THE GOLD STANDARD OF INTELLIGENCE · SPECIALIST STUDY
Estimate a hidden population through social networks without hiding the visibility assumptions
Network scale-up estimation asks a general-population sample how many people they know in a hidden population and scales those reports by personal network size. The arithmetic is only the beginning. The gold standard defines “know”, estimates degree, tests barrier and transmission effects, separates sampling from structural uncertainty, and reports decision bounds rather than a deceptively precise count.
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
| Component | Question | Failure mode |
|---|---|---|
| Target size | Who is in the hidden population? | Boundary ambiguity |
| Personal degree | How many people does a respondent know? | Recall and definition error |
| Barrier effect | Are ties mixed like the whole population? | Homophily and geography |
| Transmission effect | Would membership be known? | Stigma and concealment |
A large survey cannot repair a target group that is systematically unknown to respondents.
CHAPTER 1 OF 10 · Build the foundations
1. Define the target population and reference population
Core idea. Network scale-up needs compatible boundaries for the population being sampled and the hidden population being estimated.
Method. Specify person, geography, age, behaviour or status, time window, inclusion overlaps and whether temporary residents or institutional populations belong.
Worked example. A survey samples residents aged 18–69 but the target includes younger people; scaling to the full census population would mismatch numerator and denominator.
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. Fuzzy or changing membership can dominate the error and may require several operational definitions.
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. The target label is self-explanatory. Each estimator needs an observable membership rule.
Practical evidence check
- Ask three independent coders to classify ten boundary cases.
- Return to chapter contents
CHAPTER 2 OF 10 · Build the foundations
2. Operationalise what it means to know someone
Core idea. Respondents must use a consistent social-tie definition that is relevant to information transmission.
Method. Use a time-bounded mutual-contact rule, cognitive prompts and non-overlapping relationship categories; test comprehension before fieldwork.
Worked example. “Someone you recognise” yields far larger degree than “mutual contact in the last two years and able to contact now.” The hidden-tie count changes with it.
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. No definition eliminates recall error, and online contacts can differ from offline disclosure 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 2 is complete only when the claimed result survives that review.
Misconception to avoid. A natural-language question gives comparable answers automatically. Tie definitions are measurement instruments.
Practical evidence check
- Re-interview a subsample using an alternative definition and compare degree.
- Return to chapter contents
CHAPTER 3 OF 10 · Apply the core method
3. Estimate personal network size transparently
Core idea. The basic estimator divides reported hidden ties by respondents’ total personal network sizes, so degree error directly affects population size.
Method. Use known-population or summation methods, publish reference groups, check impossible answers, model heaping and compare estimators.
Worked example. A respondent reports knowing 8 teachers and 12 nurses among calibrated groups; the degree estimate changes when one group is geographically concentrated.
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. Known population counts can be outdated, and respondents may not know occupations or statuses accurately.
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. Degree is a stable personal trait. It depends on question, population and time.
Practical evidence check
- Compare known-population and summation degree estimates for the same respondents.
- Return to chapter contents
CHAPTER 4 OF 10 · Apply the core method
4. Diagnose barrier effects and non-random mixing
Core idea. Barrier effects occur when social networks cross target-group boundaries more or less often than proportional mixing would predict.
Method. Stratify known-population performance and hidden ties by geography, age and relevant social structure; use multilevel or scenario models rather than one universal correction.
Worked example. Urban respondents know more members of the target group than rural respondents even after degree adjustment, so a city-heavy sample inflates the naive estimate.
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. Unobserved homophily remains possible, and post-stratification cannot repair a completely missing stratum.
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. Probability sampling removes barrier effects. It samples respondents fairly but does not randomise their friendships.
Practical evidence check
- Reweight the sample and vary cross-group mixing across plausible ranges.
- Return to chapter contents
CHAPTER 5 OF 10 · Apply the core method
5. Measure transmission error instead of assuming disclosure
Core idea. A respondent can know a person socially without knowing the sensitive membership that makes that person part of the target population.
Method. Use game-of-contacts or target-population studies where ethical, ask what fraction of contacts know the status, and model disclosure by relationship type.
Worked example. Only half of sampled target-group participants report that close contacts know their status; dividing naive reports by a single universal visibility rate may still miss relationship differences.
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. Visibility studies can have their own selection bias and must protect identities.
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. Transmission error is random forgetfulness. It is structured by stigma, privacy and relationship.
Practical evidence check
- Estimate visibility separately for close family, friends and co-workers, then test sensitivity.
- Return to chapter contents
Illustrative scale-up
| Quantity | Illustrative value | Interpretation |
|---|---|---|
| Population N | 1,000,000 | Reference population |
| Average degree d | 200 | Known people per respondent |
| Reported hidden ties y | 20 | Across 100 respondents |
| Naive estimate | 1,000 | N × y ÷ total degree |
Barrier and transmission corrections can move it sharply; the numbers are an original teaching example.
CHAPTER 6 OF 10 · Apply the core method
6. Calculate the naive estimator before applying corrections
Core idea. The uncorrected scale-up ratio provides a transparent baseline against which every adjustment can be understood.
Method. Sum hidden ties and estimated degrees with survey weights, retain zero reports, show the formula and then add corrections one at a time.
Worked example. Across weighted respondents, total reported hidden ties are 20 and total degree is 20,000 in a population of one million, giving a naive estimate of 1,000.
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. Ratio estimators can be unstable with extreme degree values, clustering and small hidden-tie counts.
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. Corrections should be applied inside one black-box model. A visible baseline exposes their leverage.
Practical evidence check
- Recompute after trimming or retaining high-degree respondents and explain the difference.
- Return to chapter contents
CHAPTER 7 OF 10 · Test evidence and alternatives
7. Separate sampling error from structural uncertainty
Core idea. A conventional confidence interval conditions on design and model choices; it rarely covers boundary, degree, barrier and transmission uncertainty.
Method. Report survey variance, model interval and scenario range separately. Use probability distributions only where evidence supports them and preserve unresolved states elsewhere.
Worked example. The design interval is 850–1,200, while plausible transmission and barrier scenarios produce 700–2,600. Planning should respond to both layers.
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. Wide structural bounds may be the honest result and do not identify the correct correction factor.
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. A narrow credible interval means the hidden population is known precisely. It can be conditional on fragile assumptions.
Practical evidence check
- List every assumption excluded from the formal interval.
- Return to chapter contents
CHAPTER 8 OF 10 · Test evidence and alternatives
8. Triangulate without recycling the same visibility pathway
Core idea. Multiplier, mapping, capture–recapture or respondent-driven methods challenge NSUM only when their data pathways and blind spots are sufficiently independent.
Method. Map denominator and source dependencies, compare time windows, and investigate disagreement rather than averaging incompatible estimates.
Worked example. A service multiplier and NSUM agree because both use the same clinic-defined target boundary; community mapping reveals people outside clinic reach.
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. Independent methods may estimate different operational populations and should not be forced to coincide.
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. Three estimates are automatically stronger than one. Correlated sources can repeat the same omission.
Practical evidence check
- Draw a dependency graph for every input and mark shared sources.
- Return to chapter contents
Decision-bound report
| Layer | Contains | Use |
|---|---|---|
| Observed reports | Raw counts and zeroes | Audit collection |
| Sampling interval | Sample design uncertainty | Conditional precision |
| Structural scenarios | Degree, barrier, transmission | Robust planning |
| Triangulation | Independent methods | Challenge blind spots |
Do not fold every assumption into one opaque corrected point estimate.
CHAPTER 9 OF 10 · Handle limits and disagreement
9. Protect respondents and hidden populations
Core idea. Questions about sensitive contacts can expose stigma, create fear or prompt unsafe inference about small groups.
Method. Minimise identifiers, use private administration, train interviewers, suppress unsafe cells, consult communities and predefine disclosure controls.
Worked example. A district-by-trait table is combined into a broader region because the apparent cell could identify one local network.
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. Protection can reduce detail and power; analysts must state the loss rather than reconstruct identities indirectly.
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. The survey asks about other people, so participant risk is low. Network questions can reveal both respondent and contacts.
Practical evidence check
- Run a re-identification and harm review before release.
- Return to chapter contents
CHAPTER 10 OF 10 · Transfer and govern
10. Govern estimates as revisable planning bounds
Core idea. Hidden-population estimates are time-, definition- and visibility-bound evidence for services, not permanent facts or enforcement targets.
Method. Publish low, central and high scenarios, tie each to capacity choices, monitor external indicators and schedule revision when networks or disclosure change.
Worked example. A programme funds a safe base at the observed lower layer and reserves scalable capacity for the high scenario, then updates after an independent survey.
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. Political incentives can favour high or low counts; independent review and community participation are necessary.
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. One official point estimate improves accountability. A range with assumptions can be more accountable.
Practical evidence check
- Trace how each scenario changes resources and who bears under- or over-estimation costs.
- Return to chapter contents
Build the specialist decision record
For “The Gold Standard of Intelligence | Network Scale-Up Estimation, Barrier Effects, Transmission Error and Hidden-Population Bounds”, 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 is network scale-up estimation?
- A survey-based method that infers hidden-population size from reported social-network ties and estimated personal degree.
What is a barrier effect?
- Departure from proportional social mixing across groups.
What is transmission error?
- A hidden member is known socially but their membership is not known to the respondent.
Does probability sampling remove network bias?
- No; it addresses respondent selection, not friendship formation or disclosure.
Why report bounds?
- Structural assumptions can exceed sampling error.
Should estimates guide enforcement?
- No; use them for protective planning with community and privacy 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
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
UNAIDS and WHO, Guidelines on Estimating the Size of Populations Most at Risk to HIV
WHO Eastern Mediterranean Region, Network scale-up and proxy respondent guidance
Bernard et al., The network scale-up method for public health
Teo et al., network scale-up estimation in Singapore
This article is public-health and analytic education, not operational guidance for identifying vulnerable people. Examples are original teaching cases. Protect respondents and communities, obtain ethical oversight, minimise data and publish safe aggregates. Estimates depend on boundaries, degree, mixing, disclosure and model assumptions. Use transparent decision bounds and independent triangulation; never treat one estimate as hidden-population truth.