1
Lie Detection and the Polygraph

For as long as human beings have deceived each other, people have tried to develop techniques for detecting deception and determining truth (see, e.g., Kleinmuntz and Szucko, 1984). These techniques have almost always included interviews and interrogations to try to see through deception and reveal what a deceiver will not freely admit. In the 20th century, lie detection took on scientific aspects with the development of techniques that use measures of physiological responses as indicators of deception. The best known of these is the polygraph. This technique, which relies on physiological measurements developed early in the century, has become for many in the U.S. law enforcement and intelligence communities (including counterintelligence officials in several agencies with whom we met) the most valued method for identifying criminals, spies, and saboteurs when direct evidence is lacking.

Polygraph examinations are widely used in the United States and in some other countries (notably, Israel, Japan, and Canada) for three main purposes:

  1. They are used for preemployment screening in law enforcement and preemployment or preclearance screening in agencies involved in national security. The great majority of U.S. police departments, for example, include polygraph examinations as part of their preemployment screening batteries. Preclearance screening may involve current employ-



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement



Below are the first 10 and last 10 pages of uncorrected machine-read text (when available) of this chapter, followed by the top 30 algorithmically extracted key phrases from the chapter as a whole.
Intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text on the opening pages of each chapter. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

Do not use for reproduction, copying, pasting, or reading; exclusively for search engines.

OCR for page 11
The Polygraph and Lie Detection 1 Lie Detection and the Polygraph For as long as human beings have deceived each other, people have tried to develop techniques for detecting deception and determining truth (see, e.g., Kleinmuntz and Szucko, 1984). These techniques have almost always included interviews and interrogations to try to see through deception and reveal what a deceiver will not freely admit. In the 20th century, lie detection took on scientific aspects with the development of techniques that use measures of physiological responses as indicators of deception. The best known of these is the polygraph. This technique, which relies on physiological measurements developed early in the century, has become for many in the U.S. law enforcement and intelligence communities (including counterintelligence officials in several agencies with whom we met) the most valued method for identifying criminals, spies, and saboteurs when direct evidence is lacking. Polygraph examinations are widely used in the United States and in some other countries (notably, Israel, Japan, and Canada) for three main purposes: They are used for preemployment screening in law enforcement and preemployment or preclearance screening in agencies involved in national security. The great majority of U.S. police departments, for example, include polygraph examinations as part of their preemployment screening batteries. Preclearance screening may involve current employ-

OCR for page 11
The Polygraph and Lie Detection ees who are being considered for new assignments, typically at a higher level of clearance. They are used for screening current employees, especially in security-sensitive occupations. For example, the U.S. Department of Energy polygraph program, established in 1999, mandated polygraph examinations for about 1,300 employees in sensitive positions; a year later, the program was expanded to cover several thousand additional employees (P.L. 106-65 and P.L. 106-398). They are used in investigations of specific events, for instance, in criminal cases. Although there are many restrictions on the use of polygraph results in courts, they are often used to help direct and focus criminal investigations. These three uses of the polygraph raise very different scientific and practical questions, as discussed in this report. The polygraph continues to be the subject of a great deal of scientific and public controversy in the United States. A 1983 report by the U.S. Office of Technology Assessment examining the validity of the polygraph raised many criticisms that are still being voiced. The 1988 Employee Polygraph Protection Act sharply limited the use of polygraphs in employment settings, largely because of doubts about its validity for screening. Different courts have different sets of rules about the admissibility of polygraph evidence and even about what test must be met for such evidence to be considered admissible. Many people find polygraph testing objectionable, and there are several websites and organizations devoted to discrediting the polygraph. It is against this background of continuing controversy that the committee was given the charge to “conduct a scientific review of the research on polygraph examinations that pertain to their validity and reliability, in particular for personnel security screening.” We were also asked to “review other techniques that may be adapted for similar purposes . . . in order to allow for a comparative evaluation of the polygraph and to suggest directions for future research that may include both polygraph and other tests.” Based on our review, we were asked to present our “assessments of and recommendations for polygraph examinations for personnel security purposes” and to suggest further research.1 THE INSTRUMENT, THE TEST, AND THE EXAMINATION Polygraph testing combines interrogation with physiological measurements obtained using the polygraph, or polygraph instrument, a piece of equipment that records physiological phenomena—typically, respiration, heart rate, blood pressure, and electrodermal response (electrical

OCR for page 11
The Polygraph and Lie Detection conductance at the skin surface).2 A polygraph examination includes a series of yes/no questions to which the examinee responds while connected to sensors that transmit data on these physiological phenomena by wire to the instrument, which uses analog or digital technology to record the data. Because the original analog instruments recorded the data with several pens writing lines on a moving sheet of paper, the record of physiological responses during the polygraph test is known as the polygraph chart. A variety of other technologies have been developed that purport to use physiological responses to make inferences about deceptiveness. These range from brain scans to analyses of voice tremors; some evidence relevant to these techniques is discussed in this report. Physiological Phenomena The physiological phenomena that the instrument measures and that the chart preserves are believed by polygraph practitioners to reveal deception. Practitioners do not claim that the instrument measures deception directly. Rather, it is said to measure physiological responses that are believed to be stronger during acts of deception than at other times. According to some polygraph theories, a deceptive response to a question causes a reaction—such as fear of detection or psychological arousal— that changes respiration rate, heart rate, blood pressure, or skin conductance relative to what they were before the question was asked and relative to what they are after comparison questions are asked. A pattern of physiological responses to questions relevant to the issue being investigated that are stronger than those responses to comparison questions indicates that the examinee may be deceptive. The central issues in dispute about the validity of polygraph testing concern these physiological responses. For example, are they strongly and uniquely associated with deception, or are there conditions other than deception that could produce the same responses? Does this association depend on particular ways of selecting or asking questions, and if so, do examiners ask the right kinds of questions and make the right comparisons between the physiological responses to different questions? Is the same association of deception with physiological response observable across all kinds of examinees in all kinds of physical and emotional states? Does it depend on factors in the relationship between examiner and examinee? Is it influenced by an examiner’s expectation about whether the examinee will be truthful? In Chapters 3, 4, and 5 we discuss in more detail the theory of the polygraph and two kinds of evidence on these questions. One comes from basic psychophysiological research on

OCR for page 11
The Polygraph and Lie Detection the phenomena the instrument measures. The other comes from research on polygraph testing itself. Polygraph Test Techniques Although the polygraph instrument is the centerpiece of the technique, the ability of the polygraph test to detect deception also depends critically on other elements of the process. One is the interpretation of the polygraph chart. Interpretation normally involves comparison of physiological responses to “relevant” questions (i.e., questions about the issue that is the focus of the examination) and responses to other questions that are asked for purposes of comparison.3 Interpretation is often done by the examiner, who reviews the chart and may code it according to a standard protocol. People other than the examiner may also use such a protocol to code a chart. Chart interpretation can also be done by computer. Different polygraph techniques are defined in part by the ways the relevant and comparison questions are selected and placed in a polygraph test. A considerable portion of the empirical research on polygraph testing focuses on validating particular techniques or comparing the performance of one technique with another. Three major classes of questioning techniques are in current use. In the oldest of these, the relevant-irrelevant technique, the relevant questions are typically very specific and concern an event under investigation: for example, “Did you rob the bank on Friday?” The irrelevant questions may be completely unrelated to the event and may offer little temptation to deceive: for example, “Is today Monday?” or “Are you in New Jersey?” Stronger physiological responses to relevant than to irrelevant questions are taken as indicative of deception. Although this technique has numerous limitations from a scientific standpoint (Raskin and Honts, 2002), it is used in criminal investigations and in some federal employee security screening programs, for instance, at the National Security Agency. The second class of techniques, called control question or comparison question testing, compares responses to relevant questions with responses to other questions that are intended to generate physiological reactions even in nondeceptive examinees. In one version of this technique, the comparison questions are selected to create a temptation to deceive: for example, “Have you ever stolen a small object from your place of work?” or “Have you ever violated a minor traffic law?” Such so-called probable lie questions are presumed to be like the relevant questions in creating a level of concern related to truthfulness. For truthful examinees, this level of concern is presumed to be higher than for the relevant questions, about

OCR for page 11
The Polygraph and Lie Detection which the examinee can be truthful without much anxiety. For examinees who may be deceptive about the events under investigation, it is presumed that the relevant questions create the greater level of concern and therefore a stronger physiological response. Comparison question tests are used both for specific-event investigations and for screening. A version of the comparison question technique, the Test of Espionage and Sabotage (TES) is a staple of the U.S. Department of Energy’s employee security screening polygraph program. The third class of techniques, commonly called guilty knowledge polygraph testing, involves questions about details of an event under investigation that are known only to investigators and those with direct knowledge of the event. We refer to these tests as concealed information tests because they are applicable even when an examinee who possesses information is not guilty and even if the information is incorrect. The questions are presented in a multiple-choice format. For example, in a burglary investigation: “Where was the place of entry? Was it a: (1) front entrance? (2) kitchen door? (3) bathroom window? (4) balcony? (5) room on the second floor?” (Nakayama, 2002:50). If an examinee who denies knowledge of the event shows the strongest physiological response in several such sets of questions to the alternative that accurately describes the event, the examinee is concluded to have concealed information. Because this test format requires that the examiner have knowledge of the details of a specific event that is the topic of questioning, it cannot be used in typical security screening contexts. Appendix A provides brief descriptions of these basic polygraph questioning techniques and some of their variants. More detail is available from several sources, including the recent Handbook of Polygraph Testing (Kleiner, 2002; especially chapters by Raskin and Honts, Nakayama, and Ben-Shakhar and Elaad). Appendix B provides more detail on how security screening polygraph examinations are conducted in the U.S. Department of Energy and other federal agencies. As these brief descriptions make clear, polygraph testing techniques vary in the ways the relevant and comparison questions differ and in how these differences, combined with an examinee’s physiological responses to them, are used to make inferences about whether the person may be lying in response to the relevant questions. We return to these differences in Chapter 3. In many applications, examiners take a stronger response than to comparison questions as an indication not necessarily of deception, but of the need for further interviewing or testing to determine whether deception is occurring. The lack of such a differential response or a stronger response to comparison questions generally leads to a conclusion that a respondent is being truthful.

OCR for page 11
The Polygraph and Lie Detection Pretest Interview A polygraph test is part of a polygraph examination, which includes other components. A critical one, particularly in comparison question tests, is the pretest interview. This interview typically has multiple purposes. It explains the test procedure to the examinee. It explains the questions to be asked so that examiners and examinees understand the questions in the same way. Shared understanding is especially important for screening polygraphs that ask about general categories of behavior, such as “Have you ever revealed classified information to an unauthorized individual?” The pretest interview shapes the expectations and emotional state of the examinee during the test. It may be used to convince the examinee that the polygraph instrument will detect any deception. This process often involves a demonstration in which the examinee is asked to lie about an unimportant matter, and the examiner shows the instrument’s ability to detect the lie; these demonstrations sometimes involve deceiving the examinee.4 In comparison question testing, the interview is also used to help the examiner decide which questions to ask for comparison purposes. It is important to note that each of these aspects of the pretest interview may influence an examinee’s physiological responses to the relevant or comparison questions and, therefore, the result of the examination. Finally, the polygraph examiner is likely to form impressions of the examinee’s truthfulness, based on the examinee’s demeanor and responses in the pretest interview and during the charting. These impressions, as well as any expectations the examiner may have formed in advance of the examination, are likely to affect the conduct and interpretation of the examination and might, therefore, influence the outcome and the validity of the polygraph examination. Overall Examination A polygraph test and its result are a joint product of an interview or interrogation technique and a psychophysiological measurement or testing technique. It is misleading to characterize the examination as purely a physiological measurement technique. Polygraph examiners’ training implicitly recognizes this point in several ways. It provides instruction on the kind of atmosphere that is to be created in the pretest interview, advises on techniques for convincing examinees of the accuracy of the test, and offers guidance (in different ways for different test formats) for selecting comparison questions. Examiners are advised to control these

OCR for page 11
The Polygraph and Lie Detection details—sometimes following carefully specified procedures—because they can affect test results. Polygraph examination procedures often explicitly combine and interweave testing and interviewing. When a polygraph chart indicates something other than an ordinary nondeceptive response to a relevant question, the examiner typically pursues this response with questioning during the course of the examination. For example, the examiner may say, “You seem to be having a problem in the area of X [the relevant item]” and ask the examinee if he or she can think of a reason for having a strong physiological reaction to that question. The interview may reveal a misunderstanding of the question, which is then explained and reasked in a subsequent charting. Or if the reaction remains unexplained to the examiner’s satisfaction, the issue may be probed in more detail in the interview or with questions in a subsequent charting. Some examiners believe that an important use of polygraph testing is in helping narrow the range of issues that need to be investigated, using both polygraph and other investigative tools. The important role of interview conditions is also recognized in much of the practice and lore of polygraph testing. For example, it is widely and plausibly believed that polygraph results are different for “friendly” and “unfriendly” examinations (e.g., examiners proffered by the defense or by the prosecution in criminal cases). Presumably, examinees are more relaxed with “friendly” examiners and less likely to have responses that indicate deception on the test. When interviewers are hostile or aggressive, examinees may be less relaxed and may produce different physiological responses than those they would produce in response to calm, friendly questioning. Such effects of the interview situation are common in other settings, for example, the widely noted phenomenon of “white-coat hypertension,” in which blood pressure is believed to increase because of the context of a medical examination. These situational effects represent a challenge to the validity of any physiological test that does not adequately reduce the influence of variations in the interview situation on the physiological responses being measured or separate the effects of the situation from the effects of the condition (such as deception) that the test is intended to measure. In polygraph testing, the use of initial buffer items is intended to reduce situational effects on the examinee’s physiological responses. Comparison questions are also used to separate situational effects from the effects of deception by statistical means. Whether these procedures in fact have the desired effects is an empirical question, which is explored in this book.5

OCR for page 11
The Polygraph and Lie Detection THE LIE DETECTION MYSTIQUE In order to frame a scientific discussion about the polygraph, we consider the role of this method of detecting deception in American culture and compare it with methods of detecting deception that have been accepted in other cultures. The polygraph, perhaps more than any other apparently humane interrogation technique, arouses strong emotions. There is a mystique surrounding the polygraph that may account for much of its usefulness: that is, a culturally shared belief that the polygraph device is nearly infallible. Practitioners believe that criminals sometimes prefer to admit their crimes and that potential spies sometimes avoid certain job positions rather than face a polygraph examination, which they expect will reveal the truth about them. The mystique shows in other ways, too. People accused of crimes voluntarily submit to polygraph tests and publicize “passing” results because they believe a polygraph test can confer credibility that they cannot get otherwise. In popular culture and media, the polygraph device is often represented as a magic mind-reading machine. These facts reflect the widespread mystique or belief that the polygraph test is a highly valid technique for detecting deception—despite the continuing lack of consensus in the scientific community about the validity of polygraph testing. Ritualized Lie Detection Across Cultures Ritualized lie detection techniques in many groups, societies, and cultures through the ages share several characteristics that help create a mystique that enables the techniques to be effective. Lie detection rituals involve a socially certified administrator (an examiner or interrogator) and some device or procedure that purportedly can objectively and publicly identify lying on the part of the examinee. The administrator—in some cultures, a priest or shaman—has completed a secret or semi-secret training process. The keeping of the secrets of the ritual within a small, select group adds to the mystique (e.g., the belief that keepers of the secrets have good reason not to publicize them and should be trusted), and, consequently, adds to the power of the technique. The belief structure of the endorsing society includes beliefs about the special powers of the officials authorized to perform the ritual and about the ritual’s ability to divine or elicit concealed truths. The examinee, as a member of the society or culture, generally accepts the importance of the lie detection ritual and believes that it is very accurate. Hence, if he or she is telling the truth, there is little or no reason to fear the examination, but if he or she is lying, there is reason to fear it. Many procedures and techniques have been used in lie detection rituals, including ones that in our society would

OCR for page 11
The Polygraph and Lie Detection be regarded as quite primitive and unscientific, such as immersion in water or placing a wafer on the tongue (see Kleinmuntz and Szucko, 1984). Despite the lack of scientific evidence supporting the validity of such techniques, they apparently are useful, as judged by their ability to elicit confessions of truths that are not forthcoming when other methods are used. Some or all of this usefulness is attributed to mystique—the systems of beliefs that surround and support the techniques. The polygraph testing procedures currently used in the criminal justice system and in several government agencies in the United States and other countries fit this prototype ritual. A polygraph examiner subculture exists, complete with its own institutions (e.g., professional societies), norms, values, etc. Examiners are trained and certified expert by various training institutes, including some private ones and, importantly, by the U.S. Department of Defense Polygraph Institute. Members of the polygraph examiner culture have a particular jargon and shared lore that are generally unknown to others. They also maintain secrets because to reveal too much of their knowledge would enable targets of investigations to “beat” polygraph tests. The polygraph device or instrument is purported to have the power to discriminate lies from truths in the hands of a certified and experienced examiner. The polygraph examination follows standardized, ritual-like procedures and usually occurs in a setting designed to evoke associations with science, medicine, or law enforcement, institutions whose certified practitioners are believed to have special powers to uncover truths. Claims that polygraph testing is a scientific method, together with the establishment of research programs to improve polygraph testing, are useful for building credibility in a society that confers credibility on scientific activities. Moreover, potential examinees are assumed to believe in the validity of polygraph testing, and its validity is supported by popular culture. These similarities between current polygraph detection of deception procedures and the lie detection rituals of other and former cultures say nothing directly about the validity or invalidity of the polygraph testing for distinguishing truth from deception. They do, however, suggest that some of the value or utility of the polygraph for eliciting admissions and confessions undoubtedly comes from attributes other than the validity of the testing itself. Polygraph testing may work, in part, because it capitalizes on the mystique that is common to lie-detection rituals in many societies. Any investigation into the scientific validity of polygraph detection of deception must try to identify and distinguish between two kinds of scientific evidence: evidence bearing on the effects of the polygraph ritual and mystique and evidence bearing on the validity of polygraph testing and the polygraph device for detecting deception. Any scientific investigation must also deal with some of the cognitive

OCR for page 11
The Polygraph and Lie Detection and organizational phenomena that go along with a ritual that has a mystique, a “priesthood,” and a set of secrets. One of these is the difficulty of gaining access to information. Some information of interest to this study, such as the polygraph test records of known spies, is classified for national security reasons. Other information, such as the precise ways particular pieces of polygraph equipment measure physiological responses, is guarded by equipment manufacturers as trade secrets. Some manufacturers ignored our requests for such information, even though we offered to sign legally binding promises of nondisclosure. Information about computer scoring algorithms for polygraph tests was similarly withheld by some algorithm developers. All of this behavior makes scientific analysis difficult. Some of these “secrets” probably have good practical justification, but they are also very much like the activities of a priesthood keeping its secrets in order to keep its power. Another aspect of the polygraph mystique that creates difficulties for scientific analysis is the strong, apparently unshakeable, beliefs of many practitioners in its efficacy on the basis of their experiences. We have heard numerous anecdotes about admissions of serious crimes and security violations that have been elicited in polygraph examinations even after background checks and ordinary interviews had yielded nothing. Many of these admissions have been later corroborated by other convincing evidence, indicating that the polygraph examination sometimes reveals truths that might otherwise have remained concealed indefinitely. We do not doubt the veracity of these anecdotes. However, they do not constitute evidence that the polygraph instrument conveys information that, in the context of the polygraph test, accurately identifies the locus of deception. Rather, they signify that something in the polygraph examination can have this result. It may be the test, the interviewer’s skills, the examinee’s expectation of detection, or some combination of these or other factors. From a scientific standpoint, these anecdotes are compelling indications that there is a phenomenon in need of explanation; they do not, however, demonstrate that the polygraph test is a valid indicator of deception. Practical Implications From a practical standpoint, it can make a considerable difference whether decisions that rely on polygraph evidence are resting on a scientifically proven device and procedures (that is, on the test), on the judgments of examiners, or on the expectation that guilty examinees will be sufficiently fearful of detection to confess. For example, if the apparent successes depend only on examinees’ fear of detection and not on the test itself, the examination would fail with well-trained spies who know the test’s limitations and do not respond to the mystique.

OCR for page 11
The Polygraph and Lie Detection Polygraph examiners and the decision makers who use their reports do not always make such distinctions. The belief among many agency officials that the important questions about polygraph testing validity have already been favorably resolved makes it difficult to conduct scientific analysis of the components of polygraph testing, including the polygraph instrument itself, in those agencies. It also creates resistance to scientific evidence critical of the test’s validity among practitioners whose personal experience has convinced them of the polygraph’s utility. Finally, placing polygraphic detection of deception within the anthropological and historical context of lie detection rituals strongly suggests that the mystique will outlive current lie detection techniques, including the polygraph test. We surmise that if the mystique of lie detection no longer attaches to the polygraph, a new technique or instrument will take its place and assume its mystique. Indeed, some people argue that the mystique has already been dispelled, as exemplified by the controversy over polygraph security screening that led to the request for this study. It is therefore not surprising that in the current context of heightened concern about espionage and terrorism, there is a lot of publicity about new devices and techniques for the psychophysiological detection of deception. This interest reflects both the need for security and at least latent doubts about the validity of polygraph testing procedures. As discussed in this report, the scientific criteria that should be used to evaluate new devices and procedures are the same as those that apply to the polygraph. SCIENTIFIC ISSUES Detecting Deception and Eliciting Truth For a criminal investigator or a counterintelligence officer, detecting deception and eliciting truth are opposite sides of the same coin. It does not matter whether deception is detected in an interviewee’s physiological responses or whether truth is elicited in the form of an admission or revealed by a combination of physiological responses and further interrogation and investigation. Such distinctions are not made in official reports on polygraph screening programs. What matters most to investigators and is reported to Congress are the number of examinees who were ultimately “cleared,” the number subjected to adverse personnel actions, and the security violations revealed. From a scientific standpoint, however, detecting deception and revealing truth are two distinct purposes of polygraph examinations or any other technique for the psychophysiological detection of deception. The polygraph test is advocated as an accurate psychophysiological indicator of deception. The polygraph examination, which includes the test and the interrogation surrounding it, is a tool for revealing truth. To

OCR for page 11
The Polygraph and Lie Detection evaluate the accuracy of polygraph tests, it is imperative to distinguish several different roles of the polygraph test in polygraph examinations, some of which do not depend on whether the test provides a valid indicator of deception. One role of the polygraph test is to help elicit admissions from people who believe, or are influenced to believe, that it will accurately detect any deception they may attempt. This role is demonstrated most clearly when a polygraph examination is terminated because of an admission before any charts are done. Such an examination can be thought of as an interrogation interview conducted in the presence of a polygraph. In this case, the polygraph test has a useful role independently of whether it can accurately detect deception: it is effective if the examinee believes it can detect deception. Admissions of this kind provide evidence of the value of the polygraph examination for investigative purposes, but they do not provide evidence that the polygraph test accurately detects deception. Another role of the polygraph is to test cooperation with an investigative effort. Sometimes a polygraph examination is terminated or leads to an assessment that the examinee is deceptive because of detected or suspected countermeasures during the test.6 If an examinee is judged to be using countermeasures, that is taken as evidence that the examinee is not cooperating with the investigation, particularly if the test protocol asks the examinee not to use countermeasures. Noncooperation is in turn taken as a reason to suspect deception. Holding aside the question of whether such inferences are valid, the use of the polygraph in this way does not depend on the scientific validity of the test. A third role of the polygraph test is to influence the conduct of a polygraph interview. A polygraph examiner who detects what he or she believes to be deceptive responses during the polygraph test normally conducts the remainder of the interview differently than an examiner who sees no signs of deception. Such an examiner may ask more probing questions, do additional charting, shift to a different type of polygraph test protocol, or take a more confrontational attitude in the interview in an effort to elicit an admission or to “clear” the examinee of suspicion. In this situation, it is impossible without careful experimental analysis to disentangle the effect of polygraph validity from other elements of the interaction in the examination. Finally, polygraph chart readings may be used directly to make inferences about truthfulness or deceptiveness. Assessments of the scientific validity of the polygraph test as a technique for the psychophysiological detection of deception should properly be made on test outcomes that depend only on chart scoring.7 However, it can be difficult or impossible to consider chart results in isolation because of the likelihood that the examiner’s behavior during the test is affected by prior expectations, the

OCR for page 11
The Polygraph and Lie Detection pretest interview, and his or her initial interpretations of a chart. Despite such difficulties, it is important to distinguish between the use of the polygraph as a diagnostic test of deception, in which the charts are scored and decisions are made on the basis of the score, and its use as part of an interrogation procedure. Purposes of Polygraph Testing As we note at the beginning of this chapter, polygraph testing and interviewing are used for three main purposes: event-specific investigation, employee screening, and preemployment (or preclearance) screening. These different purposes are reflected in different kinds of questions that are asked in polygraph tests. For an event-specific investigation, the polygraph is used to investigate a specific incident, such as a crime or a specific act of sabotage or espionage. In this case, it is possible to ask relevant questions that are highly specific, such as “Did you plant the bomb that exploded at location X on June 12?” or “Was the murder committed with a knife?” Relevant questions like these are highly specific to a known event about which a guilty person may have a strong motive to lie or to conceal information. For employee screening, the polygraph is used with current employees who may have committed acts prohibited by their employer or by law, but there is usually no specific known act that is the focus of the examination. Relevant questions in a security screening context might include “Have you released classified information to any unauthorized person?” or “Have you had any unreported contacts with a foreign government representative?” Some analysts believe that such questions, because they do not refer to specific past events, are more similar to comparison questions than are the relevant questions that can be asked in an event-specific investigation. For this reason, it has been argued that it is inherently more difficult to discriminate deception from truthfulness in a screening context (Murphy, 1993). For preemployment screening or preclearance screening of employees being considered for new job assignments, the polygraph is used to try to determine the potential for future acts. For example, when someone is given a polygraph examination as part of an application to do intelligence work or for a new assignment that requires access to classified information, the employer’s concern may be with the potential that the person may commit an act in the future that he or she is not at present in a position to commit. In this situation, “relevant” questions can only be about unspecific past acts that are different in kind from the ones of greatest concern. Deception can be inferred from the polygraph in the same way it is done in screening current employees. However, in making

OCR for page 11
The Polygraph and Lie Detection inferences from indications of deception, it is necessary to make one additional assumption: that a person who is deceptive about certain undesirable past acts is at risk for committing different kinds of undesirable acts in the future.8 Some polygraph test situations, which can be described as focused screening situations, do not fit neatly into the above three categories because they have attributes of both the screening and the specific-incident investigation purposes. An example might be the investigation of a fairly large group of individuals who are suspected of involvement with a known terrorist organization. Such investigations are like typical screening situations in that there is no known specific incident that can be the focus of questioning, but they are like specific-event investigations if it is possible to ask specific questions about the organization, its leaders, or the places in which it operates. Strong physiological responses to such specific questions might indicate that the examinee has information about the terrorist organization and should be investigated more fully regarding possible ties to it. If the answers to such questions are likely to be known only to the investigators and to the organization’s members and close associates, the situation is amenable to the use of tests of the concealed information type, which are not otherwise considered to be applicable to screening situations. The ability of polygraph testing to uncover the deceptions of interest and to serve broader law enforcement or national security goals may depend on the purpose of the test and the kinds of acts that are the subject of the relevant questions. It is plausible that the task of the polygraph is easiest in event-specific investigation and hardest in preemployment screening. The possibility that accuracy depends on the purpose of the test makes it unwise to assume that accuracy estimates calculated from data when the polygraph is used for one purpose are pertinent to its use for a different purpose. Our study focused on the use of polygraph examinations for employee and preemployment screening. However, one of the critical limitations of the available research is the extreme paucity of studies that directly address the validity of the polygraph for current or preemployment screening. Most of the scientific research considered in this report deals with the use of the polygraph for event-specific investigations. Unfortunately, the relevance of such research for the screening context is not self-evident. As we note in Chapter 2, the sorts of decisions made in screening contexts (e.g., a forecast of whether a job applicant might pose a future risk) and event-specific investigations (e.g., an assessment of whether a suspect is truthful when denying a crime) are so fundamentally different that even the best event-specific research may not be relevant to the validity of the polygraph for employee or preemployment screening.

OCR for page 11
The Polygraph and Lie Detection CONTEXT OF POLYGRAPH TESTS Polygraph examinations are not the only source of information used to determine an examinee’s truthfulness or deceptiveness. In event-specific investigations, a variety of techniques of criminal or security investigation are used, and it is often these that lead to the selection of the individuals (suspects) for polygraph testing. In pre-employment screening, employment questionnaires and interviews, as well as background checks, may supplement information from polygraph tests. In employee screening, periodic or occasional polygraph examinations may be supplemented by interviews and investigations, especially if the polygraph test result is inconclusive or shows a significant response that remains unexplained. In short, information from polygraph examinations may be combined in many ways with information from other sources in judging truthfulness or deception. Policy decisions on the use of the polygraph must therefore consider not only the information that can be gained from the polygraph alone, but also the value it may add to what can be learned from other available investigative techniques. Furthermore, besides the additive value of polygraph information, the polygraph test may influence or be affected by other forms of investigation in known and unknown ways. For example, evidence about a crime may identify certain suspects who are then given a polygraph test, or a polygraph test result may lead an ongoing investigation to focus on one person and turn away from others. Such interactions can make it difficult to separate the effects of the polygraph test from those of concurrent investigative methods. The value, or utility, of polygraph testing does not lie only in its validity for detecting deception. It may have deterrent value, for instance, if people do not take certain actions because they fear that a polygraph examination will uncover them. It may help focus an investigation on particular aspects of a case highlighted by an examinee’s physiological responses. And, as noted above, polygraph testing may elicit admissions or confessions of undesired activity from people who believe they are better off to admit certain activities voluntarily than to submit to a polygraph test and risk being accused of these or more serious activities, as well as being accused of deception. These admissions or confessions may occur during the polygraph examination, either before charts are collected or in response to an examiner’s questions about the charts. These kinds of utility do not depend on validity in the sense that polygraph policies may yield deterrence, admissions, and confessions when a potential examinee believes that the polygraph will detect or has detected deception, even if scientific evidence does not support such a belief.9 We discuss utility in more detail in Chapter 2, along with its relationship to the investigation of validity.

OCR for page 11
The Polygraph and Lie Detection STRUCTURE OF THIS BOOK This book reviews the scientific evidence on the validity of polygraph testing, giving special attention to the use of the polygraph for employee screening for national security purposes. To do this, we consider all the available scientific evidence on polygraph validity, as well as evidence on a number of alternative techniques and technologies for detecting deception. Chapter 2 discusses the concept of validity as it applies to the psychophysiological detection of deception, distinguishes validity from utility, and explains the measure we have chosen as an index of the accuracy of the polygraph. It covers issues of definition and measurement that are important for understanding how we conducted this study but that may not be of interest to readers concerned mainly with its results. Chapter 3 discusses theories of the polygraph and summarizes the basic scientific knowledge, mainly in psychology and physiology, relevant to polygraph validity. A solid scientific base is necessary if one is to have confidence in the validity of psychophysiological detection of deception across a wide range of settings, and the chapter evaluates this scientific base. Chapters 4 and 5 summarize and evaluate the evidence on the accuracy with which polygraph tests detect deception in experimental simulations and field settings. Chapter 6 discusses a number of alternative techniques for detecting deception that have been suggested as supplements to or replacements for the polygraph and evaluates the research on them. Chapter 7 discusses the issues raised by using polygraph evidence for making practical decisions, particularly in security screening processes, including the issue of combining polygraph evidence with other sources of information. Chapter 8 presents the committee’s conclusions about the validity of polygraph testing and its recommendations about the use of the polygraph in employee security screening. It also presents a set of guidelines that should be considered in evaluating emerging techniques for lie detection and offers recommendations for future research. The appendixes provide technical detail and documentation of certain points of the study and are designed for technically oriented readers.

OCR for page 11
The Polygraph and Lie Detection NOTES 1.   More specifically, the work plan for the study calls for the National Research Council “to conduct a scientific review of the research on polygraph examinations that pertain to their validity and reliability, in particular for personnel security screening. The review would include what is known about the effect of medications, sleep deprivation, and illnesses on the physiological responses measured. . . . “The panel would review other techniques that may be adapted to similar purposes, such as research on facial expressions and voice stress analysis, in order to allow for a comparative evaluation of the polygraph and to suggest directions for future research that may include both polygraph and other tests. The panel will not, however, independently review and assess these other techniques nor assess the use of the polygraph in conjunction with other techniques. . . . “The report would present the panel’s assessments of and recommendations for polygraph examinations for personnel security purposes and the panel’s suggestions for further research.” 2.   Some standard definitions of key terms can be found in the U.S. Department of Energy’s Polygraph Examination Regulations (10 CFR, Part 709.3): Polygraph means an instrument that (1) Records continuously, visually, permanently, and simultaneously changes in cardiovascular, respiratory, and electrodermal patterns as minimum instrumentation standards; and (2) Is used, or the results of which are used, for the purpose of rendering a diagnostic opinion regarding the honesty or dishonesty of an individual. Polygraph examination means a process that encompasses all activities that take place between a polygraph examiner and individual during a specific series of interactions, including the pretest interview, the use of the polygraph instrument to collect physiological data from the individual while the polygraph examiner is presenting a series of tests, the test data analysis phase, and the post-test phase. Polygraph test means that portion of the polygraph examination during which the polygraph instrument collects physiological data based upon the individual’s responses to test questions from the examiner. Our usage is consistent with these definitions. 3.   There is much debate in the polygraph research literature on the relative validity of control question or comparison question tests vis-à-vis other kinds of tests, particularly guilty knowledge or concealed information tests, which are not based on the same kinds of comparisons. Notwithstanding this scientific issue, all polygraph tests involve comparison of physiological responses to questions that bear directly on the issue being investigated with responses to other questions, however named, that are used for purposes of comparison. 4.   These demonstrations are commonly referred to as stimulation tests or acquaintance tests. They are normally described to the examinee as procedures designed to acquaint the examinee with the equipment and to determine whether the examinee can make the physiological responses used in the test. The examinee is connected to the polygraph equipment and asked to pick a card or select a number within a specified range. He or she is then asked to respond “no” to each of a series of questions of the form, “Was the number 4?” After the series of questions, the examiner, who in some versions of the demonstration has knowledge of the examinee’s choice by a subterfuge such as a stacked deck of cards, reviews the chart with the examinee and shows that the polygraph was able to detect deception when the examinee did lie.

OCR for page 11
The Polygraph and Lie Detection 5.   Some researchers make the plausible claim that comparison questions are a more effective protection against situational effects in concealed information formats than in comparison question tests because an examinee who lacks the concealed information will be unable to discriminate between the relevant and comparison questions and will therefore not have a different physiological response to the relevant question. 6.   Countermeasures are actions taken by an examinee to influence the physiological responses being measured and thereby produce a test result that indicates truthfulness. 7.   When a polygraph test is scored from a chart alone, scorers are normally provided with the questions that were asked and the temporal point on the chart when each question was asked. 8.   This assumption must be made in any preemployment screening test and is not unique to polygraph screening. 9.   It is also possible for polygraph examinations to result in false confessions, just as with other interrogation techniques (Kassin, 1997, 1998). False confessions should probably be counted as evidence against the utility of polygraph examinations.