Forensic Lie Detection Flashcards
Who Lies?
Daily Telegraph (1994)
1 in 4 tell important lies daily
1 in 12 claim never to lie
Problems with social desirability
Who Lies?
Royal & Sun Alliance (2007)
Do we tell more lies a decade later?
45% were prepared to lie to their boss.
41% admitted they would lie to their friends.
20% said they had lied to an insurance company
38% claimed they had never told a lie
Who lies?
DePaulo et al 1996
Using diary
Students lie x2 day 1 in 3
Community members lie 1x day / 1 in 5
Who do we lie to?
Romantic partners 1 in 3
Spouses 1 in 10
Student - Mother 1 in 2
How do we tell our lies?
Face to Face 85%(s)/72%(c)
Telephone 14%(s)/28%(c)
Written 1%(s)/<0.05%(c)
When to use Lie Detection
- white lies
- more serious lies
Who uses Lie Detection?
- police (lie detection)
- other professions (no less important)
How do we measure detection accuracy?
Standard Methodology
How good are you?
Human Lie Detectors
Accuracy rates in general
Average 54% (Bond & DePaulo 2006)
Lies - 47% accuracy
Truth - 61% accuracy
Human Lie Detectors
Professionals
Ekman & O’Sullivan (1991)
Secret Service - 64.12 Federal Polygraphers - 55.79 Robbery Investigators - 55.79 Judges - 56.73 Psychiatrists 57.61 Special Interest 55.34 College students 52.82
Truth Bias vs Lie Bias
Truth Bias
Everything is the truth unless we have a reason to believe otherwise.
Cognitive short-cut for non-professionals
Truth Bias vs Lie Bias
Lie Bias
Ekman (2001)
Police and social workers “think that nearly everyone they see is guilty and everyone is lying”
No one ever tells the truth
Possibly not using correct cues
Lie Signs
Subjective
Lack of eye contact
Global Deception Research team 2006 - in 51 of the 58 countries we studied, gaze aversion is more prevalent than any other belief about lying.
71.5% of respondents worldwide believe that liars would avoid eye contact.
43% mention first cue of lying
Nervous body shifting 65.2%
Lie Signs
Objective Cues
- no consensus
- lack of eye contact?!
- body lang unreliable
the honest baseline…
Physiological Lie Detection
Tyes
1) Voice stress analysis (Psychological Stress Evaluator)
2) The Polygraph
3) Brain Imaging
Voice Stress Analysis
Definition
Lie detector tests that analyse voice changes in telephone conversations are being used to catch benefit cheats.
Voice Stress Analysis
How does it work?
Records voice using a microphone and based on tenent non-verbal, low frequency content of the voice conveys information and psychological state of speaker.
Aims to differentiate between stressed and non-stressed outputs in response to stimuli with high stress seen as indication of deception.
The Polygraph
How does it work?
Measures and records physiological indices: BP, pulse, respiration, skin conductivity while a person is asked a series of questions.
Deceptive answers will produce physiological responses.
The Polygraph
Main Question Schedules
Control Question Technique
Guilty Knowledge Test
Other tests e.g. truth control
Control Question Technique
Key advocate - Charles Honts
Compares responses to relevant questions. Control questions are designed to control for effect of generally threatening questions.
Used to arouse subjects concern about past truthfulness while relevant questions than to control questions lead to diagnosis of deception.
Control Question Technique
Types of Question
Neutral
Control
Relevant
Criticisms of CQT
Innocent suspects and the control questions
Subjective role of examiner
Problems with scoring
Ethical problems
False positives
Countermeasures
Guilty Knowledge Test
AKA peak of tension.
Protection of innocence.
Test of knowledge that only a guilty person could possess.
MCQs
e.g. could be asked to read out list of six weapons. Show physiological response when reading out the correct weapon.
Guilty Knowledge Test
Types of Question
MCQ
Answer No to all
Ignore first question responses.
Criticisms of GKT
Limited applicability
- only used where facts are not widely available
- limited question generation
Memory fallibility
False Negatives
Accuracy of the Polygraph
Does it work?
Types of outcome
Good result
False Positive
False Negative
Field / Experimental Study
Accuracy of Polygraph
Stats
Vrij 2000
Guilty Correctly Classified
Lab CQT - 66%
Lab GKT - 96%
Field CQT - 72%
Field GKT - 96%
Problems of Accuracy
Vrij 2000
False Positive
Lab CQT - 13%
Lab GKT - 4%
Field CQT - 21%
Field GKT - 4%
False Negative
Lab CQT - 9%
Lab GKT - 18%
Field CQT - 10%
Field GKT - 41%
Applying the Polygraph
Polygraph in court
- not UK
- Limited USA
- Japan routinely
Used by some agencies
- USA gov’t
- UK Secret Services?
- Other countries?
Brain Scanning and Lie Detection
- current play?
- fMRI
e. g. Langleben (2008)
Brain Wave Monitoring (P300 Event-Related Brain Potentials)
e.g. Farwell 1991
BPS Working Party 2004
Evaluation of Lie Detection Techniques?
Issues in accuracy and reliability