REAL-WORLD INTERPRETIVE

PSYCHOLOGICAL WARFARE AND INFORMATION INTEGRITY

Gulf War POW Survey Origin, Channel Percentages, and Sampling Limits

A page-specific review of RAND's 250-POW survey discussion, the 14 March 1991 report path, front-line weighting, missing instrument, and causal limits.

Read the working brief Choose a learning path

REAL-WORLD INTERPRETIVE Editorial synthesis or explanation derived from sources and explicitly open to correction.

Confidence: Moderate
Sources: 3
Independent corroboration: 0
Jurisdiction: International / historical source context
Date range: 1943–1996; review through 2026-07-26
Review: Editorial primary-source and origin-chain review complete — independent archival, rights, translation, affected-community, legal, accessibility, and specialist review pending

Estimated reading: 14 minutes
Review date: 2026-07-26
Stable route: /psyops/cases/gulf-war-pow-survey-origin-and-sampling-limits

Source boundary · evidence boundary

SOURCE-BOUNDED EDUCATION — NOT OPERATIONAL GUIDANCE. Artifact survival, cataloging, distribution, availability, exposure, attention, interpretation, belief, behavior, operational outcome, and strategic effect remain separate. No page grants targeting, audience profiling, message optimization, surveillance, cyber, publication, production, activation, or tuning authority.

LEVEL 1

ORIENTATION

Why this matters

REAL-WORLD INTERPRETIVE

One-sentence brief

The often-repeated 98/88/70 shorthand hides channel-specific figures and a direct warning that the selected sample did not represent all Iraqi troops in the theater.

REAL-WORLD INTERPRETIVE

Three key points

An archive catalog record or surviving artifact establishes only the propositions its metadata and custody support.
Repeated statistics remain one origin chain until an independent underlying record or method is retrieved.
Output, distribution, selected-sample self-report, behavior, operational outcome, and strategic effect are separate.

LAB

LOCAL-ONLY EFFECT-CLAIM AUDIT

Check whether an influence claim exceeds the evidence

Causal-evidence prompt — not an effectiveness score

This deterministic workbench uses fixed declarations only. It does not accept a real audience, person, account, location, message, campaign, dataset, URL, or target. It performs no persuasion analysis, vulnerability scoring, forecasting, legal judgment, publication action, or operational recommendation.

Claimed rung and causal evidence Highest outcome currently claimed Exposure evidence Counterfactual or comparison design Alternative explanations and confounders Who produced the result claim? Duration of the observed outcome

Publication safeguards The exact outcome, unit of analysis, and time window are named. Baseline conditions and relevant base rates are recorded. Negative, null, contradictory, and missing findings are preserved. Blowback, unintended audiences, displacement, and opportunity costs are considered. Origin-chain independence and evaluator conflicts are disclosed. Review date, correction trigger, and reopening condition are defined. The output denies targeting, publication, specialist, operational, production, and tuning authority.

REVIEW OUTPUT

Only activity or output is established

Distribution, impressions, repetition, engagement, disruption, or an institution’s own after-action report cannot by themselves establish persuasion, behavior change, operational success, or strategic effect.

Highest defensible rung: Activity or output
Review state: Incomplete
Authority: Review prompt only

Required next actions

Define the exact outcome, population or unit, baseline, time window, and counterfactual.
Record competing explanations and separate sponsor assessment from independent evaluation.

Rule trace

PE-01 Output does not establish exposure.
PE-02 Exposure does not establish persuasion or behavior.

Boundary: even a methodologically strong historical study does not create a universal effect law, a persuadability score, a target ranking, a campaign recommendation, specialist approval, publication authority, or production authority.

JavaScript is unavailable.

Use the static method on this page: identify the exact claimed outcome, record distribution and exposure separately, name the unit and time window, establish a baseline and counterfactual, test alternative explanations, preserve null findings, disclose evaluator conflicts, and stop public wording at the highest rung directly supported.

LEVEL 2

WORKING BRIEF

Evidence, context, and limits

REAL-WORLD VERIFIED

Origin path

RAND identifies an Operation Desert Storm POW report dated 14 March 1991 as the survey summary and says it was cited in the VII Corps G2 study, The 100-Hour Ground War: How the Iraqi Plan Failed, pages 60–61.

The source path is more precise than WIP.40.
The report itself remains unavailable.
Official Army sources confirm the study's existence.

REAL-WORLD INTERPRETIVE

Channel-specific figures

RAND reports exposure of 98 percent to leaflets, 58 percent to radio, and 34 percent to loudspeakers; belief of 88, 46, and 18 percent; and reported influence on surrender or desertion of 70, 34, and 16 percent.

The shorthand omits six percentages.
Reported influence is retrospective self-report.
Different constructs require different wording.

REAL-WORLD VERIFIED

Front-line weighting and nonrepresentativeness

RAND cautions that the sample was weighted heavily toward front-line infantry, where leaflet density was greater, and was not representative of all Iraqi troops in the KTO.

Selection is outcome-linked.
Exposure opportunity differed by unit.
Population generalization is unsupported.

REAL-WORLD INTERPRETIVE

Missing method and ethical context

The questionnaire, raw data, codebook, translation, interview conditions, nonresponse, and comparison group remain unavailable. Interviews with prisoners and line crossers require attention to detention, survival, coercion, expectations, and prior intent.

Method gaps remain open.
Self-report is not a counterfactual.
Affected-person dignity matters.

LEVEL 3

COMPLETE DOSSIER

Limitations, game links, and review context

DISPUTED / MULTIPLE ACCOUNTS

Known limitations and gaps

The site did not interview recipients, prisoners, veterans, archivists, translators, or affected communities for this increment.
RAND material is linked and cited but not republished because its product page prohibits unauthorized online posting.
The original 14 March 1991 questionnaire, raw responses, codebook, and full VII Corps redacted study were not retrieved.
Automated validation and editorial source review are not specialist approval or accessibility certification.

REAL-WORLD VERIFIED

Source, archive, and method records

No external source is required to understand the evidence method.

REAL-WORLD INTERPRETIVE

Glossary cross-reference

Terms are defined for this site’s evidence method, not as universal legal or clinical definitions.

selected prisoner sample sample weighting population representativeness origin record causal attribution gap

REAL-WORLD INTERPRETIVE

Decision matrix

Origin-chain and evidence-boundary matrix
Evidence available	Safe treatment	Do not infer
Catalog or finding aid	Record institution, identifier, date, description, custody, and access limitations.	Distribution, audience exposure, interpretation, or effect.
Institutional output table	Attribute the table, unit, categories, rounding, source, and missing ledger.	Unique readers, population reach, or persuasion.
Selected prisoner sample	State selection, sample size, front-line weighting, interview context, and missing instrument.	Representative population estimate or causal effect.
Operational outcome	Reconcile counts and expose battlefield, material, command, coercive, and prior-intent explanations.	A fixed causal share for one channel.

REAL-WORLD INTERPRETIVE

Publication audit checklist

Item identity

Is the institution, identifier, item state, date, language, and rights/access state explicit?

Pass condition: The record does not manufacture missing item metadata.

Origin path

Is the earliest accessible origin and every derivative step visible?

Pass condition: Repeated reports are not counted as independent corroboration.

Sample boundary

Are selection, weighting, denominator, translation, interviewer, and comparison gaps stated?

Pass condition: The sample is not generalized beyond its evidence.

Effect boundary

Are output, distribution, exposure, belief, self-report, behavior, outcome, and effect separate?

Pass condition: No stage silently establishes the next.

Reopening

Are correction, supersession, access, and new-evidence triggers recorded?

Pass condition: A future reviewer can update the claim without erasing prior state.

LEVEL 4

RESEARCH EDITION

Sources, methods, and stable links

REAL-WORLD INTERPRETIVE

Linked reports

Measuring PSYOPS EffectivenessProtected research reference; no public report route in this release.
Global Psychological Warfare HistoryProtected research reference; no public report route in this release.

REAL-WORLD VERIFIED

Method and corrections

This page follows the public method for provenance, confidence, source independence, alternative accounts, limitations, review state, and visible correction.

Read the method View corrections

CONTINUE

Related learning

Psychological warfare and information integrityA 1945 Indochina Leaflet Proposal: Preparation Is Not Proof of UseA NARA-based artifact record for a July 1945 leaflet prepared for possible use in French Indochina, with explicit preservation of uncertainty over whether it was ever used.Psychological warfare and information integrityAmerica's Playboy No.1 — Finding-Aid Evidence and Missing ContextA bounded Australian War Memorial finding-aid record for a wartime enemy leaflet title, copy relationships, image description, rights limits, and unresolved production and reception evidence.Psychological warfare and information integrityGulf War Leaflet Counting Table, Delivery Categories, and Origin GapA close reading of RAND Table 10.1, its 29.1-million rounded total, delivery categories, 4th POG attribution, and the missing operational ledger.Psychological warfare and information integrityGulf War Leaflets: Output, Distribution, Reach, and Effect Are Different ClaimsA source-chain review of RAND's rounded 29.1-million leaflet table, its 4th POG briefing attribution, the missing distribution ledger, and the boundary between output and effect. Learning pathPSYOPS Origin Chains: From Artifact Record to Survey and Effect ClaimA seven-step path through archive identity, explicit non-use uncertainty, finding-aid limits, output-table provenance, selected-sample methods, count reconciliation, and correction.