The Physiological Detection of Deception

William M. Waid

Martin T. Orne

The observation that behavior involving conflict and emotion is typically accompanied by marked physiological changes has been the basis of one of the oldest and most widely used psychological technologies -- the physiological detection of deception. Jung (1906) originally envisioned the basic principle as an aid in diagnosis and psychotherapy. Items on a word-association test that evoked larger than usual galvanic skin responses were scrutinized as possible areas of conflict for the patient. Although interest in such an approach in therapy has only recently been rekindled (Abrams 1973), the adoption of analogous techniques in criminal investigation was swift and widespread (Marston 1917). Today their use extends to routine screening of employees for possible malfeasances (Lykken 1974, 1981).

The idea that the detection of deception is merely a kind of psychometric instrument, or a psychophysiological technique, must be considerably revised to take into account the social nature of both deception and its detection. The investigation of social and other factors in deception had become a fundamental area of inquiry in psychology. Researchers such as Ekman and Friesen (1974) and DePaulo and Rosenthal (1979) have devoted particular attention to the involuntary gestures and vocal changes that accompany deception; others, such as Barland and Raskin (1973) and Waid and Orne (1981), have focused on accompanying physiological changes as exemplified by the polygraph -- “lie detector” -- test.

The controversy about the accuracy of current polygraph test techniques in scientific journals as well as in congressional hearings has had the salutary effect of sharpening the debate concerning the precise scientific status of the polygraph test as a means of assessing the veracity of testimony (Lykken 1974, 1978, 1979; Podlesny and Raskin 1977; US Senate 1978). Although it is important for public policy purposes to obtain valid estimates of the accuracy of the techniques, it would be unfortunate if the current controversy obscured the need to understand the precise processes by means of which physiological responses may accurately reflect veracity, deception, or other psychological states. Current polygraph testing procedures, which appear to be less accurate than claimed by some field polygraphers, do not necessarily represent the ultimate state of the art.

The physiological detection of deception involves comparing physiological responses to relevant questions (which evoke a lie from a deceptive person) with responses to control questions. The exact nature of these questions, particularly the control questions, depends on the examiner and the purpose and type of test, as do the collection and use of the physiological data. Some recent reviews have focused on the psychometric issues involved -- that is, the nature of the questions asked, the structure of the test, and the validity and reliability of different approaches (Lykken 1974, 1981) -- while others have focused on identifying the most accurate physiological response or combination of responses (Podlesny and Raskin 1977).

The purpose of the research described below, however, was to examine the underlying cognitive, social, and personality processes involved in the physiological response accompanying deception. Deception is only one of many social behaviors in which physiological arousal plays a role, and there is no specific pattern of physiological response to deception (Orne et al. 1972). Understanding how fundamental processes contribute to physiological response would help clarify the reciprocal effects of social, psychological, and physiological processes in general. The control of such processes, whether experimentally or statistically, might improve the accuracy of polygraph test results. Before describing this research, we shall review the nature of field polygraph testing and the scientific controversy surrounding it.

Field polygraph testing

An individual who takes a polygraph test usually does so at the request of the police, his attorney, or even his employer; an individual may, of course, volunteer to take a test to clear himself of suspicion. Typically, though, the

William M. Waid is Research Psychologist at the Institute of Pennsylvania Hospital and Clinical Assistant Professor of Psychology in Psychiatry at the University of Pennsylvania. He received his B.A. from the University of New Mexico and his M.S. and Ph.D. from Rutgers. He has also published on the physiological correlates of delinquency proneness, and is the editor of a forthcoming book entitled Sociophysiology. Martin T. Orne is Director of the Unit for Experimental Psychiatry at The Institute of Pennsylvania Hospital and Professor of Psychiatry at the University of Pennsylvania. He received his B.A., M.A., and Ph.D. from Harvard University and his M.D. from Tufts University Medical School. The editor of the International Journal of Clinical and Experimental Hypnosis for the past 20 years, he has studied hypnosis, EEG alpha biofeedback, and sleep. The research reported here was supported in part by a grant from the Institute for Experimental Psychiatry. This article is an updated version of a chapter published in Advances in Experimental Social Psychology, vol. 14 (Academic Press, 1981). Address: The Institute of Pennsylvania Hospital, 111 North 49th St., Philadelphia, PA 19139.

402

403 1982 July -- August

Figure 1. In a typical polygraph examination, a blood-pressure cuff is attached to the subject’s arm, electrodes for recording the electrodermal response are attached to the fingertips, and sensors for recording respiration go around the chest and abdomen. (Photograph by William M. Waid)

person is suspected of a crime and failure to be judged truthful on the polygraph test can have such serious consequences as dismissal by his employer, indictment, trial, and perhaps prison.

The polygraph examiner reminds the suspect of his right not to take the test, and has him sign a consent form testifying that he takes the test “voluntarily.” In a properly conducted test the examiner presents the entire process as an opportunity for the suspect to demonstrate his truthfulness, and thus his innocence (Reid and Inbau 1977). The examiner portrays himself as helping the subject prove his truthfulness, and the polygraph as a scientific instrument that will record the subject’s physiological reactions, implying that deception would be pointless because the record would reveal it.

Usually the examiner carefully discusses the questions he will ask with the subject. The test contains a number of irrelevant items such as, “Are you in the United States?” Also included are certain control questions -- for instance, “Besides what you told me about, have you ever stolen anything?” --designed to elicit an emotional response. Finally, there are three or four questions relating to the specific purpose of the test.

Only after the wording of the questions has been discussed with and accepted by the subject does the actual examination begin. The typical commercial polygraph records respiration, a cardiovascular measure, and the electrodermal response (EDR). Partly because the inflated blood pressure cuff used for the cardiovascular measure soon becomes uncomfortable, the examination itself is limited to about 12 questions, and takes 3 - 4 minutes. The examiner should wait 20 - 25 seconds between questions to allow the three measures to return to a fairly stable baseline, so that the response to one question does not affect the next response. During the actual test the subject is usually seated to that he cannot see the examiner or the record on the machine, as shown in Figure 1. Typical records from a polygraph test are illustrated in Figure 2.

In actual tests in the field, deception is frequently diagnosed merely by inspection of the records, not by quantitatively comparing the measures or by analyzing each measure separately. Laboratory and recent field studies have consistently found that the EDR is the most accurate of the physiological measures (Orne et al. 1972). Most of the findings discussed below will be from the EDR records, with cardiovascular and respiratory data presented only when they make some point not otherwise obvious.

Before a standard test, the examiner administers a “stimulation” test. He instructs the subject to choose one of six to eight cards, look at it, and conceal it on his person without the examiner seeing it. A polygraph test is then conducted in which the subject is asked in turn if he has each one of the cards; the subject is told to answer no to each question. The ostensible purpose of the test is to assess the individual’s typical response to deception, but the actual purpose is to demonstrate to the subject that the polygraph can indeed distinguish between truthful and deceptive answers. Such a demonstration is intended to increase the subject’s belief in the accuracy of the test, and subsequently to enhance the physiological responses of the deceptive subject and to reduce the fear of false detections among truthful subjects. As we shall discuss in more detail below, the subject’s belief in the accuracy of the polygraph does indeed affect his physiological responses. For this reason, examiners frequently arrange the stimulation test so that they know which card the subject has and can guarantee the accuracy of the test.

Although the stimulation test is seen by the field examiner as a demonstration, it is also the perfect lie, or more precisely “guilty-knowledge,” detection test, since its relevant questions and control questions are perfectly

404 American Scientist, Volume 70

matched except for deception. In other words, if the subject did not know which item was the relevant one, he would respond about equally to all the items. If some item of information, such as color of the stolen envelope in Figure 2, is known only to the guilty person among the subjects, the guilty-knowledge technique may be used. This procedure involves a series of questions asked in an order previously established with the subject; the crucial item is typically placed third or fourth. In addition to a larger response to this item on the part of the guilty subject, there may be a gradual increase in the amount of tension elicited by each item as the crucial one is approached, and then a sudden relaxation once it is passed.

The crucial feature of the guilty-knowledge test is that an innocent subject does not know which item is the relevant one and thus should not respond to it more than to the control items, whereas in the typical field test it is obvious which questions are relevant. Reid and Inbau (1977) provide many examples of ingenious uses of the guilty-knowledge test. Laboratory investigations, which -- until recently -- generally used a variation of this procedure, are therefore of practical as well as theoretical significance.

The effectiveness of the guilty-knowledge test has been proved in many laboratory demonstrations, and is

Figure 2. Of the three measures recorded in a typical polygraph test, the electrodermal response, the middle tracing in both tests above, is the most accurate. The other measures are respiration (top tracing) and relative blood pressure (bottom tracing). In these records from an actual theft case, the subject’s deception was indicated only by the peak of tension shown in the electrodermal response measure. The case involved the disappearance of a blue envelope containing a considerable sum of money. The subject, who claimed not to have seen the envelope and not to know its color, was asked (top) if the missing envelope was brown (1), red (2), blue (3), yellow (4), or gray (5). The same questions were then repeated (bottom). Observe the very pronounced peak of tension in the electrodermal response tracing each time question 3 was asked. Confronted with these polygraph records, the subject admitted taking the money and returned it in its blue envelope (After Reid and Inbau 1977).

supported by the sparse field of evidence regarding the technique. In contrast, when the control-question test normally used in the field is administered by police polygraph examiners and scored by inspection, rather than quantitatively, it appears to identify correctly about 75% of deceptive subjects, but at the expense of classifying 49% of truthful subjects as deceptive (Horvath 1977); the chance level of accuracy would be 50%. Similarly, a study by Barland and Raskin (1973) of field polygraph tests in criminal investigations found that 98% of the deceptive subjects were correctly identified, but 55% of the innocent subjects were erroneously identified as deceptive. To improve the accuracy of present field procedures, research is needed on the conditions under which the physiological record alone can be used to distinguish between deception and truth. The studies discussed below attempt to identify the factors that affect the physiological record.

What evokes the physiological response?

In field polygraph testing, the actual guilt or innocence of subjects does not become established for a long time -- if at all. The outcome of the case may in fact be affected by the polygraph results, since subjects who appear truthful may not be investigated as thoroughly as those who appear deceptive (Orne et al. 1972). A further problem is that the majority of cases are not solved. Consequently, investigators have turned to the experimental laboratory to study what evokes the physiological response accompanying deception.

The type of deception attempted by experimental subjects varies. In some studies, subjects commit a mock crime. In others, subjects memorize certain information and attempt to convince the examiner that they have not done so or lie about their personal identity and history. In still others, a procedure similar to the field stimulation test is used. Subjects are usually told, “This is a study in lie detection. Only very stable and emotionally mature individuals are able to fool the lie detector. We want you to try your best to fool the machine. Good luck!”

As in applied social research in general, experimental results permit more reliable analysis of the variables being studied, but may not be completely applicable to nonlaboratory settings. For example, the experimental polygraph subject cannot be expected to suffer the same degree of apprehension as the field subject. However, this difference is not an obstacle to experimental research since deception is detectable under laboratory conditions. As Davis (1961) noted, lying even about rather trivial matters, according to instructions from an experimenter, yields high rates of detection. Davis interpreted this to mean that the detectability is a general phenomenon that should also occur in situations quite different from criminal investigations, such as counterintelligence.

One basic question about detectability has been whether the physiological response is caused by the act of deception itself or by some other factor. In a study addressing this question, Gustafson and Orne (1965a), using a procedure based on the field stimulation test, instructed one group of students to say no after each question. A second group was also instructed to attempt

405 1982 July -- August

to appear innocent, but was to remain silent. In both groups deception was detected significantly above chance levels, but more deception was identified when the subjects responded verbally. Davis (1961) reported similar results, as did Kugelmass and colleagues (1967), whose subjects were instructed to answer yes to every question.

A recent study by Dawson (1980) also indicates that the physiological response is a result of the attempt to deceive, rather than the verbal answer per se. Dawson’s subjects were asked a series of questions twice, answering immediately one time but waiting eight seconds before answering the other time. Significant discrimination between deceptive and nondeceptive subjects was achieved in the immediate answer test, and with the physiological responses to the questions, but not to the answers, in the delayed answer test. The EDRs in the immediate answer test were the largest; in the delayed answer test the EDRs to the questions were about 60 - 80% the size of those in the immediate answer test, and for the answers only, the size was considerably smaller.

It might also be hypothesized that the physiological response is caused by the content of the relevant questions, independent of the attempt to deceive. Gustafson and Orne (1963) exposed two groups of subjects to a stimulus, and in subsequent polygraph tests in which the subjects remained silent, told one group to try to deceive the examiner about their exposure to the stimulus, while deception was not mentioned to a second group. Significant identification of the relevant stimulus was obtained with the group motivated to deceive, but not with the unmotivated group. These results indicate that, at least with relatively unimportant items of information, it is the effort to deceive, rather than the content of the question, that results in the detectable physiological response.

Nonetheless, lie detector tests usually do involve inherently arousing subject matter, and Thackray and Orne (1968) found that the content of the questions does affect detectability. The subjects in this study were told to disguise their identity, and not to admit to recognizing any of the words they had previously memorized. Deception about first and last names and date of birth was significantly more detectable than was deception about the memorized words. Similar results were recently presented by Stern and colleagues (1981).

These findings support Davis’s (1961) theory that the content of the relevant questions usually has arousal properties that produce a detectable physiological response independently of an effort to deceive, or fear of consequences. A person’s name is a stimulus to which he is thoroughly conditioned to respond, and deception about which is therefore relatively detectable. By analogy, although any question about a crime would evoke arousal, the more serious the crime, the greater the arousal when questioned about it (Horvath 1977).

Physiological reaction, and thus detectability, may be increased in field polygraph tests because the subject is well aware that his physiological responses are being monitored. If we could obtain the physiological responses of an individual without his awareness, how effectively could we detect deception? A preliminary investigation of this problem involved administering polygraph tests to subjects who were told that the recording equipment, located in an adjacent room, was turned off (Orne et al. 1972). The subjects were asked a series of questions and were requested to answer no to each question, ostensibly to prepare them for an official test which followed. The difference in detection rates between the first and second tests, while in favor of the second, only approached statistical significance (p ‹ 0.10).

No systematic effort was made to establish whether or not the subjects in this study believed that their responses were not recorded in the first test, but Thackray and Orne (1968) found that detectability did not vary significantly from an official test when subjects were (mistakenly) convinced that their responses were not being recorded. In this study, after several electrodes were attached to the subject, the examiner explained that some of the sensors would have to stabilize over a period of time before the test could take place. The subject and the examiner then discussed the questions, supposedly in order to avoid confusion and surprise during the test, but actually so that the subject’s EDRs could be recorded by a polygraph in another room. Once the questions had been discussed, the official examination took place. When the subjects were later asked if they believed a recording was made during the discussion period, only one subject out of twenty-eight expressed such a belief.

Waid and colleagues (1979a) also reported that, whether unaware or aware that their responses were being recorded, deceptive subjects tended to give significantly larger EDRs to the relevant question, while innocent subjects did not. However, the deceptive individuals were more detectable when they knew they were being tested.

There was no plausible way to reverse the order of the two tests in any of the studies, which means that the tests may not be exactly comparable. Nonetheless, the difference in detectability between the tests in each study is so slight that reversing the order should not significantly affect the results. The findings suggest that the arousing nature of the question together with the attempt to deceive may be sufficient to evoke a detectable response. The response may be enhanced, however, by an awareness that one’s responses are being monitored.

Detectability might be further enhanced by informing the subject about the effectiveness of his attempted deception. Feedback, or knowledge of results, has pervasive effects in a broad range of psychological processes. The field stimulation test, described above, is a vivid form of feedback, but what effect does it actually have?

To investigate this question, Gustafson and Orne (1965b) told subjects the outcome of one test before conducting the second. In the first test, subjects selected one of several cards, mistakenly believing that the examiner did not know which card they had chosen. They were then instructed to reply no when asked if they had each of the cards. After the first test, regardless of the subjects’ actual physiological responses, the experimenter told half the subjects what their card had been -- thus indicating that they had been detected. The experimenter named the wrong card to the other subjects, thereby leading them to believe that their deception had gone undetected. Each subject was then tested again with a new card; the results are presented in Table 1.

406 American Scientist, Volume 70

Table 1. The effect of feedback on detectability

Subjects who were told that they had been detected produced augmented physiological responses the next time that they were required to lie, making their detection much easier. Exactly the opposite result occurred with subjects who thought they had deceived the polygraph. If a subject believes that he has successfully deceived the polygraph once, he becomes less aroused when he lies subsequently. However, if a subject is led to believe that he has responded emotionally -- i.e., physiologically -- to a particular stimulus, that stimulus become more arousing to him. Bradley and Janisse (1981) recently reported that this finding is also applicable to control-question polygraph tests. The more often the subjects in this study believed they had been detected on the stimulation test, the more likely they were to be correctly identified as deceptive or truthful on subsequent control-questions tests.

These initial findings about the physiological detection of deception suggest that there are several behavioral and cognitive components of the interaction between examiner and subject, each of which independently produces a detectable response: the act of deception, the effort to deceive, and any inherently arousing content of the questions. Therefore, the polygraph test has a number of different psychological components that elicit a physiological response, which may be further enhanced by the use of feedback in the form of the stimulation test.

Decreasing detectability

In addition to merely attempting to deceive, a guilty individual faced with a polygraph test may make explicit preparations to cope with the stress of the test and escape detection. Trying to suppress physiological reactions inevitably fails, nor does consciously producing augmented responses to control items -- for instance, by thinking emotional thoughts, taking deep breaths, or clenching the teeth or sphincter -- appear to be successful (Lykken 1960). However, unobtrusive mental countermeasures and some of the tools available for handling day-to-day stress and anxiety, such as tranquilizers, may reduce physiological responses to the polygraph test generally, and to relevant questions specifically.

Research on the psychophysiology of attention and memory suggests that detectability may be affected by how a subject mentally processes the test questions. For example, Corteen (1969) reported that the words that subjects were later able to remember had produced significantly larger EDRs when presented than words the subjects later forgot. Similar results have been reported by Maltzman and co-workers (1966), Sampson (1969), and McLean (1969). Presumably, the more intensely a subject focuses on a stimulus, the greater the EDR and the more likely the stimulus is to be remembered later. These studies of the role of arousal in human memory suggest the hypothesis that subjects who avoid fully processing the questions, perhaps by engaging in an unobtrusive but distracting mental activity during the test, such as counting backward by sevens, would show reduced EDRs and consequently would be less detectable.

We investigated this hypothesis by conducting a polygraph test with three kinds of questions: critical, control, and irrelevant (Waid et al. 1981b). As hypothesized, the mean EDRs to critical and control questions that were later recalled were significantly larger than EDRs to critical and control questions that were not recalled. Thus, whatever attentional or information-processing mechanisms are associated with later recall are also related to larger EDRs.

We next sought to determine whether this phenomenon would account for some of the differences among subjects in physiological detectability, That is, are subjects who recall many questions, particularly critical ones, detected more frequently than subjects who recall few questions?

We calculated the correlation of detectability with the recall of critical and control questions, and found that although the correlation was not significant among innocent subjects, the more a guilty subject tended to recall critical rather than control questions, the more frequently he was detected (r = 0.40, p ‹ 0.02). In addition, the more irrelevant questions a guilty subject later recalled, the more often he was detected (r = 0.48, p ‹ 0.01). Apparently, memory for these questions indicates both the intensity of attention or information processing that the subject devotes to all the stimuli of the test, and the consequent likelihood of responding physiologically and being detected.

The results are consistent with the following model: The more attention a subject pays to a given stimulus, the larger the EDR it evokes and the more likely it is to be recalled later. Guilty subjects who pay more attention to the critical than to the control stimuli have a greater physiological response to the critical stimuli and thus are more detectable. Finally, recall of the irrelevant questions appears to serve as an independent index of the subject’s general attentiveness, a factor that also correlates with the number of detectable responses. Thus, evasive mental countermeasures might decrease a subject’s detectability by reducing his attentiveness.

The use of tranquilizers is another possible way of reducing detectability. Professional polygraphers have asserted that drugs are rarely used as countermeasures and that if they were used, the drugs would produce readily discernible effects that would prompt the examiner to postpone the test. Despite such claims, polygraphers have had no objective way of knowing whether or how often drugs have been used as countermeasures, since no plasma or urine analysis for such drugs has ever been reported in the context of polygraph tests. It would be prudent to assume that many guilty parties could obtain tranquilizers, the most widely prescribed and abused of all drugs, prior to polygraph testing. Tran-

407 1982 July -- August

quilizers are even available in jail, where illicit traffic in such drugs thrives.

We studied the effects of meprobamate, a tranquilizer, on guilty-knowledge polygraph test results, and whether the examiner could identify tranquilized subjects on the basis of their behavior and appearance (Waid et al. 1981a). Guilty subjects were divided into three groups: the first group was given no medication; the second group was told they were being given a tranquilizer that would help them avoid detection, but received placebo; and the third group, also told that they were being given a tranquilizer, received 400 mg of meprobamate. An experimenter who did not know which subjects were guilty or which were tranquilized conducted the polygraph tests.

As Table 2 shows, most meprobamate subjects were mistakenly classified as truthful, nor did the examiner’s identification of tranquilized subjects approach significant accuracy. Although meprobamate may affect overt behavior, the examiner has little or no previous experience of the subject, and without being able to compare the subject’s behavior during the test with his normal state cannot reliably detect the drug’s effects (Waid et al. 1981a). These results challenge both the widely held view among polygraph examiners that tranquilizers might depress a subject’s EDRs uniformly but would not reduce the difference between responses that leads to detection, and the view that tranquilized subjects can be recognized. In fact, such drugs may pose a serious threat to the detection of deception by means of the polygraph.

We have also examined the effects of hypnosis, another possible countermeasure, on polygraph test results. Self-hypnosis somewhat reduced anxiety, reported by subjects, that was evoked by the general context of the polygraph test, but, unlike tranquilizers, it did not reduce detectability (Waid and Orne 1981). Corcoran and co-workers (1978) found that subjects who received training in hypnosis or biofeedback were significantly less detectable after the training than before, whereas a control group tested at the same times did not show a significant decline in detectability. Nonetheless, the change in the first two groups may be the result of a placebo effect. Although the control group had no contact with the laboratory in the four weeks between the two tests, the other two groups returned several times for training. Rovner and colleagues (1979) have reported that subjects who received extensive information about the nature of “lie detection” and who were given two practice tests were later less detectable than a control group. The effects found by Corcoran and colleagues, therefore, may reflect the subjects’ amount of experience with the polygraph setting. It is particularly unlikely that detectability was affected by hypnosis per se, since the subjects were not selected for their susceptibility to hypnosis.

Individual differences

Detectability may be influenced by differences in personality; one dimension likely to have an effect is socialization. Poorly socialized individuals tend to be impulsive, lack restraint, and to have a history of conflict with other people. This behavior patterns suggests that deception is not unusual for such indi-

Table 2. The effect of a tranquilizer on detectability

viduals, and that they might be less aroused while attempting deception and consequently less detectable.

Waid and co-workers (1979b) found that deceptive subjects who were undetectable were significantly less socialized, as measured by the socialization scale of the California Psychological Inventory (Gough 1964), than those whose deception was detectable. Among innocent subjects, the highly socialized were more responsive electrodermally throughout the test, which caused some of them to be misclassified as deceptive. The difference in responsiveness does not seem to be a result of being tested; even when poorly socialized subjects were unaware that that their responses were being recorded, they were less detectable (Waid et al. 1979a).

Another personality variable relevant to dectability is electrodermal lability, or the frequency with which EDRs occur, either spontaneously or in reaction to a series of simple stimuli such as tones (Crider and Lunn 1971). Although in situations involving stress of some sort the frequency of response may be increased, individuals tend to maintain their relative degree of lability (see, for example, O’Gorman and Horneman 1979).

Since the EDR is the most accurate measure in the detection of deception, electrodermal lability might affect the accuracy of the polygraph test. A guilty person who usually gives few EDRs spontaneously would seem less likely to respond detectably during the test. Indeed, the less a labile the subject, whether innocent or guilty, the fewer questions on which he was classified as deceptive (Waid and Orne 1981).

Few other personality variables have been examined in the context of the polygraph test. McCarron (1973) found the subjects scoring high on a depression scale were less responsive physiologically on a polygraph test; their detectability per se was not reported, however. Ingersoll (1977 diss.) and Giesen and Rollison (1980) have reported results on socialization consistent with the data discussed above, using different measures of antisocial tendencies. Although Raskin and Hare (1978) found no effects on detectability of either clinical,

408 American Scientist, Volume 70

diagnoses of sociopathy or of low scores on Gough’s (1964) socialization scale in a prison sample, Lykken (1978) has discussed a number of problems with that study that may explain the lack of effects.

Social variables

By its very nature, lying is a social phenomenon. Ironically, the inherently social variables influencing the physiological changes accompanying deception have been the subject of even less explicit research than personality variables. Obvious factors such as the relative social status, sex, age, or ethnic origin of the polygraph examiner and subject might well affect responses, particularly in the case of innocent subjects.

Studies in Israel (Kugelmass and Lieblich 1968), the United States (Sternbach and Tursky 1965), and Japan (Lazarus 1966) suggest that there are ethnic differences in physiological responses to stress. Other research suggests that a subject might be more responsive physiologically if the examiner is from a different ethnic background (see, for example, Kugelmass and Lieblich 1968). Such increased responsivity might enhance the likelihood of detection if the subject is deceptive, but it might also increase the chance of falsely identifying an innocent subject as guilty (Waid and Orne 1980).

The ethnic groups we were able to compare, based on surnames of native-born American college students, were English, German, Irish, Italian, Jewish, and Scottish (Waid and Orne 1981). When subjects from these groups were given a polygraph examination by a former police captain of Irish heritage, the Irish subjects gave significantly smaller EDRs than did the other subjects, and also had the lowest number of detections of any group.

Such results might be caused by ethnic prejudices but, particularly with the exclusively Western European origins of the ethnic groups studied here, are more likely to be based on subtle differences in behavior among the groups. Polygraph subjects might respond, probably unconsciously, to unfamiliar gestures, tones of voice, and other cultural mannerisms.

Finally, another social factor that several studies (Gerard and Rabbie 1961; Stotland et al. 1971) prompted us to examine is family configuration -- i.e., how many siblings the subject has, and whether they are older or younger than the subject. Archival studies indicate that individuals from large families may be more likely to turn to delinquency (Galle et al. 1972; Rahav 1980), and thus might be less aroused by deception and less detectable; experimental studies have revealed that first-born children show greater tendencies toward dependency and conformity than do later born siblings (Warren 1966), which suggests that firstborns might be more aroused by deception.

We found that of 15 guilty subjects, the 7 who were firstborns were detectable on 4.6 questions, whereas the other 8 subjects were detectable on only 2.9 questions (Waid and Orne 1981). Similarly, the 6 subjects from families with fewer than three children were detected on a mean of 4.7 questions, while those from larger families were detected on only 3.0 questions. (The single only child in the sample was detected on 6 questions.) These variables had no effect on the responses of the 15 innocent subjects. Neither variable was significantly related to age, socialization, electrodermal lability, or to other measures previously found to affect detectability. Although family size is related to delinquency in large archival samples, the personality processes involved may not be measured on the socialization scale.

Our findings illustrate that the detection of deception tends to be affected by a variety of social and psychological influences. As is the case with traditional psychological techniques, the problems involved in the detection of deception are endemic and the physiological techniques employed are somewhat effective, though imperfect. Thus, their use is likely to continue. It is hoped, however, that findings such as those reported here will lead to a more realistic understanding of the complex processes by which physiological responses may serve to identify deception in different types of individuals and under varying conditions.

References

Abrams, S. 1973. The polygraph in a psychiatric setting. Am. J. Psychiatry 130:94-97.

Barland, G.H., and D.C. Raskin. 1973. Detection of deception. In Electrodermal Activity in Psychological Research, ed. W.F. Prokasy and D.C. Raskin. Academic Press.

Bradley, M.T., and M.P. Janisse. 1981. Accuracy demonstrations, threat, and the detection of deception: Cardiovascular, electrodermal, and pupillary measures. Psychophysiol. 18:307-15.

Corcoran, J.F.T., M.D. Lewis, and R.B. Garver. 1978. Biofeedback-conditioned galvanic skin response and hypnotic suppression of arousal: A pilot study of their relation to deception. J. Forensic Sci. 23:155-62.

Corteen, R.S. 1969. Skin conductance changes and word recall. Brit. J. Psychol. 60:81-84.

Crider, A., and R. Lunn. 1971. Electrodermal lability as a personality dimension. J. Exp. Res. in Personality 5:145-50.

Davis, R.C. 1961. Physiological responses as a means of evaluating information. In The Manipulation of Human Behavior, ed. A.D. Biderman and H. Zimmer. Wiley.

Dawson, M.E. 1980. Physiological detection of deception: Measurement of responses to questions and answers during countermeasure maneuvers. Psychophysiol. 17:8-17.

DePaulo, B.M., and R. Rosenthal. 1979. Telling lies. J. Personality and Social Psychol. 37:1713-22.

Ekman, P., and W.V. Friesen. 1974. Detecting deception from the body or face. J. Personality and Social Psychol. 29:288-98.

Galle, O.R., W.R. Gove, and J.M. McPherson. 1972. Population density and pathology: What are the relations for man? Science 176:23-30.

Gerard, H.B., and J.M. Rabbie. 1961. Fear and social comparison. J. Ab. and Soc. Psychol. 62:586-92.

Giesen, M., and M.A. Rollison. 1980. Guilty knowledge versus innocent associations: Effects of trait anxiety and stimulus context on skin conductance. J. Res. in Personality 14:1-11.

Gough, H.G. 1964. Manual for the California Psychological Inventory. Palo Alto: Consulting Psychologists Press.

Gustafson, L.A., and M.T. Orne. 1963. Effects of heightened motivation on the detection of deception. J. App. Psychol. 47:408-11.

_______. 1965a. The effects of verbal responses on the laboratory detection of deception. Psychophysiol. 2:10-13.

_______. 1965b. The effects of perceived role and role success on the detection of deception. J. App. Psychol. 49:412-17.

Horvath, F.S. 1977. The effect of selected variables on interpretation of polygraph records. J. App. Psychol. 62:127-36.

Ingersoll, B.D. Detection of deception in primary psychopaths. 1977 diss., Pennsylvania State Univ.

Jung, C.G. 1906. Die psychologische diagnose des tatbestandes. Juristischpsychiatrische Grezfragen 4:1-61.

Kugelmass, S., and I. Lieblich. 1968. The relation between ethnic origin and GSR reactivity in psychophysiological detection. J. App. Psychol. 52:158-62.

409 1982 July -- August

Kugelmass, S., I. Lieblich, and Z. Bergman. 1967. The role of “lying” in psychophysiological detection. Psychophysiol. 3:312-15.

Lazarus, R.S. 1966. Psychological Stress and the Coping Process. McGraw-Hill.

Lykken, D.T. 1960. The validity of the guilty knowledge technique: The effects of faking. J. App. Psychol. 44:258-62.

_______. 1974. Psychology and the lie detector industry. Am. Psychologist 29:725-39.

_______. 1978. The psychopath and the lie detector. Psychophysiol. 15:137-42.

_______. 1979. The detection of deception. Psycholog. Bull. 86:47-53.

_______. 1981. A Tremor in the Blood: Uses and Abuses of the Lie Detector. McGraw-Hill.

McCarron, L.T. 1973. Psychophysiological discriminants of reactive depression. Psychophysiol. 10:223-30.

McLean, P.D. 1969. Induced arousal and time of recall as determinants of paired associate recall. Brit. J. Psychol. 60:57-62.

Maltzman, I., W. Kantor, and B. Langdon. 1966. Immediate and delayed retention, arousal, and the orienting and defensive reflexes. Psychonomic Sci. 6:445-46.

Marston, W.M. 1917. Systolic blood pressure symptoms of deception. J. Exp. Psychol. 2:117-63.

O’Gorman, J.G., and C. Horneman. 1979. Consistency of individual differences in non-specific electrodermal activity. Biol. Psychol. 9:13-22.

Orne, M.T., R.I. Thackray, and D.A. Paskewitz. 1972. On the detection of deception: A model for the study of physiological effects of psychological stimuli. In Handbook of Psychophysiology, ed. N.S. Greenfield and R.A. Sternbach. Rinehart & Winston.

Podlesny, J.A., and D.C. Raskin. 1977. Physiological measures and the detection of deception. Psycholog. Bull. 84:782-99.

Rahav, G. 1980. Birth order and delinquency. Brit. J. Criminol. 20:385-95.

Raskin, D.C., and R. D. Hare. 1978. Psychopathy and detection in a prison population. Psychophysiol. 15:126-36.

Reid, J.E., and F.E. Inbau. 1977. Truth and Deception: The Polygraph (“Lie Detector”) Technique. Baltimore:Williams & Wilkins.

Rovner, L.I., D.C. Raskin, and J.C. Kircher. 1979. Effects of information and practice on the detection of deception. Abstracted from Psychophysiol. 16:197.

Sampson, J.R. 1969. Further study of encoding and arousal factors in free recall of verbal and visual material. Psychonomic Sci. 16:221-22.

Stern, R.M., J.P. Breen, T. Watanabe, and B.S. Perry. 1981. Effect of feedback of physiological information on responses to innocent associations and guilty knowledge. J. App. Psychol. 66:677-81.

Sternbach, R.A., and B. Tursky. 1965. Ethnic differences among housewives in psychophysical and skin potential responses to electric shock. Psychophysiol. 1:241-46.

Stotland, E., S.E. Sherman, and K.G. Shaver. 1971. Empathy and Birth Order: Some Experimental Explorations. Univ. Nebraska Press.

Thackray, R.I., and M.T. Orne. 1968. Effects of type of stimulus employed and the level of subject awareness on the detection of deception. J. App. Psychol. 52:234-39.

US Senate. 1978. Polygraph control and civil liberties protection act. Hearings before the Subcommittee on the Constitution of the Committee on the Judiciary.

Waid, W.M., E.C. Orne, M. R. Cook, and M.T. Orne. 1981a. Meprobamate reduces accuracy of physiological detection of deception. Science 212:71-73.

Waid, W.M., E.C. Orne, and M.T. Orne. 1981b. Selective memory for social information, alertness, and physiological arousal in the detection of deception. J. App. Psychol. 66:224-32.

Waid, W.M., and M.T. Orne. 1980. Individual differences in electrodermal lability and the detection of information and deception. J. App. Psychol. 65:1-8.

_______. 1981. Cognitive, social, and personality processes in the physiological detection of deception. In Advances in Experimental Social Psychology, vol. 14, ed. L. Berkowitz. Academic Press.

Waid, W.M., M.T. Orne, and S.K. Wilson. 1979a. Socialization, awareness, and the electrodermal response to deception and self-disclosure. J. Abnormal Psychol. 16:15-22.

_______. 1979b. Effects of level of socialization on electrodermal detection of deception. Psychophysiol. 16:15-22.

Warren, J.R. 1966. Birth order and social behavior. Psycholog. Bull. 65:38-49.