Lesson 5 of 6 · Uncertainty, calibration & revision
Understand a measurement with room around it
8 minute lesson · 3 checks along the way
Your goal for this lessonExplain why a precise-looking value needs an uncertainty description, and distinguish a revised estimate from a changed physical object.
Follow a linked name to explore its archive card. “Back to your lesson” brings you to the same part of the reading.
Words with a dotted underline open a short definition right here in the lesson.
A number needs its measurement story
A display can show many decimal places without establishing that every digit is reliable. A measurement depends on an instrument, a method, and conditions that are never known perfectly. describes limitations on what can reasonably be concluded from the result. It is not an admission that anything goes, and it is not automatically evidence that someone made a careless mistake. Its size and meaning need explanation.
Two important concerns are scatter and shared offsets. Repeated readings may vary because of changing conditions or noise. They may also cluster tightly around a value displaced by a problem. Repeating a measurement can help characterize variation, but averaging does not automatically remove a bias common to every reading. A separate reference or a different measurement method can expose a weakness that repetition alone leaves hidden.
Check 1 of 3
10 pointsUse the reading above. You can try again without losing points.
Work through a decision
Read the interval before comparing results
Use this invented classroom example: a crater-width estimate is 10 kilometers, with an explicitly stated plausible range from 8 to 12 kilometers. A second method gives 11 kilometers with a stated range from 10 to 12. The central numbers differ, but their difference alone does not establish that the methods conflict. You need to know how those ranges were constructed and what assumptions they include.
Neither range should be read as a guarantee that the true width lies inside it. Real uncertainty statements may describe a standard uncertainty or a coverage interval with specified assumptions. Nor does overlapping a pair of plotted bars prove that two quantities are identical. Here the ranges simply illustrate why comparing naked central values loses information. A formal comparison would also consider correlations and the measurement models.
Compare the evidence
A revised estimate and a measurement principle
Use a dated case together with definitions that explain how measurements can improve.
Apophis orbit update
- When and where
- NASA assessment: March 26, 2021 · Radar and optical observations of Apophis
- Purpose
- Refine the orbit and assess previously possible encounters.
- What it supports
- Additional measurements and orbit analysis excluded a possible 2068 Earth impact in that assessment.
- What remains uncertain
- The observations changed the allowed orbit estimates; they were not an asteroid-deflection maneuver.
Measurement terminology
- When and where
- NIST TN 1297, Appendix D1 · National Institute of Standards and Technology
- Purpose
- Distinguish repeatability, measurement error, and systematic effects.
- What it supports
- Closely repeated readings can coexist with an offset shared by the method.
- What remains uncertain
- A terminology guide does not supply an orbit, a particular risk estimate, or an uncertainty budget for this asteroid.
Would repeating the same biased reading indefinitely replace a new observation that checks the calibration?
Compare your reasoning
No. Repetition can characterize scatter while preserving a common offset. A measurement that tests a different weakness can add information that repetition lacks. The Apophis case concerns new constraints on an estimate, not a physical redirection caused by measuring it.
Check 2 of 3
10 pointsUse the reading above. You can try again without losing points.
New observations can change an orbit estimate
Apophis provides a dated example of knowledge improving. In March 2021, NASA reported that additional radar observations, combined with optical measurements and orbit analysis, ruled out a possible Earth impact in 2068. The observations tightened the estimated orbit. This lesson uses that 2021 result as a historical case; the linked report supplies its dates and stated interval rather than an undated general claim.
The radar campaign did not steer the asteroid away. It changed which paths remained consistent with the evidence. Imagine drawing several possible curves through a few uncertain position measurements. Additional well-placed observations can rule out some curves and narrow the remaining set. Real orbit fitting is more sophisticated than this sketch, but the distinction remains: the object’s trajectory and our estimate of it are different things.
Explain why your conclusion changed
A useful revision identifies its cause. Did a new become available, an instrument calibration improve, a processing error get corrected, or a assumption change? Those developments need not have the same implications. A corrected unit mistake differs from a more precise measurement that narrows an earlier uncertainty. Reporting what changed helps readers understand whether an earlier conclusion was unsupported or simply less specific.
When comparing two space headlines, preserve the publication date, measured quantity, method, and stated uncertainty. Ask whether the newer evidence addresses the actual source of disagreement. Independent agreement can strengthen confidence, while a persistent discrepancy can motivate better tests. Responsible science neither treats every estimate as permanent nor discards all previous knowledge whenever one number changes. It explains what the evidence now supports and what remains unresolved.
Check 3 of 3
10 pointsUse the reading above. You can try again without losing points.
Put it together
Your lesson progress
Each check connects to a part of the story. Complete all 3 to earn your 25-point lesson bonus.
Go to the source
Explore the evidence behind this lesson. These optional readings and videos open in a new tab.
- Read: NASA: the March 2021 Apophis assessmentA dated example of an updated estimate, not a live risk bulletin.
- Read: NASA/JPL: radar observations used in the campaign
- Read: NIST: measurement terminology and uncertaintyOptional technical reference; compare repeatability with measurement error and systematic effects.