Who Can Best Catch a Liar A Meta-Analysis

Who Can Best Catch a Liar? A Meta-Analysis of Individual Differences in Detecting Deception Michael G. Aamodt & Heather Mitchell Radford University Radford, Virginia

Detecting Deception in Police Psychology Criminal investigations – Suspects – Accusers – Witnesses Psych evaluations – – – Pre-employment Fitness for duty Insanity pleas Competence to stand trial Threat assessment Employment – Interviews – Reference checking – Internal affairs Courtroom testimony Hostage negotiation Conflict management Political survival

Old Ways of Detecting Deception India- Trial by Sacred Ass Arabs - Hot iron to tongue Chinese - Swallow rice flour Hindus - Chew rice and spit Inquisition - Chew and swallow a slice of bread and cheese Judicial torture in Europe King Solomon Source: Jack Annon

New Ways of Detecting Deception Electronic Methods – Polygraph – Voice stress analyzer Neurological Methods – Brain fingerprinting – Brain mapping Extreme Methods – Chemicals – Torture Communication – Actual words used – Paralanguage – Body language

General Research Findings People usually detect deception at slightly above chance levels Subjects have a “truth bias” when responding Training can help, but… Having a baseline is essential Listeners are better than interrogators Use of patterns rather than single cues is essential

Behavioral Indicators Will Only Be Successful If You have a baseline of behavior There is a consequence for getting caught The response is spontaneous The person does not believe the lie (e. g. , Clinton, O. J. ) The lie involves a high degree of cognitive complexity Source: Jack Annon

Scientific Inquiry Deceiving – Cues used – Individual differences Detecting Deception – Overall accuracy – Conditions affecting accuracy – Effect of training – Cues used – Individual differences in accuracy

Study of Deception is International in Nature Sweden – Pär Anders Granhag (Göteborg University) – Leif Strömwall (Göteborg University) – Maria Hartwig (Göteborg University) United Kingdom – Aldert Vrij (University of Portsmouth) – Siegfried Sporer (University of Aberdeen) United States – – – Bella De. Paulo (University of Virginia) Paul Ekman (University of California, San Francisco) Charles Bond (Texas Christian University) Canada – Stephen Porter (Dalhousie University)

Countries with Studies in Our Meta-Analysis Australia Canada China England Germany Israel Jordan Netherlands Scotland Sweden United States

Our Study Conduct a meta-analysis on individual differences in the ability to detect deception Meta-analysis is a statistical review of the literature Individual difference variables – Experience – Confidence – Sex – Personality

The Literature Review Goal: Find all relevant studies from 1970 -2003 – Others dates included when found – Study had to report correlations or a statistical test or raw data that could be converted into a correlation Method – Computer searches – Bibliography leads – Hand searches of key journals

Literature Review Results Volume – 76 studies – 9, 453 subjects Study Date – – – 1960 s (2) 1970 s (8) 1980 s (22) 1990 s (26) 2000 s (18) Source – – Journal articles (67) Dissertations (7) Master’s theses (1) Book chapters (1)

Each Meta-Analysis Contains Number of studies (k) Number of officers in the analysis (n) Mean validity coefficient (r) 95% confidence interval % of observed variance explained by sampling error – If < 75% a search for moderators was conducted

Are Professionals Better than Students? Group K N Accuracy % Criminals 1 52 65. 40 Secret service 1 34 64. 12 4 2 8 4 122 5 1 508 194 511 341 8, 876 341 32 61. 56 59. 01 55. 16 54. 54 54. 20 51. 16 40. 42 Psychologists Judges Cops Federal officers Students Detectives Parole officers

Problems in Comparing Studies The Stimuli are Different Task – Realism – Consequence of getting caught Stimulus – Length – Number of attempts – View (full body, head, voice only)

Is Confidence Related to Accuracy? 95% CI Confidence K N r L U SE% Total 31 3, 033 . 06 . 03 . 10 86% Police 11 1, 174 . 02 -. 03 . 08 100% Students 18 1, 747 . 10 . 05 . 15 100% Is confidence related to accuracy? Yes Size of the relationship? Small Can we generalize the findings? Yes Cops (k=11) Students (k=17) Other (K=2)

Is Experience Related to Accuracy? 95% CI Variable K N r L U SE% Experience 8 696 -. 07 -. 14 . 00 100% Is experience related to accuracy? Size of the relationship? Can we generalize the findings? No Yes Cops (k=7) Students (k=1) Other (K=0)

Is Age Related to Accuracy? 95% CI Variable K N r L U SE% Age 8 862 -. 08 . 05 100% Is age related to accuracy? Size of the relationship? Can we generalize the findings? No Yes Cops (k=5) Students (k=2) Other (K=1)

Is Sex Related to Accuracy? 95% CI Sex K N d L U SE% Overall 15 1, 451 . 06 -. 06 . 19 17% Law Enforcement 3 144 . 63 . 46 . 81 33% Students 12 1, 307 . 00 -. 09 38% Note: A positive “d” indicates men were more accurate than women Is sex related to accuracy? Size of the relationship? Can we generalize the findings? No No Cops (k=3) Students (k=10) Other (K=2)

Is Personality Related to Accuracy? Too few studies to determine – Openness (k=1) – Conscientiousness (k=1) – Extraversion (k=4 related, 2 on extraversion) – Agreeableness (k=2) – Neuroticism (k=2) – Other (k=7)

Analyses Still to be Conducted Track down a few missing studies Investigate moderators for Accuracy Rates – – Medium (audio, visual, written) Visual cue (face, body, legs) Presence of a baseline Number of segments viewed Enhance database for sex differences – Contact recent authors for more info Explore truth vs. lie accuracy – – Actual difference Role of truth bias and/or context

Questions? Michael G. Aamodt, Ph. D. Department of Psychology Radford University Radford, VA 24142 -6946 (540) 831 -5513 maamodt@radford. edu www. radford. edu/~maamodt

Citation Information for this Presentation Aamodt, M. G. , & Mitchell, H. . (2004, October). Who can best catch a liar? A meta-analysis of individual differences in detecting deception. Paper presented at the annual Meeting of the Society for Police and Criminal Psychology, Rome, Italy.