use an estimate rather than demand a certainty the evidence cannot provide? Missing probabilistic evidence may be inaccessible, or the effort to produce it may not have been made.1
Using probabilistic evidence calls for competence in three related areas:
- Computational thinking: understanding how data were obtained, selected, transformed, and checked before an estimate was produced.
- Inferential thinking: understanding how conclusions are drawn from data, including the assumptions and limits of tools such as hypothesis tests.
- Evidentiary thinking: deciding what an analysis supports and how much weight it deserves in the choice at hand.
These skills matter when results agree with our beliefs as well as when they challenge them. An agreeable result can escape scrutiny; an unwelcome one can prompt us to reject the data immediately. Understanding how an analysis was produced gives us grounds to assess either result. It should help us identify warranted criticism and useful evidence, rather than merely make a conclusion easier to accept.
Weather forecasts offer a familiar example of probabilistic evidence. They express the likelihood of precipitation on a scale from 0% to 100%; Banasiewicz traces the US National Weather Service’s introduction of these forecasts to 1965. Such estimates do not remove uncertainty. Their value depends on what they predict, how they were produced, and whether they are dependable enough for the decision we need to make.
When to use it
A startup founder evaluating whether to pivot based on market data that conflicts with initial assumptions.
A healthcare administrator deciding whether to adopt a new treatment protocol using statistical evidence from clinical trials.
A project manager assessing risks during a complex software rollout by distinguishing between scope of ambiguity and availability of data.
An investor analyzing probabilistic returns from different asset classes to build a diversified portfolio.
FROM THEORY TO EVERYDAY LIFE
In practice
Brand Extension Decision
A brand manager considering a brand extension faces uncertainty due to unfamiliarity with outcomes of similar efforts by other brands in the sector and lack of access to customer analytics-based demand projections.
Organizational Restructuring Reaction
During restructuring, a manager may have enough information to understand Employee X’s concerns while knowing much less about the whole department’s response. The individual’s reaction can still depend on colleagues and on how the restructuring proceeds.
Weather Forecast as Probabilistic Evidence
Weather forecasts express precipitation likelihood from 0% to 100%, serving as a ubiquitous example of probabilistic estimates used in decision-making.
A MOMENT TO REFLECT
Put it to work.
Think of a recent significant decision you faced. What aspects of uncertainty influenced your choice, and how did you reduce that uncertainty?
How might the degree of knowability differ between a personal decision (e.g., choosing a career move) and an organizational decision (e.g., launching a new product)?
When encountering probabilistic evidence that contradicts your prior beliefs, what steps can you take to ensure you evaluate it objectively rather than dismissing it defensively?
CONTINUE THE CONVERSATION
Source texts
- Evidence-Based Decision-Making How to Leverage Available Data and Avoid Cognitive Biases (Andrew D. Banasiewicz)