Military Interrogation Training Data on Rapport and Persuasion Techniques

Seventy-six hours of recorded interrogations from a high-fidelity military exercise in Sweden were segmented, coded, and analyzed. The dataset, authored by Tove Phillipsen and last updated in May 2026, contains coded behavioral data from 25 prisoners and 6 trainee interrogators. It examines how rapport behaviors (HEAAR principles) and persuasion techniques predict verified information disclosure.

Use Cases

• Compare the effectiveness of rapport-based versus persuasion-based interrogation techniques based on coded behavioral data.
• Model information yield predictors using the ORBIT ordinal yield scale mentioned in the description.
• Analyze detainee responsiveness variability suggested by exploratory random-slope analyses.
• Train models to assess interpersonal interaction quality in high-stakes communication scenarios.

Strengths

• Data is based on 76 hours of recorded interrogations from a realistic military training exercise.
• Information yield was verified against a known exercise scenario ground truth.
• Analysis used Bayesian cumulative-probit regression with random intercepts for interrogator, detainee, and segment.

Limitations

• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.
• The dataset is small in size (215.4 KB), indicating limited scope.

Provenance

Source

figshare

Collection Method

Recordings from a high-fidelity military exercise were segmented, coded, and analyzed.

Freshness

Last updated 2026-05-25 10:30:39

Geography

Sweden