THE GOLD STANDARD OF INTELLIGENCE · SPECIALIST STUDY
Use recruitment order and degree depletion without pretending the hidden population size is identified automatically
Successive-sampling population-size estimation uses the ordered degree pattern in respondent-driven sampling to learn about a finite hidden population sampled without replacement. The gold standard states the recruitment model, degree measurement, seeds and prior, tests how high-degree people are depleted, and reports when the data cannot distinguish population size from network or recruitment assumptions.
This guide belongs to the Intelligence subject guide, the Gold Standard of Education, and the How X Works learning library. Choose a reading route, work the examples, then use the transfer checks to test whether the method survives a changed context.
Choose a reading route
Build the foundations · chapters 1–2
Apply the core method · chapters 3–5
Test evidence and alternatives · chapters 6–7
Handle limits and disagreement · chapters 8–9
Transfer and govern · chapters 10
The whole problem at a glance
| Signal | Model use | Main vulnerability |
|---|---|---|
| Recruitment order | Tracks depletion through the sample | Non-random recruitment |
| Reported degree | Approximates inclusion propensity | Recall and definition error |
| Sample fraction | Creates finite-population correction | Unknown population size |
| Prior on N | Stabilises weak identification | Can dominate inference |
The same ordered degrees can support different N values under different network and recruitment assumptions.
CHAPTER 1 OF 10 · Build the foundations
1. Start from finite sampling without replacement
Core idea. When a respondent-driven sample is a substantial fraction of the hidden population, recruited members cannot be sampled again and inclusion probabilities change over time.
Method. Define target population, sample size, recruitment order and without-replacement approximation; compare with the small-sample with-replacement limit.
Worked example. In a population of 500, recruiting 250 people depletes many high-degree nodes; a with-replacement approximation misses that finite-population effect.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Real recruitment is not uniform successive sampling and network components can remain unreachable.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 1 is complete only when the claimed result survives that review.
Misconception to avoid. RDS is always a tiny fraction of the population. In local groups it may not be.
Practical evidence check
- Calculate the sample fraction under several plausible N values.
- Return to chapter contents
CHAPTER 2 OF 10 · Build the foundations
2. Use ordered degrees as depletion evidence
Core idea. Degree-biased recruitment tends to bring highly connected people earlier; declining reported degree can therefore contain information about N.
Method. Preserve exact recruitment order, define personal degree consistently, plot degree by order and separate seeds from recruited participants.
Worked example. Mean degree falls from 28 early to 18 late, consistent with depletion, but the same pattern could arise if early seeds came from a more connected subgroup.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Flat degree order provides little size information and may reflect large N, weak degree-proportional recruitment or measurement noise.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 2 is complete only when the claimed result survives that review.
Misconception to avoid. Any downward trend identifies population size. It is evidence only under a recruitment model.
Practical evidence check
- Permute recruitment order and compare the diagnostic signal.
- Return to chapter contents
CHAPTER 3 OF 10 · Apply the core method
3. Write the successive-sampling likelihood explicitly
Core idea. The model approximates recruitment as probability proportional to degree among people not yet sampled, so the ordered degree sequence informs the unseen degree distribution and N.
Method. State inclusion mechanism, degree distribution model, treatment of unsampled units and computational approximation; retain simulation settings.
Worked example. Early high-degree observations are more likely under degree-proportional sampling, while late composition reflects who remains. Candidate N values imply different depletion rates.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Network dependence, coupons and nonresponse violate the idealised drawing process.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 3 is complete only when the claimed result survives that review.
Misconception to avoid. The estimator is design-unbiased because it has a likelihood. Validity remains conditional on the approximation.
Practical evidence check
- Simulate from the fitted model and compare degree-by-order summaries.
- Return to chapter contents
CHAPTER 4 OF 10 · Apply the core method
4. Elicit and disclose the prior on population size
Core idea. SS-PSE is often weakly identified, so prior information about N can materially shape the posterior.
Method. Use external counts, service bounds or transparent scale priors; publish prior quantiles and run sceptical alternatives.
Worked example. A known minimum of 800 service contacts rules out smaller N, while a diffuse upper prior allows a long posterior tail.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. External bounds can share coverage bias with the RDS sample or be politically influenced.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 4 is complete only when the claimed result survives that review.
Misconception to avoid. A Bayesian interval lets the data speak without assumptions. The prior is part of the evidence contract.
Practical evidence check
- Plot prior and posterior together under at least three priors.
- Return to chapter contents
CHAPTER 5 OF 10 · Apply the core method
5. Model the hidden degree distribution carefully
Core idea. Inference about unsampled people requires a model for population degrees, including low-degree members who are least likely to appear.
Method. Fit several plausible degree families, retain heaping and top-coding information, and test posterior predictive degree tails.
Worked example. Observed degrees cluster at 5, 10 and 20 because respondents round; a smooth model mistakes heaping for population structure.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Completely isolated eligible people cannot be reached by link tracing and may sit outside the model’s effective population.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 5 is complete only when the claimed result survives that review.
Misconception to avoid. The sample degree histogram represents the population after weighting. It is already selection-biased.
Practical evidence check
- Compare observed, weighted and posterior population degree distributions.
- Return to chapter contents
Illustrative depletion pattern
| Wave | Mean reported degree | Interpretation |
|---|---|---|
| Seeds | 35 | Chosen, not probability sampled |
| Early recruits | 28 | High-degree people appear sooner |
| Middle | 20 | Some degree depletion |
| Late | 18 | Weak additional decline |
The values are an original teaching example; a decline can also reflect seed composition or changing recruitment.
CHAPTER 6 OF 10 · Test evidence and alternatives
6. Treat seeds as a separate design layer
Core idea. Seeds are commonly selected purposively and can influence early composition and recruitment paths even when later estimators aim to reduce seed dependence.
Method. Condition on seeds or exclude them from the successive-sampling likelihood, track trees and repeat analyses across seed sets.
Worked example. All seeds come from one high-degree service network; the first waves show strong degree decline that weakens when those trees are analysed separately.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Many waves do not guarantee escape from a disconnected or highly homophilous component.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 6 is complete only when the claimed result survives that review.
Misconception to avoid. The estimator automatically washes out seeds. Seed bias can persist through network structure.
Practical evidence check
- Re-estimate after dropping each seed tree in turn.
- Return to chapter contents
CHAPTER 7 OF 10 · Test evidence and alternatives
7. Audit recruitment behaviour and degree error
Core idea. Participants do not recruit uniformly from all ties, coupon limits constrain paths, and reported degree can refer to different eligible networks.
Method. Record coupon outcomes, recruitment preference, nonresponse and degree-question stages; model or bound deviations from degree-proportional recruitment.
Worked example. Participants recruit close friends rather than random eligible contacts, overrepresenting one subgroup even after degree adjustment.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Behavioural data are incomplete and asking for detailed networks can increase risk.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 7 is complete only when the claimed result survives that review.
Misconception to avoid. More precise weights repair non-random recruitment. Model misspecification can make precision misleading.
Practical evidence check
- Run sensitivity to homophily, differential recruitment and degree misreporting.
- Return to chapter contents
CHAPTER 8 OF 10 · Handle limits and disagreement
8. Use diagnostics as model checks, not proof
Core idea. Posterior predictive checks, simulated recruitment paths and comparison with known sample features can reveal mismatch but cannot validate unseen population structure.
Method. Compare recruitment depth, degree order, tree balance, subgroup transitions and repeated-sample behaviour under fitted scenarios.
Worked example. The model reproduces mean degree but not long single-group recruitment chains, indicating missing homophily.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. A flexible model can fit observed summaries while remaining unidentified about N.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 8 is complete only when the claimed result survives that review.
Misconception to avoid. A good posterior predictive plot proves the population estimate. It tests selected observable implications only.
Practical evidence check
- List the summaries not used in fitting and test them separately.
- Return to chapter contents
Sensitivity report
| Input | Low scenario | High scenario |
|---|---|---|
| Population size prior | Tight external bound | Diffuse upper tail |
| Degree error | Reliable reports | Heaping and underreporting |
| Seed handling | Condition out seeds | Treat as modelled entrants |
| Recruitment | Near degree-proportional | Strong preference |
A credible interval conditional on one model is not full structural uncertainty.
CHAPTER 9 OF 10 · Handle limits and disagreement
9. Report conditional and structural uncertainty separately
Core idea. A posterior interval conditions on the prior, degree model and recruitment approximation; structural alternatives can be much wider.
Method. Publish posterior under the main model, sensitivity across priors and mechanisms, observed-count lower bound and decision consequences.
Worked example. The main posterior is 1,200–2,000, while plausible degree-error and seed scenarios extend to 3,500. Capacity planning uses staged bounds.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Very weak data may leave the upper tail largely prior-driven; saying unresolved is preferable to trimming it silently.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 9 is complete only when the claimed result survives that review.
Misconception to avoid. One Bayesian interval contains all uncertainty. It does not include unmodelled structure.
Practical evidence check
- Create an assumption-to-range table and identify the dominant lever.
- Return to chapter contents
CHAPTER 10 OF 10 · Transfer and govern
10. Govern SS-PSE through triangulation and community review
Core idea. Population-size estimates affect resources and stigma; they require ethical oversight, safe data, independent methods and participation by affected communities.
Method. Compare service multipliers, mapping, network scale-up and SS-PSE with dependency analysis; publish safe aggregates and revision triggers.
Worked example. A high SS-PSE estimate conflicts with independent mapping because one seed network spans two jurisdictions; the target boundary is corrected before resources are allocated.
Changed-context transfer. Changed-context transfer: apply the same reasoning structure to a school or public-service decision. Preserve the formal relationship, then replace the population, authority, privacy duties, evidence and remedy; never copy the domain conclusion.
Limits. Agreement can be circular when methods share lists or definitions, and public detail can expose networks.
Evidence and review. For this chapter, keep a dated evidence record: the input, rule or assumption, result, accountable owner and a condition that would change the conclusion. Work an ordinary case and a boundary case, preserve rejected alternatives, and require an authorised independent reader to reproduce where judgement entered. Chapter 10 is complete only when the claimed result survives that review.
Misconception to avoid. A sophisticated estimator is politically neutral. Boundaries, priors and publication choices carry values.
Practical evidence check
- Trace one estimate into a decision and later outcome review.
- Return to chapter contents
Build the specialist decision record
For “The Gold Standard of Intelligence | Successive-Sampling Population-Size Estimation, Degree Distributions, Seed Dependence and Finite-Population Correction”, begin with one bounded decision rather than an abstract claim. Name the affected people, data or institutions; the time window; the authority for acting; and the conclusion the method is allowed to support.
Preserve source material before interpretation. Record definitions, exclusions, transformations and the person responsible for each judgement. A reviewer should be able to separate an observed fact from an inferred result, a policy choice and a presentation shortcut.
Work one ordinary case in full. Show the input, every material rule or calculation, the output and the action that follows. Then write what the same evidence would mean under the simplest credible alternative method.
Keep an explicit unresolved state. Missing, protected, conflicting or unevaluated information must not become zero, false, consent or absence merely because a form expects one value.
Finish the record with a claim ledger. For every sentence that could drive action, identify its supporting source, inferential step, confidence or legal status, intended reader and expiry condition. This keeps descriptive evidence, formal consequence, professional judgement and institutional authority from collapsing into one impressive-looking conclusion.
Run the boundary-case audit
Select a case near the method’s domain edge: a missing input, a disputed definition, an endpoint, a sparse group, a resource limit or a legal deadline. Boundary cases reveal the contract that routine examples can conceal.
Change one assumption at a time and retain every output. Mark what stays stable, what reverses and what becomes undefined. If several assumptions move together, report that dependence instead of attributing the change to one convenient cause.
Search deliberately for a counterexample to the strongest claim. A method that works on the published illustration can still fail under a permitted input, an alternative grouping, a changed unit or an inaccessible evidence pathway.
Translate the counterexample into plain language: who would receive a different decision, what evidence would be hidden, and which safeguard or narrower claim repairs the problem.
Include an adversarial but plausible case. Let a missing group, ambiguous identifier, extreme weight, conflicting rule or urgent deadline interact with normal system incentives. The aim is not theatrical failure; it is to learn whether ordinary pressure makes the method conceal uncertainty, shift burden or reward a misleading shortcut.
Test independent reproduction
Give a reviewer the bounded inputs, public method contract and permitted access, but not the desired answer. Ask them to reproduce the result and identify every place where judgement, local knowledge or unavailable evidence was required.
Compare the reviewer’s intermediate states, not only the final number or label. Matching outputs can hide different denominators, inference paths, missing-data treatments or authorities that will diverge on the next case.
Where reproduction fails, classify the cause: ambiguous definition, undocumented transformation, incompatible implementation, inaccessible source, calculation error or legitimate alternative interpretation. Each class needs a different remedy.
Retain the disagreement and resolution. A corrected record should show what changed, why it changed and which prior outputs or decisions must be revisited.
Run the reproduction on a changed but structurally similar case after the first discrepancy is repaired. A method is not transferable merely because its authors can explain the original example. The new case tests whether the public contract carries the reasoning or whether hidden expertise is still doing essential work.
Design correction, challenge and remedy
Assign a visible route for correcting inputs, challenging assumptions and reviewing the application of the method. Technical accuracy without a usable remedy leaves the affected person carrying an error they cannot contest.
Separate urgent live-case repair from later system learning. A current decision may need pause, reassignment, fresh evidence or a human review before the organisation finishes its longer investigation.
Protect confidential and privileged material during review, but do not use confidentiality as a reason for unrecorded authority. Layer public explanation, protected oversight and individual reasons according to lawful access.
Verify that correction reaches cached results, derived records, dashboards and future decisions. Closing a complaint is not enough if the old conclusion continues to travel.
Measure remedy from the affected person’s position. A revised database row is useful, but it is incomplete when a hearing, service allocation, published claim or downstream model still relies on the old result. Record notification, restored opportunity, compensation or procedural repair where the governing framework provides it.
Set review and retirement triggers
Review after a change in definitions, population, law, source access, measurement scale, software, vocabulary or incentives. A stable formula can become invalid when the world or the meaning of its inputs changes.
Monitor leading signals as well as final outcomes: coverage gaps, timeouts, overrides, complaints, disagreement, missingness and boundary frequency. These often reveal failure before a headline metric moves.
Predefine thresholds for recalibration, suspension and retirement. Do not force a method to answer after its assumptions fail merely because replacing it is inconvenient.
At retirement, preserve the versioned contract, evidence, known limitations and migration decision. State which historic results remain interpretable and which require warning, restatement or withdrawal.
Schedule one prospective review before confidence fades or controversy forces it. Assign named owners for evidence, implementation and rights impact, and invite at least one reviewer outside the design team. The review should be able to narrow the claim, change the method or stop use without requiring proof of catastrophic failure.
Questions readers ask
What does successive sampling add to RDS?
- It models sampling without replacement and finite-population depletion.
Why does degree order matter?
- High-degree people tend to appear earlier under the approximation.
Can it estimate N without a prior?
- Information may be weak; prior and structural assumptions often matter greatly.
Do many waves remove seed bias?
- Not necessarily, especially with homophily or disconnected components.
What is the hard lower bound?
- At least the number of distinct eligible people observed, subject to linkage and boundary rules.
How should results be used?
- As transparent, revisable planning bounds with ethical safeguards.
Practise, verify and continue
Choose one consequential case. Reproduce the complete method, work one ordinary and one boundary case, test the strongest rival explanation, and state the evidence that would reverse or suspend the result. Build an evidence pack containing original inputs, definitions, calculations or rule steps, alternatives rejected, uncertainty, reviewer comments and the final action. Give the pack to a reader who did not design the method. Ask that reader to reproduce the result, identify one missing assumption and construct a case that would make the method return unresolved. Revise the public explanation only after the evidence pack and the operational process agree.
Move the method into an education or public-service setting. Keep the reasoning structure while replacing domain authority, population, privacy, evidence and correction routes. Start with a low-stakes historical example, then compare the specialist method with a transparent baseline. Check whether the same people, records and decisions remain inside scope; whether the scale or vocabulary still means the same thing; and whether a person affected by error can understand and challenge it. Pilot under human review, record reversals and burden, and stop the transfer if the formal resemblance hides a different ethical or legal problem.
Related eduKate reading
Respondent-driven sampling, network size and recruitment homophily
Collection requirements, source validation and analytic gaps
Competing hypotheses and diagnostic evidence
Negative reporting and absence-of-evidence safeguards
Analytic lineages, evidence provenance and judgement revision
Authoritative sources and scope
Handcock, Gile and Mar, Estimating Hidden Population Size using Respondent-Driven Sampling Data
Gile, Improved Inference for Respondent-Driven Sampling Data
Gile and Handcock, Respondent-Driven Sampling: An Assessment of Current Methodology
RDS software documentation, including Gile successive-sampling estimation
This article is statistical and public-health education, not operational guidance for identifying vulnerable people. Examples are original teaching cases. SS-PSE depends on target boundaries, recruitment, degree, seeds, priors and network reach. Protect participants, minimise data, report prior sensitivity and structural uncertainty, and never treat one posterior as hidden-population truth.