Lie Detection Procedures with Recommendations on their Suitability for Use in Court - Essay Example

? Psychology Critical Evaluation of Psychological Research on Lie Detection Procedures with Recommendations on their Suitability for Use in Court University Date Abstract This paper revisits the question of lie detection in the criminal justice system via a critical review of psychological research on this topic from the latter quarter of the twentieth century onwards. It outlines three critical issues in lie detection. The first of these is the identification of specific cues and indicators that are believed to occur when lying takes place, including both verbal and nonverbal elements. The second issue reviewed is that of methods of analysis that have customarily been used in this field including polygraph tests and observation of videotaped interviews. The third issue is the question of how to analyse the data received, and the difficulties encountered in obtaining reliable results. In the light of the evidence presented here, this paper concludes that useful information in crime investigation can be gathered using these methods, but the level of error is high, and in particular the variation between individual suspects, different contexts, and individual human evaluators, even when highly trained, means that results are unreliable. The probability of misleading information being introduced to a court context is therefore unacceptably high, and it is recommended that no change be made to the current ban on lie detection evidence in court. Introduction. In a criminal justice context, lie detection procedures are based upon the assumption that people who are deliberately attempting to deceive law enforcement personnel will demonstrate observable behaviour that is different from the behaviour of people who tell the truth when being questioned. Psychologists generally accept this basic premise but there has been considerable debate about how to design methods for detecting these differences, how to devise systems for interpreting the results. There is a considerable body of research devoted to the task of establishing exactly what aspects of observable behaviour are linked with lying, determining how this behaviour can be recorded, measured and interpreted, and deciding how reliable any conclusions are likely to be. This paper presents a critical review of recent research on this branch of psychology, under three main headings of 1) cues and indicators of lying, 2) methods of analysis, and 3) evaluating results and reliability issues. On the basis of this evidence a number of recommendations are then made regarding the extent to which lie detection techniques could or should be admissible in a court of law. 1 Cues and indicators of lying. Lying is a part of normal human behaviour, since people often misrepresent their own true feelings in order to avoid upsetting other people, and to represent themselves in positive way. Everyone expresses positive opinions, for example, when required to comment on another person’s appearance, and these socially acceptable lies are accomplished very convincingly all the time. The context of criminal justice system is more concerned with the deliberate intention to mislead and usually brings much higher stakes into the act of producing deceitful evidence by means of lies. Research in the United States in the 1970s and 1980s into the detection of deceit has produced rather disappointing results: “In every study reported, people have not been very accurate in judging when someone is lying.” (Ekman and O’Sullivan, 1991, 913) Even more worrying was the finding that professional lie catchers such police officers and customs officials were no more accurate than college students in determining whether people were lying or not. (Kraut and Poe, 1980, 796-797) Since these results were little better than what might be expected through chance, research psychologists began to focus on determining exactly which observable cues indicated lying. Ekman is recognised as one of the first to design experiments which started to pinpoint once and for all which cues were clear indicators of lying. Using his experience as a medical professional, he and his colleagues started off by investigating videos of patients suffering from severe depression to try and work out whether or not they were telling the truth about improvements in their mood. This is an important area for psychologists, because such patients are released from hospital on the basis of whether or not doctors believe the assurances that patients give. (Ekman, 2001, 16-18) They discovered when watching an interview recording in slow motion mode the existence of “micro expressions” and “micro gestures.” This led to a series of further investigations looking for nonverbal cues that human beings may display in fleeting fractions of a second. Ekman also detected verbal features such as higher pitch of the voice, more pauses and more speech errors and attributed those to fear of detection, surmising also that the more serious the consequences of detection, for example in a murder case, the more obvious these cues would be. (Ekman, 2001, 251-253) This work established the fact that nonverbal cues are an important dimension which must not be ignored in lie detection work. A useful meta-analysis of articles on cues to deception was conducted by DePaulo et al. (2003) They note the four categories derived by Zuckerman et al.(1981) namely i) generalised arousal – indicated by greater pupil dilation, increased blinking, more frequent speech disturbances and higher pitch; ii) specific affects, or feelings, experienced during deception – such as guilt, indicated by evasiveness, less eye contact; iii) cognitive aspects of deception – indicated by longer delays in answering questions, more speech hesitations and fewer illustrative hand movements, as subjects concentrate more on constructing their story and iv) attempted control of verbal and nonverbal behaviours – indicated by less spontaneity and a discrepancy between verbal and non-verbal elements of their behaviour. (Zuckerman et al., 1981, quoted in DePaulo et al., 2003, 75) A feature of all four categories is that they include often tiny variations which are difficult to measure, and would need to be correlated against some norm or other in order to determine whether or not they are significant. Both of these are huge difficulties for researchers designing experiments on lie detection, and they transfer also to any practical application in the criminal justice arena. These categories also mix verbal and nonverbal elements, suggesting that both are important in lie detection work. 2 Methods of analysis. The use of machines such as the polygraph is often presented as a means of detecting lies, but in actual fact all that the machine can do is detect physical changes in the body such as sweating, respiration and blood pressure. These changes can be related to emotional changes, but they are not necessarily indicators of deceit. (Ekman, 2001, 51) The problem with all such mechanistic methods of lie detection is that they focus too much on physical cues and usually do not correlate them with verbal cues. There is also a significant stress and arousal factor, which polygraph operators often exploit to put pressure on subjects, and this undoubtedly can result in a significant false positive effect. An experiment conducted using polygraph techniques on both psychopathic and non-psychopathic prison inmates found that “in situations in which there is genuine concern about the outcome of a control question polygraph examination, a substantial proportion of innocent subjects may react more strongly to the relevant questions, producing deceptive test results.” ( Patrick and Iacono, 1989, 353) This is a very worrying feature of the polygraph method, as indeed is the likelihood that some subjects who demonstrate pathological behaviour may not react in a typical way to standard questions, and this could result in false negatives. The link between physiological cues and deceit is therefore far from clear. Video recordings of witness statements are perhaps a more reliable collector of cues, since they can be observed again and again, or slowed down to observe the tiniest changes in behaviour. There is some difficulty, however, in deciding which features of a person’s behaviour, or cues, are relevant to the subject being discussed, and which might be due to unrelated factors. It is well known that human personalities differ greatly, and that people express their thoughts and emotions in a wide range of possible ways. The human mind is complex, and there is no sure way to connect observed behaviours, even when involuntary and unconscious, with particular instances of guilt or innocence, or deliberate attempts to lie about something. A small study involving a sample of only five student observers first trained these individuals in spotting a range of verbal and nonverbal cues and then measured their ability to make a “quick global assessment” or “rapid judgement” of videotapes made of 26 liars and 26 truth tellers (again drawn from the student population, although checks were made to ensure that the observers were not acquainted with the videotaped volunteers). (Vrij, Evans et al., 2004, 285) This method of training had an emphasis on detecting lies implicitly, through body language, although explicit and verbal cues were also sought. The variables that observers were trained to look for were extensive, namely: latency period between question being asked and answer being given; hand and finger movements; speech hesitations such as “ah” and “mm”; quantity of details given; contextual embeddings e.g. time and location; reproduction of conversation; description of other’s mental state; visual details; auditory details; spatial information; temporal details; cognitive operations such as thoughts and reasonings. The results indicated that observers achieved an accuracy rate of 74% overall, which is rather high, with the exception of their ratings of the last variable, cognitive operations, and this variable was therefore dropped from the study. This looks impressive, but the tables showing the detail of analysis reveals considerable variations which would be very significant in a real criminal investigation. Truths were detected with an 84% accuracy rate, while lies were detected with a 65% accuracy rate. Total accuracy rates for the five observers varied from a low of 56% to a high of 86% (Vrij, Evans et al., 2004, 249-251) These considerable ranges suggest that lie detection based on video recordings is by no means an exact science. 3 Evaluating results, and reliability issues. The studies reviewed above reveal that there is considerable variation both in the use of equipment and methods, and in the ability of individuals to detect lies and deception. Vrij notes in a survey of dozens of empirical studies a range of between 20% and 80% accuracy in detecting lies. These articles cover experiments involving students, police officers and laypersons, with both laypersons and professionals attaining an average accuracy of around 56%. (Vrij, 2000, 163) This evidence suggests that application of truth and lie detection systems in criminal justice is fraught with risks. A small number of studies have demonstrated higher accuracy rates than the average and it seems likely that by eliminating errors, and improving training of observers, better results can be obtained. In an empirical test (Ekman and O’Sullivan, 1991, 919-920) involving both professional lie-catchers and laypersons, it was noted that younger people did better than older people in detecting lies, and that secret service personnel were indeed more skilled than other groups. The most accurate detectors used a combination of verbal and nonverbal cues. Interestingly, the professional polygraphers noted that they depended on asking specific questions and analysing the polygraph machine charts in their work, which meant that they routinely did not pay much attention to the dimension of nonverbal clues. In the case of judges, the face of witnesses was often not visible to them, and so once more, nonverbal clues were not available to them. An interesting study on professional lie detector beliefs found that “a comparison of the research findings concerning actual and perceived indicators of deception shows that observers are not very knowledgeable about actual indicators of deception.” (Vrij and Semin, 1996, 66) In fact the evidence shows that observers often interpret gaze aversion, trunk movements, self-touches and shifting position as indicators of deception, when in fact deception is more often associated with a decrease in movements, due to the fact that more control is being exercised by the deceptive subject. (Vrij and Semin, 1996, 67) This study hypothesised that the people best placed to decode verbal and nonverbal cues were in fact not the people conducting the tests, but the subjects themselves. The logic of this is that people like prisoners have much more experience of feedback on their own lie attempts, and so they have a better working knowledge of which strategies and cues are successful, and which to avoid if they wish to deceive someone while being questioned. The authors conclude with a piece of timely advice on the need for training of professionals: “the fact that lie experts had incorrect beliefs about indicators of deception makes workshops on behavioural clues to deception relevant.” (Vrij and Semin, 1996, 65) These comments reveal that there may be a considerable gap between the emergent findings of the psychologists and the current practice of professionals who use these methods. Given the high stakes involved for both subjects and professionals, this is an area of considerable concern. Conclusion. The review of research above shows that there is indeed some reason to believe that carefully trained observers can identify cues that indicate deception. Videotaping suspects and conducting polygraphs can produce new evidence which can then be analysed, and used to form hypotheses with varying levels of probability. An important proviso is, however, that any professionals who use such methods need to be properly qualified and trained in the spotting and analysis of cues. An appropriate application for these methods would be to use them as a complement to other investigative procedures. Results of lie detection could, for example, guide investigators to prioritise particular lines of investigations in order to find more evidence that can be used to prove or disprove the subject’s verbal account. Lie detection systems have a number of disadvantages, however, including the fact that they consume a very high level of time and resources in order to produce a very slight result. Nonverbal cues have to be first recorded on video and then examined many times, often using slow motion and complex scoring frameworks, before any results can be obtained. Verbal cues need to be transcribed and analysed, with due regard for many variables including contextual, age and personality of the subject, and of the observers. Training is essential, and the extensive and on-going research in this field suggests that updating in this field will be necessary for decades to come, as psychologists refine their methods and discover more about deceit and its expression in human subjects. The most serious disadvantage of lie detection methods, in their current form, however, is that they produce high levels of error, and can never be relied upon to provide 100% accurate results. Findings are always open to interpretation and can be easily challenged on a number of different counts. Lie detection is improving, as researchers learn more about human behaviour, but there is still much to be discovered. The fact remains that lie detection is an inexact science, which produces huge individual variation, and for this reason it is recommended that the current ban on the use of lie detection evidence in court should remain in place. References De Paulo, B.M., Lindsay, J.J., Malone, B.E, Muhlen, L., Charlton, K. and Cooper, H. (2003) Cues to Deception. Psychological Bulletin 129 (1), 74-118. Ekman, P. (2001) Telling Lies: Clues to Deceit in the Marketplace, Politics and Marriage. 3rd edn. [first published in 1985]. New York: Norton. Ekman, P. and O’Sullivan, M. (1991) Who Can Catch a Liar? American Psychologist 46 (9), 913-920. Kraut, R.E. and Poe, SD. (1980) On the Line: The deception judgments of customs inspectors and laymen. Journal of Personality and Social Psychology 39, 784-798. Mann, S., Vrij, A. and Bull, R. (2004) Detecting True Lies: Police Officers’ Ability to Detect Suspects’ Lies. Journal of Applied Psychology 89 (1), 137-149. Patrick, C.J. and Iacono, W.G. (1989) Psychopathy, Threat, and Polygraph Test Accuracy. Journal of Applied Psychology 74 (2), 347-355. Vrij, A. (2000) Detecting Lies and Deceit: The psychology of lying and the implications for professional practice. Chichester: Wiley. Vrij, A., Evans, H., Akehurst, L., and Mann, S. (2004) Rapid judgements in assessing verbal and nonverbal cues: their potential for deception researchers and lie detection. Applied Cognitive Psychology 18 (3), 283-296. Vrij, A. and Semin, G.R. (1996) Lie Experts’ Beliefs About Nonverbal Indicators of Deception. Journal of Nonverbal Behavior 20 (1), 65-80. Zuckerman, M., DePaulo, B.M. and Rosenthal, R. (1981) Verbal and nonverbal communication of deception. In L. Berkowitz (Ed.), Advances in experimental social psychology Vol 14, 1-59. New York: Academic Press. Read More