What are the two major methods of detecting deception
- Observing and talking to people
- Physiological Techniques (Polygraph and brain measures)
On average how many lies do Americans tell per day
1-2
Behavior cues for lying
- We assume that telling a lie creates a physiological change
- Assume nervous= lying, but that’s not true
- Verbal cues are best for detecting lies
- Speak prompted hand gestures vs rhythmic gestures
DePaulo, Lassiter, and Stone (1982) lying cues study
Message source participants told 4 descriptions to detection participants
- Describe someone you like
- Describe someone you dislike
- Describe someone you like as if you dislike them
- Describe someone you dislike as if you like them
Half of detection participants told to focus on vocal tone
- Produced better truth detection
Verbal cues to lying
- Higher pitched voice
- Speech disturbances
- Slower rate of speech
Content cues to lying
- Liars provide fewer details than truth-tellers
- Liars’ Stories are less compelling than truth-tellers (less plausible, lacking logical structure, more discrepancies, less engaging, less fluent, more nervous, tense)
- Truth-tellers are more likely to spontaneously correct their stories and more likely to admit lack of memory
Adams & Harpster (2008) 911 call experiment
- Analyzed 100 911 calls (50/50 for guilty and innocent)
INNOCENT - Request help
- Corrects misinterpretations
- Rude and demanding
- More emotion in voice
- Quickly speeking
- Cooperative
GUILTY - Irrelevant details
- State victim is dead
- Polite and patient
- Blame or insult victim
- Less emotion in voice
Microexpressions
Brief, uncontrollable facial expressions that reflect the emotion a person is feeling
Paul Ekman’s take on microexpressions
- Believed that certain people were especially good at lie detection
- Narrowed 12,000 professionals down to 29 people who could detect at 80-90%
- Theory was debunked
How were microexpressions studied
- Had nurses watch gruesome medical films and pretend they were pleasant
- Looked to see how they tried to conceal it
- Focused on containing their facial expressions
- Mircoexpressions still visible
Can we reliably detect deception?
- Laypeople detect at a rate of 54% accuracy
- Trained professionals detect at a rate of 55% accuracy
- Secret service are better than others
- Women better at detecting lies than men with people they are close with
Why are we so bad at detecting lies
- People rely on stereotypic beliefs about lying (gaze aversion and fidgeting)
- Truth-bias: People judge more messages as truthful than deceptive
- The differences between truth-tellers and liars are small
High-stakes lies
- Officers are better at catching high-stakes lies
- They are more familiar with settings that involve high-stakes lies
- Lies are easier to detect when the liar is motivated
The Othello Error
Too readily interpreting signs of nervousness as deception
The Polygraph Technique
- Not Admissible in Canadian court
- Polygraph: Device used to record an individual’s autonomic nervous system response (emotional states and arousal)
- physiological states associated with lying similar to those associated with anger, anxiety, embarrassment and fear
- Records beathing rate, heart rate and skin conductance (sweating)
The Comparison Question Test
10 yes/no questions
NEUTRAL QUESTIONS
- Asks about respondent’s identity, personal background etc.
RELEVANT QUESTIONS
- Asks about the crime being investigated
- Specific questions
PROBABLE-LIE COMPARISON QUESTIONS
- Emotionally arousing for any respondent (guilty or innocent)
Guilty if physiological arousal relevant > Comparison
Innocent if physiological arousal relevant< Comparison
Issue with set up of CQT
- If innocent Emotional aroused because they witness it
- If innocent physiological reaction to big claims due to fear of being convicted or gruesome details
- If guilty may react more to novel questions then crime they committed
- Guilty suspect may be habituated to details for crime by thinking about the crime
The Concealed Information Test
- Technically does not detect deception, determines if person knows details of the crime
- Multiple choice questions - one critical option and multiple foils
- Guilty: Larger physiological response to the correct option than the incorrect options
- Skin conductance is most common physiological response
ISSUES WITH CONCEALED INFORMATION TEST - may not remember details if guilt
- Community member may already know details
Lab studies for accuracy of polygraphs
Participants try to lie about a mock crime
- Stakes are low, lying is easier
Field Studies for accuracy of polygraphs
- Compare accuracy of original examiners to blind evaluators
- Blind evaluators only read charts, not swayed by other cues
- Difficulty establishing ground truth
Accuracy of CQT
- Confessions used to classify suspects as innocent or guilty (but could be false)
- Among guilty: 84-92% accurate
- Among innocent: 55-78% accurate
- 9 - 24% suspects falsely identified as guilty
- Experts don’t accept it as accurate but lay people do
Polygraph Countermeasures
Techniques used to conceal deception
- After 30 min of instruction 50% of guilty individuals can pass a polygraph
PHYSICAL
- Biting tongue, pressing toes on the floor
MENTAL
- Counting backward by 7 from 200 or any complex mental test will decrease physical arousal between questions
Brain Fingerprinting
Uses EEG to evaluate differences in event related potentials
Facial thermal imaging
Uses changes in facial temperatures to determine if an individual is lying
- warmer face if lying
- Not enough evidence for technique to be used in court
