Choose your reading route
eduKateSG · Why Mathematics? · Zeta Scores, Combined Standard Uncertainty and Laboratory Claims
Start with the question, audit the worked example, or jump to the assumptions and practice. Every chapter stays open for uninterrupted reading.
Understand the question
Define the quantity and follow the mechanism.Route 2 · Chapters 4–7
Reproduce the example
Audit the numbers one step at a time.Route 3 · Chapters 8–13
Check assumptions
See where design and calibration matter.Route 4 · Chapters 14–18
Explore applications
Connect the method to a concrete investigation.Route 5 · Chapters 19–22
Study and verify
Practise, inspect failure cases and read further.
Full chapter index · Practice and checking · Mathematics Learning Hub
Two results can be close in ordinary units yet inconsistent with the small uncertainties claimed for them. The zeta score compares a participant result with an assigned value using the combined standard uncertainties of both. It tests whether the observed difference fits the uncertainty statements.
This is a powerful reason to learn mathematics: a difference and its uncertainty must be interpreted together. Zeta can reveal an understated uncertainty budget even when a proficiency z score looks satisfactory.
Full chapter index
Chapters 1–3 · Understand the question
Chapters 4–7 · Reproduce the example
Chapters 8–13 · Check assumptions
Chapters 14–18 · Explore applications
Chapters 19–22 · Study and verify
Let x be the participant result with standard uncertainty u(x), and X the assigned value with standard uncertainty u(X). Under the simple independent form, zeta = (x – X)/square root of[u(x) squared + u(X) squared].
ISO 13528 defines zeta as a standardised performance measure using the participant result, assigned value and their combined standard uncertainties.
Suppose x = 104 units, X = 100 units, u(x) = 1.5 units and u(X) = 0.5 unit. The numerator is 4 units.
The denominator is square root of 1.5 squared plus 0.5 squared = square root of 2.5, about 1.5811 units. Zeta is 4/1.5811, approximately 2.5298.
| x | X | u(x) | u(X) |
|---|---|---|---|
| 104 | 100 | 1.5 | 0.5 |
The positive sign indicates that x is above X. The magnitude shows the difference in combined-standard-uncertainty units.
Providers specify evaluation conventions. IAEA reports commonly distinguish satisfactory, questionable and unsatisfactory regions using magnitudes around 2 and 3, while stressing investigation of uncertainty estimates.
| Step | Calculation | Result |
|---|---|---|
| Difference | 104−100 | 4 |
| Denominator | √(1.5²+0.5²) | 1.5811 |
| Zeta | 4/1.5811 | 2.5298 |
If sigma_pt were 3 units, the ordinary z score for the same 4-unit difference would be 1.333. The zeta magnitude is 2.530 because the combined stated standard uncertainty is only 1.581.
The two scores answer different questions: fitness-for-purpose performance versus consistency with reported uncertainties.
Participant uncertainty should include relevant contributions from calibration, preparation, recovery, blanks, environment, resolution and repeatability where applicable. Using only replicate standard deviation can be too narrow.
An IAEA comparison report notes that questionable zeta results can indicate underestimated uncertainties and that replicate precision often omits major contributions.
Zeta uses standard uncertainties in the denominator. If a laboratory reports expanded uncertainty U with coverage factor k, convert to u = U/k when appropriate and documented.
Never combine one expanded uncertainty with one standard uncertainty without conversion.
Normalised error often combines expanded uncertainties and applies a magnitude-one criterion in metrology comparisons. Zeta uses combined standard uncertainties and proficiency-testing interpretation conventions.
The algebra is closely related, but the scale and decision rule differ.
z-prime uses sigma_pt together with u(X). Zeta replaces sigma_pt with the participant's u(x). Therefore a result can have an acceptable z′ but a concerning zeta if the participant claims very small uncertainty.
Keep the denominator labels visible.
If x and X share a calibration source, the variance of their difference includes -2cov(x,X). Positive correlation can reduce the uncertainty of the difference because common effects cancel.
Use the provider's covariance treatment. Independence is not guaranteed by two separate reports.
A large zeta can arise from a shifted result, understated uncertainty, reference-value problems, item instability or ignored covariance. It does not reveal the cause.
Investigate raw data, uncertainty budgets, units, calibration, preparation and provider information.
Calculate from full-precision values. If zeta is 2.996, displaying 3.0 can change apparent category depending on the rule. The provider should state rounding and inclusivity.
Keep both signed and absolute values.
Plot zeta by analyte or level. Repeated positive scores may indicate systematic bias; large values only at low levels may indicate a concentration-dependent uncertainty problem.
Avoid averaging signed scores that cancel.
Two groups report a result and a carefully justified standard uncertainty for a shared safe measurement. Calculate the difference and combined uncertainty scale.
Then enlarge one uncertainty budget by adding a missing contribution and observe how zeta changes. Explain why honest uncertainty is better than a cosmetically small number.
Use columns for x, X, u(x), u(X), covariance assumption, denominator, zeta and category. Validate units and coverage conversions.
Store the uncertainty budget or a link to it beside the score.
Validate non-negative uncertainties and a positive variance of the difference. Use a covariance matrix when shared components exist. Test the worked example and unit-conversion invariance.
Return each denominator contribution for audit.
Students can connect square roots, variances, uncertainty and fair comparison. Ask why a small claimed uncertainty raises the magnitude when the difference stays fixed.
Parents can reinforce that uncertainty is not ignorance to hide; it is quantified knowledge about limits.
Zeta scores appear in proficiency testing, analytical laboratories, calibration and quality assurance. The broader skill is checking whether a claim of precision matches observed differences.
Use the Mathematics Learning Hub and How Mathematics Works for connected foundations.
The normalised-error article uses expanded uncertainties in a related comparison. The Birge-ratio article checks collective scatter against uncertainty claims.
Set x = 102 and recalculate. Double every result and uncertainty in a unit conversion and verify invariance. Convert U(x)=3 with k=2 to u(x). Introduce positive covariance and calculate the changed denominator under the stated model.
Then compare zeta with a z score using sigma_pt = 3.
### Is zeta the same as z potential? No. Zeta here is a proficiency-testing performance score, unrelated to electrokinetic zeta potential.
### Does a large zeta prove the uncertainty is wrong? No. It signals inconsistency requiring investigation; the result, reference, dependence and uncertainty model all need review.
### Why use standard uncertainties? They provide a common one-standard-deviation scale for variance combination.
### Can a satisfactory z score coexist with a large zeta? Yes. Fitness-for-purpose dispersion and reported measurement uncertainty are different denominators.
Use the official ISO 13528:2022 definitions, the Eurachem PT guide, the IAEA TECDOC proficiency analysis and the 2026 IAEA interlaboratory comparison report. The worked values are original teaching data.
Zeta makes uncertainty accountable. It asks whether a reported difference is believable given what both sides say they know. That is mathematics serving honesty, comparability and improvement.
Extended practice laboratory
Practice case 1: Move the participant result
Setup: Use the article’s zeta-score example and change one declared input for case 1. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 2: Move the assigned value
Setup: Use the article’s zeta-score example and change one declared input for case 2. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 3: Increase participant uncertainty
Setup: Use the article’s zeta-score example and change one declared input for case 3. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 4: Decrease participant uncertainty
Setup: Use the article’s zeta-score example and change one declared input for case 4. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 5: Increase reference uncertainty
Setup: Use the article’s zeta-score example and change one declared input for case 5. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 6: Convert expanded to standard uncertainty
Setup: Use the article’s zeta-score example and change one declared input for case 6. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 7: Double all unit values
Setup: Use the article’s zeta-score example and change one declared input for case 7. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 8: Reverse subtraction
Setup: Use the article’s zeta-score example and change one declared input for case 8. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 9: Add positive covariance
Setup: Use the article’s zeta-score example and change one declared input for case 9. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 10: Add negative covariance
Setup: Use the article’s zeta-score example and change one declared input for case 10. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 11: Compare PT z
Setup: Use the article’s zeta-score example and change one declared input for case 11. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 12: Compare z-prime
Setup: Use the article’s zeta-score example and change one declared input for case 12. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 13: Compare En
Setup: Use the article’s zeta-score example and change one declared input for case 13. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 14: Check a small denominator
Setup: Use the article’s zeta-score example and change one declared input for case 14. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 15: Reject zero denominator
Setup: Use the article’s zeta-score example and change one declared input for case 15. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 16: Plot by concentration
Setup: Use the article’s zeta-score example and change one declared input for case 16. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 17: Inspect repeated positive signs
Setup: Use the article’s zeta-score example and change one declared input for case 17. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 18: Audit calibration contribution
Setup: Use the article’s zeta-score example and change one declared input for case 18. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 19: Audit recovery contribution
Setup: Use the article’s zeta-score example and change one declared input for case 19. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 20: Audit blank contribution
Setup: Use the article’s zeta-score example and change one declared input for case 20. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 21: Check rounding
Setup: Use the article’s zeta-score example and change one declared input for case 21. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 22: Preserve signed results
Setup: Use the article’s zeta-score example and change one declared input for case 22. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 23: Document coverage conventions
Setup: Use the article’s zeta-score example and change one declared input for case 23. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.
Practice case 24: Write an investigation conclusion
Setup: Use the article’s zeta-score example and change one declared input for case 24. Task: Recalculate from raw values, preserve units and state which scoring or design assumption changed.
Check: Check the relevant invariant, separate the number from the claim it supports, and name the investigation or source needed next. Rebuild the relevant counts before interpreting the coefficient, preserve the stated design, and write one sentence separating the numerical result from the claim it can support. Then vary one defensible assumption, calculate again, and explain whether the conclusion changes because of new data, a new reference model, or a new decision purpose. Finish by naming the unit, tolerance, and evidence needed for transfer. Record one graph or table that would expose failure hidden by the summary.