Choose your reading route
eduKateSG · Why Mathematics? · Rotating Panel Surveys: Sample Overlap, Composite Estimation and Change Measurement
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 8–14
Reproduce the example
Audit the numbers one step at a time.Route 3 · Chapters 15–21
Check assumptions
See where design and calibration matter.Route 4 · Chapters 22–28
Explore applications
Connect the method to a concrete investigation.Route 5 · Chapters 29–34
Study and verify
Practise, inspect failure cases and read further.
Full chapter index · Practice and checking · Mathematics Learning Hub
Why is mathematics important when a survey must measure both the current level and how it changes over time without interviewing the same people forever? Rotating panel designs keep part of the sample, replace part and use overlap to improve comparisons. Composite estimators combine current and past information while balancing precision, respondent burden and panel effects.
Did You Know? The U.S. Current Population Survey follows a 4–8–4 pattern: four months in sample, eight months out, then four more months in sample.
Full chapter index
Chapters 1–7 · Understand the question
Chapters 8–14 · Reproduce the example
Chapters 15–21 · Check assumptions
Chapters 22–28 · Explore applications
Chapters 29–34 · Study and verify
Independent samples refresh representation and burden but make change estimates noisy. Complete overlap improves comparisons but creates burden, attrition and panel conditioning.
A rotation design replaces selected groups on a schedule. Overlap induces positive correlation between adjacent estimates, often reducing variance of differences.
The Census Bureau describes CPS households as interviewed for four months, out for eight, then interviewed for four more; 75% overlaps month to month and 50% year to year. Read the official methodology.
Suppose two monthly samples each contain 800 units and 600 are shared. The overlap rate relative to either sample is 600/800=75%.
| Sample at t-1 | Sample at t | Shared | Overlap rate |
|---|---|---|---|
| 800 | 800 | 600 | 75% |
For the 600 shared units, calculate individual changes and their average. Common stable differences cancel, which can lower change variance.
The 200 outgoing and 200 incoming units help refresh coverage and estimate the current level. Ignoring them may waste level information.
A composite estimator blends a direct current estimate with an updated previous composite plus a matched-sample change. Coefficients control memory and refresh.
The Census Bureau describes rotation as a compromise between complete overlap and independent samples, with composite estimation for levels and changes. Read the official research report.
If the direct current estimate is 52, the previous composite is 50 and matched change is 1, an updated value is 51. With blend alpha=0.6, composite=0.6×52+0.4×51=51.6.
| Direct current | Updated prior | Alpha | Composite |
|---|---|---|---|
| 52 | 51 | 0.6 | 51.6 |
More direct weight responds quickly but may be noisier. More carried-forward weight smooths noise but can lag sudden real change.
Optimal coefficients depend on variances and covariances across rotation groups and time. Treating estimates as independent discards the reason overlap helps.
Repeated respondents may answer differently because of survey experience. Rotation limits exposure but does not automatically remove time-in-sample effects.
Returning panels can have different response propensities than new panels. Weighting and nonresponse adjustments should recognise month-in-sample.
Groups entering at different times may face different field conditions, seasonal effects or modes. Diagnostics compare estimates by rotation group.
A design excellent for month-to-month change may not be best for current level. Objectives and coefficients should state which precision matters most.
Overlapping panels do not remove seasonal patterns. Seasonal adjustment is a separate time-series procedure applied to suitable estimates.
Replicate systems must preserve rotation overlap and composite recursion. A standard error for a direct cross-section is not the standard error for a composite change.
A cohort follows the same people for substantive longitudinal analysis. A rotating survey deliberately refreshes panels to sustain repeated cross-sectional inference.
Track overlap, entry and exit groups, time-in-sample response, attrition, group estimates, covariance, composite coefficients, revisions and respondent burden.
Reproduce 75% and 51.6, simulate independent and overlapping samples, compare change variance, add panel effects and test coefficient sensitivity.
Overlap cannot fix coverage error, nonresponse, panel conditioning, poor seasonal adjustment or sudden structural breaks. Smoothing can hide rapid change.
Use rows for months and columns for rotation groups, sample status, weights, estimates, matched change, overlap and composite coefficients.
Simulate panel entry and exit, maintain persistent unit IDs, estimate covariance, recursively update the composite and compute design-aware replicate variance.
Labour statistics, longitudinal education surveys, market research and public-health surveillance use rotating designs.
Compare no overlap, 75% overlap and complete overlap for estimating a stable level and a sudden change under equal interview budgets.
Overlap percentage is not response rate, repeated cross-sections are not automatically longitudinal cohorts, and a smoother series is not necessarily more accurate.
CHAPTER 27 OF 34 · Explore applications
27. Connection to recent mathematics articles
The nonresponse-weighting guide explains response differences. The successive-difference replication guide connects ordered repeated estimates and variance.
Students should map which units overlap before calculating change. Parents can ask whether repeated school surveys follow the same pupils or refresh the sample.
- Calculate 75% overlap.
- Calculate matched change.
- Reproduce 51.6.
- Change alpha.
- Diagnose panel conditioning.
### Why rotate samples? To balance continuity, refresh and burden.
### Does overlap reduce every variance? No.
### Is a composite a moving average? It can resemble one but uses survey estimates and overlap structure.
### Can smoothing lag change? Yes.
Report rotation pattern, overlap, entry and exit rules, response by time-in-sample, direct and matched estimates, covariance, coefficients, revisions, variance and panel effects.
The same unit contributes to adjacent periods, so stable unit differences cancel in a change estimator.
Confirm persistent IDs, intended overlap, refresh rules, response and panel effects, covariance-aware coefficients, separate level and change objectives, replicate variance and structural-break sensitivity.
Continue through the Mathematics Learning Hub. Recalculate one design quantity and explain which source of uncertainty it represents.
Extended practice laboratory
Practice case 1: Define periods
Setup: Start from the article’s rotating-panel mechanism for case 1. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 2: Define panel groups
Setup: Start from the article’s rotating-panel mechanism for case 2. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 3: Map month one
Setup: Start from the article’s rotating-panel mechanism for case 3. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 4: Map month two
Setup: Start from the article’s rotating-panel mechanism for case 4. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 5: Count overlap
Setup: Start from the article’s rotating-panel mechanism for case 5. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 6: Calculate overlap rate
Setup: Start from the article’s rotating-panel mechanism for case 6. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 7: Calculate matched change
Setup: Start from the article’s rotating-panel mechanism for case 7. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 8: Calculate direct level
Setup: Start from the article’s rotating-panel mechanism for case 8. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 9: Update prior composite
Setup: Start from the article’s rotating-panel mechanism for case 9. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 10: Choose alpha
Setup: Start from the article’s rotating-panel mechanism for case 10. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 11: Calculate composite
Setup: Start from the article’s rotating-panel mechanism for case 11. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 12: Change alpha
Setup: Start from the article’s rotating-panel mechanism for case 12. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 13: Use independent samples
Setup: Start from the article’s rotating-panel mechanism for case 13. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 14: Use full overlap
Setup: Start from the article’s rotating-panel mechanism for case 14. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 15: Add attrition
Setup: Start from the article’s rotating-panel mechanism for case 15. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 16: Add panel conditioning
Setup: Start from the article’s rotating-panel mechanism for case 16. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 17: Compare rotation groups
Setup: Start from the article’s rotating-panel mechanism for case 17. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 18: Estimate covariance
Setup: Start from the article’s rotating-panel mechanism for case 18. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 19: Calculate change variance
Setup: Start from the article’s rotating-panel mechanism for case 19. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 20: Add seasonal pattern
Setup: Start from the article’s rotating-panel mechanism for case 20. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 21: Audit IDs
Setup: Start from the article’s rotating-panel mechanism for case 21. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 22: Audit response
Setup: Start from the article’s rotating-panel mechanism for case 22. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 23: Audit variance
Setup: Start from the article’s rotating-panel mechanism for case 23. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 24: Name limits
Setup: Start from the article’s rotating-panel mechanism for case 24. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.
Practice case 25: Write conclusion
Setup: Start from the article’s rotating-panel mechanism for case 25. Task: Change the named quantity, recompute the dependent result and compare with the original.
Check: Preserve the target population and sampling design, verify the weighting identity, state one diagnostic, record units and compare one defensible alternative with clear reasoning. Show the arithmetic, name the response and predictor units, and separate the numerical result from the claim it supports. Then vary one defensible assumption, calculate again and explain whether the conclusion changed because of the data, the fitted model or the validation purpose. Finish with one compact table or graph that could reveal a failure hidden by the summary.