Polygraph Background Theory Types Accuracy. Physiological detection of deception (PDD) Use physiological measurements as an index of deception Not behavioral.

PolygraphBackgroundTheoryTypesAccuracy

Physiological detection of deception (PDD) Use physiological measurements as an index

of deception Not behavioral Directly measure arousal or other cognitive

processes

What is a polygraph?

NOT a lie detector Poly = many, graph = write Machine that records multiple continuous

measures of autonomic nervous system arousal Galvanic skin response (GSR) Thoracic and abdominal respiration Blood Pressure Heart rate

The “lie detector” refers more to the test used Relevant/Irrelevant test Rising Peak of Tension Comparison Question Test Directed Lie Test Concealed Information Test

Polygraph - History

William Moulton Marston (1893 – 1947) Student of Hugo Münsterberg

at Harvard Discovered correlation

between blood pressure and arousal during lying

Polygraph - History

John Augustus Larson Rookie police officer in the

Berkeley, CA, police department

Ph.D. in physiology from UC Read Marston’s article

“Physiological Possibilities of the Deception Test”

Improved test through continuous recording of blood pressure

Polygraph – History (Larson, cont…) First real-world application

“Cardio-pneumo-psychograph” Berkeley sorority house - 1921

Items including an expensive ring had been stolen from rooms Helen Graham

“No sooner had he brought up the subject of the diamond ring and stolen money – “The test shows you stole it. Did you spend it?” – than Graham’s record showed a precipitous drop in blood pressure before beginning what looked to be an alarming rise, along with skipped heartbeats and an apparent halt in her breathing.” – Alder, The Lie Detectors.

Married Margaret Taylor, one of the other suspects

Polygraph - History

Leonarde “Nard” Keeler Through connections with

Berkeley police chief, August Vollmer, was introduced to Larson (1930s)

Worked on developing his own polygraph while “studying” at Berkeley and UCLA

Created first polygraph school in Chicago in 1948

Autonomic nervous system (ANS) Part of the peripheral nervous system controlling

visceral or automatic functions Sympathetic and parasympathetic nervous systems

General theory behind polygraph Arousal Increased ANS activity

Sweating Respiration changes Vasoconstriction Pulse rate Blood pressure

Specific patterns of arousal during questioning could indicate guilt or lying

Polygraph – Modern version

Modern polygraphs are now computerized Allow for more accurate and automatic (unbiased) analysis

Main Measures Galvanic skin response (sweating) Respiration

Thoracic and Abdominal Blood pressure Pulse oximeter

Measures percentage of oxygenated hemoglobin Pad(s) to measure subject movement

Polygraph – Relevant/Irrelevant Test Earliest method of polygraph testing Two kinds of questions

Relevant Deal with issue at hand

Irrelevant Deal with outside facts or details

Assumption: A liar or guilty person will be more aroused by relevant

questions than Irrelevant ones, while an innocent person will show no difference So, if arousal(relevant) > arousal(irrelevant) = lying

Polygraph – Searching Peak of Tension (POT) Developed by Keeler Can be used when specific details of a crime are unknown to the

investigator Suspect is presented serially with potential relevant clues

Areas in which a body may be located Amounts of money that may have been stolen

Assumption: A guilty person will react strongest when the correct alternative is

chosen An innocent person may simply become more aroused as the

test goes on, but will not show a significant sudden increase in arousal to one alternative

Polygraph – Comparison Question Test Most common method of polygraph interrogation

Developed by John Reid Begins with extensive pre-test interview Three kinds of questions:

1. Relevant E.g. “Did you kill Nicole Brown Simpson”

2. Comparison (aka probable lie) E.g. “Have you ever physically harmed someone”

3. Irrelevant Is your name Orenthal James Simpson?

Polygraph – CQT (cont…)

Assumption: A liar become more aroused by lying to the relevant

questions than the comparison questions An innocent person will be more aroused by the

comparison questions Arousal(relevant) > Arousal(comparison) = guilty

Uses: Criminal investigations Employee screening Security clearances

Polygraph – Directed Lie Test (DLT) Same kind of questions as CQT, only subject

is instructed to lie to all the comparison questions

Assumption: Guilty person will show more arousal lying to

relevant questions Innocent person will show more arousal lying to

comparison questions

Polygraph – Concealed Information Test (CIT) AKA – Guilty Knowledge Test (GKT)

Developed by David Lykken in 1958 Rather than trying to detect arousal caused by lying, tries to detect arousal from

recognition of “guilty knowledge” from the “orienting response” Multiple-choice (serially presented) questions where the investigator knows the

correct answer “What was the weapon used to kill Mr. Boddy?”

Candlestick Rope Revolver Lead Pipe Knife Wrench

Assumption: A guilty person’s arousal will increase upon recognizing the correct alternative due to

involuntary orienting response Innocent person will not be able to discern the correct alternative from the others

Polygraph – CIT (cont…)

Lykken advocates 4 – 6 questions with 4 – 6 multiple-choice answers in each

Reduces theoretical false positive rate with addition of each question 1/5 > 1/25 > 1/125, etc…

Scoring 2 points if Probe is largest, 1 if second largest Total up points at the end

For 6 questions, 12 is perfect score Lykken used cutoff of 7

Polygraph - Accuracy

R/I Extremely poor

CQT 83 - 89% for guilty subjects 53 – 75% for innocent subjects

12 – 47% incorrectly classified (falsely accused of guilt) DLT

One study, 80% correct GKT

76 – 88% of guilty subjects 12 – 24% false-negatives

94 – 99% for innocent subjects 1 – 6% false-positives

Polygraph - Problems

CQT Based on faulty theory High false-positive rate Biased

GKT Difficult to create enough good GKT questions Not applicable in every setting

Psychopathy/sociopathy Estimates as high as 20% of criminal population

Polygraph – Problems (cont…) Countermeasures

Methods used to defeat a test Increase autonomic arousal

during certain questions Easy

Distraction techniques Difficult to identify Can be apply to any kind of

polygraph method

After 30 minutes of training, ~80% of subjects in a study by Honts et al., 1994, beat a CQT

Polygraph – Problems (cont…) Admissibility in court Daubert Standard

1. Is the scientific hypothesis testable? 2. Has the proposition been tested? 3. Is there a known error rate? 4. Has the hypothesis and/or technique been

subjected to peer review and publication? 5. Is the theory upon which the hypothesis and/or

technique based generally accepted in the appropriate scientific community?

Polygraph – So why is it still used? Effective at soliciting confessions

General belief of the infallibility of the machine “Psychological third-degree”

Employee Screening Can no longer be required due to Employee

Polygraph Protection Act of 1988

Polygraph – Famous misses

Julius and Ethel Rosenberg Passed nuclear secrets to

Soviet Union

Aldrich Ames CIA officer Convicted of spying for

Soviet Union

An actual CIT polygraph record

Additional Resources

A Tremor in the Blood – David Lykken

Handbook of Polygraph Testing – Murray Kleiner

The Lie Detectors: The History of an American Obsession – Ken Alder

Antipolygraph.org