March 25, 2004

1
March 25, 2004 PSYC 430 - 001Forensic
Psychology IntroductionCredibility
Assessment Dorothee Griesel, Dipl.-Psych.Departme
nt of Psychology University of British
Columbia The slides will be on the Yuille lab
website http//www.psych.ubc.ca/7Ejyuille/cours
es.htm
2
Outline

• Intro
• Non-verbal
• Physiological Approach The Polygraph
• CQT
• GKT
• P300
• Research
• Admissibility in Court
• Behavioral Indicators of Lying Behavior
• FACS
• Other behaviors
• Verbal Credibility Assessment Tools
• SVA / CBCA
• Moderators of the strength of deception cues

3
Intro

• Reasons for credibility assessments
• CJS Allegations, e.g., child (sexual) abuse
  suspect denies guilt
• Mental health Suicide assessment
• I/O Test of employees loyalty
• National security, e.g., immigration,
  counterintelligence
• Types of statements
• Real memory
• Deceptive attempt
• Historically wrong but subjectively true
  narrative (without misleading intention)

4
Intro cont

• Definition of a lie
• Target is deliberately misled.
• Target is not notified about the liars
  intention.
• Forms of lies
• Concealment
• Falsifications
• Telling the truth falsely.
• Telling a half truth.
• Incorrect-inference dodge
• No behavior is diagnostic of lying. Deception
  detection techniques are based on the assumption
  that the cognitive and emotional correlates of
  deception cause certain behavioral changes.
  However, not even behavioral changes are
  diagnostic of lying!
• Ways lies betray themselves
• Leakage
• Deception clues

(see Ekman, 2001)
5
Psychophysiological Approach The Polygraph

• Measures skin conductance, heart rate, blood
  pressure, respiration
• Relevant Irrelevant Test (RIT)
• irrelevant questions
• relevant questions
• Also uninvolved, truthful individuals recognize
  relevant questions as relevant and respond
  physiologically.
• Control Questions Test (CQT)
• relevant questions
• control questions
• 3 repetitions
• problematic assumptions
• most widely used
• purpose narrow the number of suspects down,
  lower costs of investigations

6
Polygraph cont
7
Polygraph cont

• CQT accuracy research - 2 errors
• false positive (FP) rate of those who are truly
  innocent but diagnosed as guilty
• false negative (FN) rate of those who are truly
  guilty but diagnosed as innocent
• ? Sensitivity proportion of correctly
  identified examinees having knowledge about crime
  details among all guilty examinees
• ? Specificity proportion of non-informed
  participants among all non-informed participants

8
Polygraph cont

• CQT accuracy research studies
• Raskin (1988) Laboratory study
• 97 accuracy for guilty subjects, 93 for
  innocent
• Raskin (1992) Comparison of lab and field
  studies
• 7 FN in lab, 10 FN in field studies
• Patrick Iacono (1991) Cases of 5 years from
  federal police examiners in Vancouver
• hit rate 55 for innocent, 98 for guilty
  subjects
• Fiedler et al (2002)
• The current CQT practice is simply not the kind
  of procedure that should be sold in the name of
  scientific psychology (p. 323).

9
Polygraph cont

• Directed Lie Test (DLT)
• Subjects instructed to answer no to the control
  questions.
• Control questions are the same for all
  participants.
• Guilty Knowledge Test (GKT)
• also Concealed Knowledge Test
• Key assumption There is some information about
  the episode that is known only to the
  investigator and to people who did participate in
  the episode.
• Multiple choice format
• Instruction to answer no to every alternative
• Conditions for proper use
• Applications

10
Polygraph cont

• GKT research
• Only 2 field studies (see MacLaren, 2001)
• Sensitivity 50
• Specificity 98
• Laboratory findings (see Elaad, 1998)
• Range of correct identifications of innocent
  participants 79-100
• Range of correct identifications of guilty
  participants 61-100

11
Polygraph cont

• Countermeasure problem
• Honts et al. (1994) Physical and mental
  countermeasures
• Drugs

12
Polygraph cont

• Event-related brain potentials (ERP)
• GKT rationale
• without relying on ANS responses
• The correct alternative will activate a cognitive
  processing revealed by the appearance of the
  P300.
• Oddball paradigm

13
Polygraph cont

• ? Control for false positives by applying GKT
  procedure For innocent subejct, crime-relevant
  stimuli are indistinguishable from irrelevant
  stimuli.
• ? Control for false negatives by applying
  control questions, that is the target stimuli.
• ? Countermeasures safe!
• Farwell Donchin (1991)
• Laboratory experiment ? No FPs, no FNs, 12.5
  insufficient information.
• Field study ? 100 in cases where determination
  was made, again 12.5 indeterminate.

14
Polygraph cont
15
Polygraph cont

• Admissibility of the polygraph in court
• Daubert criteria
• Testablity
• Known error rates
• Peer review and publications
• General acceptance
• Used in criminal investigations, but not
  permitted as evidence in criminal federal courts.
• Admissible in civil court if both sides agree.
• US Polygraph banned as employment selection tool
  expt. CIA, military, police forces

16
Behavioral Approach

• Lies fail for 2 reasons
• Thoughts ? cues mostly in the liars speech
  (latency, pauses)
• Emotions ? cues in the liars face (leakage,
  deception clues)
• Lying about feelings (e.g., conceal jealousness)
• Feelings about lying (e.g., fear of being caught,
  stress, guilt, duping delight)
• Again, no behavior(al change) is diagnostic of
  lying! ? Alternative explanations should be
  considered.
• Frequent errors (Ekman, 2001)
• Brokaw hazard
• Othello error
• Ekman demonstrated that observers from illiterate
  as well as literate cultures detect the same
  emotions in the same pictures.

17
Behavioral Approach cont

• Facial Action Coding System (FACS Ekman
  Friesen, 1978)
• anatomically based
• objective
• observational
• Action units (AU)
• Emotional dictionary
• Reliability
• Validity
• Lie detection

18
Behavioral Approach cont

• Other behavioral indicators of lying (see
  DePaulo et al., 2003)
• Blinking more
• Less eye contact
• Less illustrators (modifying or supplementing
  speech)
• More adaptors (e.g., rubbing)
• ? due to nervousness!
• Motivational impairment effect Under a high
  motivation to succeed at lying, some of these
  behaviors change into their opposite.
• ? over-controlled movements
• Behaviors are also idiosyncratic, e.g., moderated
  by personality characteristics.
• ? Knowledge necessary for adequate
  interpretations.
• Deception cues in speech higher pitch, more
  hesitant, less fluent, short response, slow
  speech.

19
Verbal Credibility Assessment

• Verbal deception cues arise from the same reasons
  as nonverbal lying indicators emotion, content
  complexity, and attempted control.
• Undeutsch hypothesis
• Statement Validity Analysis (SVA)
• Interview
• Step Wise Interview Protocol (Yuille, 1990a)
• Cognitive Interview (Fisher Geiselman, 1992)
• Recommended reading Vrij (2000)

20
SVA cont

• Criteria Based Content Analysis (CBCA)
• General characteristics
• Logical structure
• Unstructured production
• Quantity of detail
• Specific content of the statement
• Contextual embedding
• Description of interaction
• Reproduction of conversation
• Unexpected complications during the incident

21
SVA cont

• Peculiarities of the content
• Unusual details
• Superfluous / peripheral details
• accurately reported details misunderstood
• Related external associations
• Accounts of subjective mental state
• Attributions of perpetrators mental state
• Details characteristic of a particular act

22
SVA cont

• Motivation-related contents
• Spontaneous corrections
• Expressing insecurities
• Admitting lack of memory
• Providing reasons for lack of memory
• Raising doubts about ones own testimony
• Self-deprecation
• Pardoning the perpetrator
• Stylistic features
• Reporting style
• Theme related changes
• Rigid repetitions
• ? A semi-standardized approach!

23
SVA cont

• Validity Checklist 2 sets of items
• Attempt to consider alternative explanations!
• ? Re the witness account characteristics and
  the interview
• Psychological characteristics Appropriateness
  of language, Appropriateness of knowledge,
  Presence of affect, Appropriateness of affect,
  Spontaneous gestures, Susceptibility to
  suggestion.
• Interview Characteristics Adequacy of the
  interview, Suggestive or leading questions,
  Pressure or coercion.
• Motivation Context of the original report,
  Motives to report, Pressures to report.
• ? Re other evidence
• Medical evidence, Consistency with the laws of
  nature, Consistency with other statements,
  Consistency with the other evidence.

24
SVA cont

• SVA reliability
• Inter-item consistency only r .15
• Interrater reliability
• up to 100
• rs gt .95
• kappa analyses ? suggest the exclusion of 5 items
  (see Horowitz et al., 1997)
• SVA validity children
• Yuille (1988) A lab study
• Elementary school children tell a true and a
  false story. Blind assessors rated their
  statements according to CBCA.
• Supporting the Undeutsch hypothesis, over 90 of
  the true stories and almost 75 of the false
  stories could be correctly classified based on
  the presence of the three plus two other CBCA
  criteria.
• The procedure produced more false positive than
  false negative errors (a truth bias).

25
SVA cont

• Esplin, Boychuk, Raskin (1988) first field
  study
• several techniques to establish ground truth
  applied
• 20 confirmed vs. 20 unconfirmed cases of child
  sexual abuse
• comparison of mean CBCA scores
• ? 15/19 CBCA criteria significantly more present
  in confirmed cases
• ? 7 CBCA criteria in each confirmed case
• 100 accuracy in classifying doubtful and truthful
  statements.
• CBCA seems to work with adult witnesses as well.
• CBCA is difficult with crime suspects.

26
SVA cont

• Unresolved issues of CBCA
• Coaching similar to countermeasure problem
• Inductivistic approach (searching for signs of
  truthfulness rather than searching for deception)
• Absence of decision rules
• Yuille (1990b) recommended presence of the first
  5 CBCA criteria plus any 2 other
• Insufficient training of experts
• Cultural issues, e.g., certain criteria were seen
  more often in white peoples true statements than
  in blacks.

27
SVA cont

• Admissibility in court
• SVA required in Germany for cases of CSA.
• Also used in the Netherlands and Sweden.
• Inconsistent use in North America
• often accepted in civil but not in criminal
  courts.
• SVA has been admitted in Canadian courts,
  although it remains an issue.
• Honts (1994) Daubert criteria are fulfilled for
  SVA.
• Currently used to guide investigations and to
  gather information.
• Framework reflecting our knowledge of memory,
  deception, interviewing techniques, and cognitive
  abilities of adult and child victims of crime.
• Validity checklist never evaluated.

28
Moderators of the Strength of Deception Cues

• Being a professional does not matter much in most
  cases.
• Poor cue utilization
• Confidence does not equal accuracy.
• Repeated interrogations raise confidence.
• Observing vs. interrogating
• Training should include outcome feedback.
• Left hemisphere damage has positive influence on
  lie detection abilities.
• Gender / Attractiveness The opposite genders
  lies are better detected.
• Personality social anxiousness, self-monitoring,
  tendency to distrust people
• High motivation ? disregard nonverbal cues
• Knowing the liar does not necessarily help.
• Psychopathy do not perform better on the
  polygraph.

29
References

• DePaulo, B. M., Lindsay, J. J., Malone, B. E.,
  Muhlenbruck, L., Charlton, K., Cooper, H.
  (2003). Cues to Deception. Psycholgical Bulletin,
  129(1), 74-118.Ekman, P. (2001). Telling lies.
  New York W.W.Norton Company.
• Ekman, P., Friesen, W. V. (1978). Facial Action
  Coding System. Palo Alto, CAq Consulting
  Psychologists Press.
• Elaad, E. (1998). The challenge of the concealed
  knowledge polygraph test. Expert Evidence, 6,
  161-187.
• Farwell, L. A., Donchin, E. (1991). The truth
  will come out Interrogative polygraphy ("lie
  detection") with event-related brain potentials.
  Psychophysiology, 28(5), 531-547.
• Fiedler, K., Schmidt, J., Stahl, T. (2002).
  What is the current truth about polygraph lie
  detection? Basic and Applied Social Psychology,
  24(4), 313-324
• Fisher, R. P., Geiselman, R. E. (1992).
  Memory-enhancing techniques for inverstigative
  interviewing The Cognitive Interview.
  Springfield, IL Charles C Thomas.
• Honts, C. R. (1994). Assessing children's
  credibility Scientific and legal issues in 1994.
  North Dakota Law Review, 70, 879-903.
• Honts, C. R., Raskin, D. C., Kircher, J. C.
  (1994). Mental and physical countermeasures
  reduce the accuracy of polygraph tests. Journal
  of Applied Psychology, 79(2), 252-259.
• MacLaren , V. V. (2001). A quantitative review of
  the guilty knowledge test. Journal of Applied
  Psychology, 86(4), 674-683.
• Patrick, C. J., Iacono, W. G. (1991). Validity
  of the control question polygraph test the
  problem of sampling bias. Journal of Applied
  Psychology, 76(2), 229-238.
• Raskin, D. C., Kircher, J. C., Horowitz, S. W.,
  Honts, C. R. (1988). Recent laboratory and field
  research on polygraph techniques. In J. C. Yuille
  (Ed.), Credibility assessment. Dordrecht,
  Netherlands Kluver Academic Publishers.
• Vrij, A. (2000). Detecting lies and deceit The
  psychology of lying and the implications for
  professional practice. Chichester Wiley.
• Yuille, J. C. (1988). The systematic assessment
  of children's testimony. Canadian Psychology,
  29(3), 247-262.
• Yuille, J. C. (1990a). Adult "step wise" assault
  interview protocol. Unpublished Manuscript.
  Vancouver, University of British Columbia.
• Yuille, J. C. (1990b). Use of the criteria-based
  content analysis, Unpublished manuscript.
  University of British Columbia. Vancouver.