§4.3

Uncertainty for Managers

Executives do not need confidence intervals because they enjoy statistical ritual. They need them because decisions happen before certainty arrives. A visual interval is a management tool: it says whether an apparent difference is large enough, precise enough, and stable enough to deserve attention.

The executive question: when should uncertainty change the decision?

The Progresso soup pattern is visually strong. Share falls in warm months and rises in the soup season. But a manager still needs uncertainty language because the data is a panel of store-months, not a magical census of every possible market condition. Store coverage changes over time. Store-level behavior varies. A chart should show the finding and the limits around the finding.

Figure 1 shows monthly mean Progresso share across store-months with intervals. The intervals are narrow because there are many store-month observations. That is useful, but it is not the same as causal certainty.

Seasonal share intervals are narrow — but narrow is not causal

Dot area encodes coverage: active store-months range from 6,643 to 7,984. A precise mean built on thin coverage still deserves a second look.

Region × season on one shared axis

A real dot-and-whisker on a single share axis. Winter share sits above non-winter in every region, but the East operates at a different level entirely.

Figure 1. Progresso share is reliably lower in the summer months, but the intervals describe store-month variation, not the causal effect of price.

Figure 1 should be read in two passes. First, read the pattern: Progresso share is lower in summer and higher in the broader soup season. Second, read the caveat: the interval is about observed store-month variation. It does not answer what would happen if Progresso changed price tomorrow.

This distinction is central to the rest of the book. Statistical precision does not repair a weak comparison. A very narrow interval around a biased estimate is still a biased estimate.

Coverage belongs near the chart

The soup panel is unbalanced. Stores enter and exit the observed panel; monthly active store counts vary. That does not make the data unusable. It means the coverage note belongs near the figure, not buried in an appendix.

Good visualization treats coverage as part of the evidence. If a December point is based on many more active stores than a June point, the reader should know that before interpreting the chart as pure seasonality. The same principle will matter later in model monitoring, survey analysis, A/B tests, and AI evaluation.

Figure 2. Uncertainty language is useful only when the chart explains what kind of uncertainty the interval represents.

Rule	Interpretation
Rule 1	Narrow intervals can still be biased if the comparison is wrong.
Rule 2	Coverage changes over time, so active store counts belong near the chart.
Rule 3	Store-month intervals describe observed variation, not the causal effect of price.

Figure 2 is deliberately close to the chart rather than hidden in a technical appendix. It teaches the reporting habit: every interval should answer what is varying, what is being averaged, and which decision would change if the interval were wider.