Should an approximate report become one exact number?
A public post saying setup took between twenty and forty minutes gives an interval, not a measured thirty-minute duration. Replacing it with a midpoint may simplify a table while erasing the author’s uncertainty. Preserve ranges and one-sided bounds unless a clearly labeled analysis genuinely requires a derived value and its assumptions are defensible.
Preserve timing ranges and at-least bounds in supplied setup accounts. Keep original wording and label any justified midpoint transformation as derived rather than observed.Identify the quantity type
Classify each reported value as exact as stated, approximate, bounded or an interval. Include the original phrase beside structured fields. Exclude an assumption that round numbers are measured precisely; the research can record how the author expressed a quantity without certifying the underlying measurement process or improving its precision through formatting alone.
Store the supported bounds
For a stated interval, retain lower and upper values with the unit. For at least, retain the lower bound and leave the upper bound unknown. Do not enter a missing bound as zero or infinity in a public summary without explaining its technical meaning, and do not convert an approximation into an exact point estimate by dropping its qualifier.
Check downstream calculations
Review whether aggregation or sorting requires a point value. If so, either use a method that respects the bounds or avoid the calculation. A midpoint is a derived assumption, not a reported observation. Keep original values available so readers can assess whether the chosen transformation changes the conclusion or hides meaningful uncertainty in the evidence.
Deliver precision-aware findings
Present ranges and bounds directly when possible, with exact reviewed counts of the kinds of reports available. State that public timing accounts are self-reported rather than controlled measurements. If the owner needs a precise setup benchmark, identify the need for an appropriate study rather than presenting rounded or midpoint-filled anecdotes as an exact performance distribution.
Keep the boundaries clear.
- A range contains uncertainty that should not disappear when the account enters a table.
- Examples illustrate evidence interpretation, not live product tests or guaranteed platform access.
Start with a read-only check: ask your agent to report actual source coverage before collecting records or exporting an audience.
Primary references
Documentation reviewed 2026-11-13. Product capabilities, client availability and plans can change. Third-party marks identify their products, not an affiliation.