Home Page > NAS Polygraph Review



Lack of Scientific Evidence for the Reliability of Polygraphs

Polygraphs have long been controversial as a tool for determining the veracity of persons subjected to them. As recently as last year, the U.S. Supreme Court, in a decision upholding the absolute ban on the use of polygraph tests in court-martial proceedings said, "There is simply no consensus that polygraph evidence is reliable. To this day, the scientific community remains extremely polarized about the reliability of polygraph techniques."1 The Court went on to say that, "Although the degree of reliability of polygraph evidence may depend upon a variety of identifiable factors, there is simply no way to know in a particular case whether a polygraph examiner's conclusion is accurate, because certain doubts and uncertainties plague even the best polygraph exams."2

The skepticism of the scientific community to polygraph testing is evidenced by a peer-reviewed paper published in 1997 in the Journal of Applied Psychology.3 This paper summarized the results of a survey of two groups of scientists from relevant disciplines: members of the Society for Psychophysiological Research (SPR) and Fellows of the American Psychological Association's Division 1 (General Psychology). The SPR had been polled by polygraph advocates in the past and the responses to that poll had been interpreted to suggest that the scientific community accepted polygraphy as a legitimate investigatory tool.4

In the more detailed survey, only 36 percent of the SPR members agreed that the polygraphic lie detection technique alluded to in DOE's plans, the Control Question Test, was "based on scientifically sound psychological principles or theory."5 Less than 30 percent of SPR members said that they would advocate that courts admit into evidence the outcome of Control Question Tests.6 Almost all SPR respondents to the survey indicated their judgment that the Control Question Test can be "beaten" by augmenting one's responses to the control questions.7 Fellows of the American Psychological Association (APA) were even more skeptical of polygraphy. Only 10 percent of APA respondents thought polygraphy to be an objective procedure, "relatively independent of differences among examiners in skill and subjective judgment."8

The skepticism of the scientific community is consistent with the lack of scientific evidence for polygraphs as a screening technique. In 1983, the Congressional Office of Technology Assessment (OTA) found that "the available research evidence does not establish the scientific validity of the polygraph test for personnel security screening."9 This OTA report was reviewed, prior to its release, by a balanced panel of scientific experts in the area. The OTA concluded that "while there is some evidence for the validity of polygraph testing as an adjunct to criminal investigations, there is very little research or scientific evidence to establish polygraph test validity in screening situations, whether they be preemployment, preclearance, periodic or aperiodic, random, or 'dragnet.'"10 The intervening 16 years has not produced new research on the validity of the polygraph that would cast doubt on the OTA conclusion. Even the world-wide web pages of the American Polygraph Association, which promotes the use of polygraphs, admit that "there has been only a limited number of research projects on the accuracy of polygraph testing in the pre-employment context...."11 To my knowledge, there are no peer-reviewed papers in the scientific literature that describe these "research projects." This lack of specific studies on polygraph testing as a screening device is important because, as the OTA noted in its report, "The accuracy [of polygraphs] obtained in one situation or research study may not generalize to different situations or to different types of persons being tested."12

In addition to the lack of scientific evidence for the validity of polygraphs in counterintelligence screening, the public track record of the technique in detecting espionage does not inspire confidence. According to published reports, Aldrich H. Ames, convicted of spying for the Soviet Union, passed numerous polygraph tests. So did Larry Wu-Tai Chin, a career CIA analyst convicted in 1986 for spying for China. In the case of Dr. Wen-Ho Lee, it has been reported that he both passed and failed a polygraph test.

In light of this background, the statement in the Department's Federal Register notice that, "DOE is aware of no scientific studies that establish that polygraph examination results are unreliable for use as an investigative tool, as DOE has today proposed to use them," seems both inaccurate and inappropriate as a basis for rule making. If DOE is to engage in reasoned rule making on instituting polygraph requirements, it has an affirmative duty to produce scientific studies and impartial evidence that counterintelligence screening by polygraph is effective. The courts, in interpreting the Administrative Procedure Act, have ruled that agency failure to provide public notice of the data on which the agency chooses to rely in a rule making is a "critical defect in the decisionmaking process."13 If DOE is aware of a body of scientific studies to support its proposal, it should make it available so that the public can scrutinize the methodology of such studies and raise legitimate problems in their comments to the agency. It is not appropriate or reasonable, in this rule making, to leave the public ignorant of DOE's reasons for believing that its proposed rule will be effective or, worse, to take the position that it is up to the public to prove false DOE's seemingly unsupported assertions.

Problems in Defining Who Is Covered by the Rule

Aside from the fundamental question of whether DOE's use of polygraphs for counterintelligence screening is reasonable, the provisions of the proposed rule specifying categories of individuals to be screened raise additional questions about the proposal. The proposed rule defines a number of categories of personnel who would be subject to the polygraph requirement. Included in the list are positions in the Personnel Assurance Program (PAP) and the Personnel Security Assurance Program (PSAP). These programs are defined in 10 C.F.R. Part 711 and subpart B of 10 C.F.R. Part 710. It is very difficult to see why these categories of personnel should be swept up in counterintelligence polygraphs.

In addition to the proposed rule including categories of personnel that seem unlikely to possess detailed nuclear design information, the rule does not address which individuals in these various categories will be subjected to polygraph screening. The text of the proposed rule states that all individuals in such positions will be "eligible" for polygraph screening. The preamble of the rule, though, states that some subsequent "internal process" will result in a determination of the positions for which polygraph screening will be required, and section 709.4(c) of the proposed rule states that the actual criteria for identifying specific positions that warrant polygraph examination will be determined subsequent to the final promulgation of the rule, and not as part of the rule making process. This raises an additional issue regarding the adequacy of this proposed rule. The public, which is being asked to comment on the rule, is unable to determine whether this is a rule requiring polygraph examination of a mere handful of DOE contractor employees or for thousands of such employees. No description of the criteria to be used in the proposed "internal process" is provided, nor does DOE solicit comment on this key issue. Yet, this issue of the scope of the proposed polygraph screening program would appear to be central in the rule making.

Inadequate Protection of Rights to Counsel and Privacy

The Department's description of the steps of the polygraph examination raise other troubling issues about fairness and preservation of individual privacy. As defined in the proposed rule, a "polygraph examination" encompasses "all activities that take place between a polygraph examiner and examinee..." including "the pretest interview, the use of the polygraph instrument..., the test data analysis phase, and the post-test phase."15 The pre-test interview and the post-test phase apparently include a process in which the person being examined is encouraged to reveal any potential infractions or violations of DOE security regulations, as well as other information that may be of a private nature. Yet, while the rule states that the polygraph examiner must "advise the individual of the individual's privilege against self-incrimination"16 before beginning a polygraph examination, the person being examined is not permitted to have legal counsel or other representative present to protect his or her legal rights.17 I do not believe that it is acceptable to put DOE contractor employees in a situation where they might have questions about their legal rights, but not have access to counsel.

In addition to problems the rule raises with respect to access to legal counsel, the rule envisions the use of audio and video recording devices to make a permanent record of the polygraph examination, including disclosures during pre-test and post-test interviews with persons being examined. The proposed rule states that "all polygraph examination records," including such recordings, will be maintained in a permanent system of records.18 While a video recording of a polygraph session might be useful as a check on inappropriate practices in an examination, I question the need for long-term retention of private information that is disclosed in a polygraph examination, but is not relevant to counterintelligence issues. I believe that the proposed rule should provide for the destruction of such non-essential information.

The Problem of Inconclusive and False Positive Results

A considerable amount of concern has been directed at the proposed use of counterintelligence polygraphs because of the likelihood that it will result in a significant number of "false positive" results. I share these concerns, which are very concisely set forth in the OTA report:

"OTA concluded that the mathematical chance of incorrect identification of innocent persons as deceptive (false positives) is highest when the polygraph is used for screening purposes. The reason is that, in screening situations there is usually only a small percentage of the group being screened that might be guilty. So, in a typical situation, there may be, perhaps, one person per 1,000 engaged in unauthorized activity. Therefore, even if one assumes that the polygraph is 99 percent accurate, the laws of probability indicate that one guilty person would be correctly identified as deceptive but 10 persons would be incorrectly identified (false positives)."19

It should be recognized that the published estimates of polygraph accuracy in the peer-reviewed literature are much lower than 99 percent. Further, the proposed DOE rule specifies that if a counterintelligence polygraph results in "unresolved issues" (a larger class of results than just those polygraphs judged to indicate deception), an in-depth interview, additional polygraph examination, and potentially a comprehensive investigation must be undertaken. Thus, DOE's proposed rule is likely to result in a high percentage of situations in which innocent persons are judged to require follow-on examination and investigation.

I recognize that DOE's policy will be that a variety of information, in addition to polygraph results, will be used to assess the eligibility of an individual with "unresolved issues" to obtain or retain access to highly classified information. However, given the demonstrated political sensitivity of the issue of espionage at DOE laboratories, it is hard for me to believe that any amount of subsequent investigation could clear a person who had been judged to "fail" a counterintelligence polygraph examination in the eyes of many in Congress. The person involved and DOE would be in the position of having to prove a negative to a skeptical press and a partisan Congress.

Further, scientific study of polygraphs suggest that the type of personnel employed at DOE laboratories may be more prone to registering false positive results on a polygraph test. The OTA report cited one study that found that the majority of false positives were among subjects with college degrees.20 Since a polygraph test directly measures sensory arousal, I would expect that persons who are especially conscientious or scrupulous about their answers to questions about their past activities (a class of persons in which I believe many top scientists might be found) would be particularly likely to register ambiguous results on a polygraph.

Effect on Hiring and Retention of Personnel

There can be no question that DOE's proposed rule on counterintelligence polygraph examinations is strongly opposed at the working level in the DOE nuclear weapons laboratories. Given the above concerns, I share this opposition. I believe that DOE would be ill-advised to conclude that this opposition is either superficial, transient, or reflective of "arrogance" by laboratory scientists.

The employees of the DOE laboratory system created our nuclear deterrent, a cornerstone of our national defense. Theirs is a strong record of accomplishment, of loyalty to the United States, and of intellectual integrity on which the country has depended for over 50 years. I believe that the opposition by DOE laboratory scientists to DOE's rule is a result of subjecting the rule to the same intellectual scrutiny that they are accustomed to bringing to their other national security duties.

The DOE defense-related laboratories are having an increasingly difficult job with the recruitment and retention of the best scientists and engineers. Work at the DOE laboratories does not carry the financial rewards of high-technology jobs in the private industry, and in my discussions with laboratory leaders over the past few year, [sic] I have been concerned about DOE's continuing ability to attract the best and brightest. These concerns have been echoed by the Commission on Maintaining United States Nuclear Weapons Expertise, chaired by Admiral Henry G. Chiles, Jr. USN (Retired), in its March 1999 report to Congress and the Secretary of Energy. These concerns lead me to believe that this new rule is ill-timed for two reasons.

  1. It may accelerate the retirement of the cadre of nuclear weapons designers who have actual experience in underground testing of weapons designs. These individuals represent a key source of expertise to be passed along to the next generation of weapons scientists and engineers. The next generation are being trained and will work on weapons design issues in the absence of being able to conduct actual tests of complete weapons. At the Los Alamos National Laboratory, designers with the bulk of experience with testing nuclear designs generally range in age from 56 to 61 years of age. By the time that the proposed rule is implemented, this cohort of designers will be over the age of 60, when early retirement does not involve a great loss of benefits. They may chose [sic] to do so, rather than have their continued employment subject to a procedure that they view as arbitrary and unscientific.
  2. It will pose a new impediment to recruitment and retention of the best young scientists and engineers. These individuals have many options in today's competetive technology marketplace--the Chiles Commission characterized the DOE as being in a "war for talent" with the private sector for these individuals. Competing employers will certainly not subject these individuals to polygraph screening, as this practice is forbidden in the private sector by the Employee Polygraph Protection Act of 1988. The DOE is, thus, instituting a new test for prospective employees that will put its laboratories at even a greater competetive disadvantage with the private sector. DOE's hope that its proposed rule "will be perceived as fair by most potential employees" is unlikely to be realized if these potential employees research the scientific literature on polygraphs prior to making their decision to accept employment.

Conclusions and Recommendations

I believe that the DOE, in this rule, has gone far beyond what I envision as being an appropriate use of polygraphs, which would be as a limited investigative tool in cases where other evidence suggests the possibility of espionage. DOE's proposed rule is flawed on several counts.

Inasmuch some of the above deficiencies cannot be remedied (particularly the lack of scientific evidence of the validity of this procedure), I would recommend that DOE should abandon its plans for polygraph screening of employees for counterintelligence purposes.

If the DOE decides, despite my recommendation, to proceed with this rule, I recommend that it reconstitute and reconvene the Chiles Commission to conduct a formal study of the rule's likely impact on the critical human resources needed to ensure the safety and reliability of the nuclear weapons stockpile. I would also recommend that the DOE seek review from the National Academy of Sciences on the weight of scientific evidence establishing the reliability of the types of polygraph screening it plans to implement. Finally, I recommend that DOE use the results of these two studies to repropose its rule, laying out both the scientific evidence for the validity of tests required by the rule and providing the criteria governing the selection of individuals for polygraph examinations, for public comment.

1United States v. Scheffer, 523 U.S. 303, 118 S. Ct. 1261, 140 L. Ed. 2d 413, 419 (1988)

2Id. at 421.

3W.G. Iacono and D.T. Lykken, "The Validity of the Lie Detector: Two Surveys of Scientific Opinion," 82(3) JOURNAL OF APPLIED PSYCHOLOGY, 426-33 (1997).

4Id. at 427.

5Id. at 430.


7Id. at 431.

8Id. at 432.

9Office of Technology Assessment, United States Congress, SCIENTIFIC VALIDITY OF POLYGRAPH TESTING: A RESEARCH REVIEW AND EVALUATION--A TECHNICAL MEMORANDUM. OTA-TM-H-15. 4 (Washington, D.C., November 1983) (hereafter OTA).

10Id. at 8.

11American Polygraph Association,<>.

12OTA, supra note 9, at 7.

13Portland Cement Assn. v. Ruckelshaus, 486 F.2d 375, 392 (D.C. Cir 1973).

1410 C.F.R. Part 711.1.

15Proposed Rule at Part 709.3.

16Id. at Part 709.24(e).

17Id. at Part 709.22.

18Id. at Part 709.26(b).

19OTA, supra note 9, at 5-6

20Id. at 85. Home Page > NAS Polygraph Review