Formal methods and statistical methods are complements
The introduction says a team can show that a system keeps performing when local conditions vary by using formal methods, as Part 2 illustrates, and that statistical work adds a complement because it reaches across more of those conditions.
The requirement in plain English
The introduction says a team can show that a system keeps performing when local conditions vary by using formal methods, as Part 2 illustrates, and that statistical work adds a complement because it reaches across more of those conditions.
What the public record says
The introduction sets formal work against statistical work: the first is illustrated in 24029-2, the second is described as a complement that covers more of the varying conditions.
Why it matters to you
The complement is the argument for the standard. A formal method can establish a property over a defined region with a strong guarantee; a statistical method can sample a space too large or too poorly specified for a proof. Neither is a substitute for the other, and a pack that presents a sample as a proof will fail review. The sentence also tells you how to structure the evidence: one section for what was proved and where, one for what was sampled and how, and one for how the two were combined.
Evidence a reviewer asks for
- A two-column method table: formal results, statistical results, and the claim each supports.
- The input region for each formal result and the sampling frame for each statistical result.
- A statement of how the two were combined, or why only one was used.
Checks to run this month
- Does the pack claim a proof where only a sample exists?
- Does it claim a sample covers a region a proof did not?
- Is the combination rule written down?
Dates
FDIS registered 2026-08-21.
Cross-references
- Part 2, the formal-methods companion
- Fitting the statistical line into a wider scheme
- A statistical approach to robustness
What this page does not state
The public text does not say which method to use for which property.