"What is truth?” asked Pilate in the trial of Jesus. John 18:38. Today the polygraph really poses the same question. In the intervening nineteen hundred years, everything from trial by ordeal to jury trial has been used to try to establish truth.
This case requires us to determine whether polygraph evidence which heretofore has not been admissible at trial may be used to assist a judge in determining whether to grant a motion for a new trial. (For a description of how a polygraph and a polygraph test work, please see the appendix.)
The Michigan test presently applied to determine the admissibility of polygraph testimony is that which we enunciated in People v Becker, 300 Mich 562, 566; 2 NW2d 503 (1942), and repeated with approval in People v Davis, 343 Mich 348, 370; 72 NW2d 269 (1955), that there be:
"testimony offered which would indicate that there is at this time a general scientific recognition of such [polygraph] tests. Until it is established that reasonable certainty follows from such tests, it would be error to admit in evidence the result thereof.”
In Becker and in Davis we concluded that "the results of such tests have not, as yet, reached the stature of evidence admissible in a court of law”. 343 Mich 370.
*358This follows the traditional test used in other jurisdictions, which, as originally set forth in Frye v United States, 54 US App DC 46, 47; 293 F 1013, 1014 (1923), requires:
"[W]hile courts will go a long way in admitting expert testimony deduced from a well-recognized scientific principle or discovery, the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs.”
The Frye court, finding no such "standing and scientific recognition among physiological and psychological authorities” did not admit testimony based on the tests. 54 US App DC 47; 293 F 1014.
In this case defendant, who seeks to introduce into evidence the results of polygraph tests, has presented distinguished testimony as to the improved nature of the polygraph itself, the reliability of polygraph operators and the claimed accuracy of polygraph tests. While it is apparent that advances have been made in the state of the art, the fact of the matter is that all submitted testimony was given by polygraph operators, polygraph teachers or others connected with the use of polygraphs.
What defendant/appellant made a strong case of showing was that those whose business is the use or operation of polygraphs have confidence in the tests. There was no testimony by disinterested and impartial experts "in the particular field [physiological and psychological] in which it belongs” (Frye test), and thus no demonstration of "general scientific recognition of such [polygraph] tests” (Michigan rule).
*359So unless Michigan changes from its traditional Davis/Frye rule, polygraph tests are still not admissible in evidence. We have declined to make such a change for the following reasons:
(1) The Frye test remains the accepted standard for the admission of polygraph testimony in most jurisdictions.
(2) Serious questions still remain concerning the state of the art of the polygraph.
(3) Even if the Frye requirements were satisfactorily met, serious policy questions remain to be resolved.
However, the question of whether the polygraph may be used to assist the judge in determining whether to grant a motion for a new trial is another matter. Less formal evidentiary requirements are commonly found at the post-conviction motion, and matters such as affidavits which may not be admissible at trial may be used to assist the judge to determine whether the request for a new trial has merit. Polygraph tests permitted in this context would be merely used to buttress the credibility of new witnesses, the evidentiary value of whose testimony satisfies traditionally strict criteria for ordering a new trial. Thus, the polygraph examination would not itself be evidence, either at the post-conviction motion or at the trial itself, should one be granted.
On this basis, and within carefully drawn and defined limitations, we hold that a judge may use, in his discretion, polygraph tests and testimony offered by defendant only to help determine whether to grant a post-conviction motion for a new trial.
Remanded for further action pursuant to this opinion.
*360I — Facts
In 1969, defendant Joseph Barbara, Jr., was convicted by a jury of extortion from Delores Lazaros, wife of Peter Lazaros. Both Peter and Delores Lazaros testified at trial. Delores Lazaros testified that defendant Barbara came to her home while her husband was in prison, and raped her and extorted money from her by threatening the lives of her husband and son. She also said she told no one about this until her husband returned from prison because she feared for his safety. At that time, Mrs. Lazaros and her husband informed the authorities.
Defendant’s claim of appeal was rejected by the Court of Appeals. People v Barbara, 23 Mich App 540, 548; 179 NW2d 105 (1970). We denied leave to appeal. 383 Mich 803 (1970). The United States District Court granted defendant’s writ of habeas corpus, but was reversed by the Sixth Circuit Court of Appeals. Barbara v Johnson, 449 F2d 1235 (CA 6, 1971). Defendant is now before our Court appealing denial of his motion for a new trial.
The basis for defendant’s request for a new trial was his claim that Lazaros lied. To support this, defendant offered two new witnesses and evidence of polygraph tests passed by himself and one of two new witnesses.
Albert Spadafore, although not called to testify at the original trial, reported he had been ready to tell an untrue version supporting Lazaros’ story, under instructions from Lazaros. There was some question as to whether Spadafore had indeed told the "untrue version” in another unrelated trial involving Lazaros.
Theodore Metropoulos, a cousin of Lazaros, testi*361fled through an interpreter that Lazaros told him his wife was not raped. He said Lazaros admitted making up the story "because this was the only way to get out of the cursed jail or prison”. Metropoulos said he had successfully taken a polygraph examination. He also explained that he came forward with his story at this time because he had learned that defendant had been sentenced to a long prison term.
The trial court refused to accept the polygraph tests in evidence but gave counsel permission to make a special record concerning the polygraph examinations given to Metropoulos and defendant. Proceedings in a 1970 case involving Peter Lazaros in Oakland Circuit Court were made part of this special record. In that case, Lazaros’ defense counsel F. Lee Bailey undertook to lay a foundation for admissibility of the polygraph evidence and to qualify Lynn Marcy as a polygraph expert.
Marcy was also Barbara’s polygraph witness in the case at bar. Marcy presented the results of the tests made on Barbara and on Metropoulos. He concluded that Barbara was not guilty of either rape or extortion and testified as to the procedures used,1 indicating that accuracy of the polygraphic examination under circumstances as given is in excess of 99 percent. The same procedure was used on Metropoulos, with the addition of an interpreter, and Marcy concluded that Metropoulos also was telling the truth.
The court observed that Lazaros had sent letters indicating that he, too, had passed a polygraph examination.
The motion for a new trial was denied. The court held, first, that Michigan law prevents the *362admission into evidence of polygraph examinations, even in a post-conviction proceeding.
The judge also ruled that, in any event, the testimony did not constitute sufficient newly discovered evidence to warrant granting a new trial, first, because Spadafore’s testimony was never part of the original trial and further, since he admitted being prepared to testify falsely, the trial court could not now rely on his testimony. In addition, although Metropoulos’ testimony "clearly tended to discredit the testimony of Peter Lazaros, his own cousin”, it did not discredit that of the complainant, Delores Lazaros. Since the conviction was primarily based on Delores Lazaros’ testimony, and the alleged attack took place while her husband was in prison, the court was not satisfied that the purported new evidence would render a different result probable on retrial.
The Court of Appeals denied leave. We granted leave to appeal January 25, 1974. 391 Mich 761 (1974).
II — Standards For a New Trial
Before a court will grant a new trial because of newly discovered evidence,
"[I]t must be shown that the evidence itself, not merely its materiality, was newly-discovered; that it is not cumulative; that it is such as to render a different result probable on a retrial of the cause; and that the party could not with reasonable diligence have discovered and produced it at trial.” People v Clark, 363 Mich 643, 647; 110 NW2d 638 (1961).
Where such evidence, however, takes the form of witnesses’ recantation testimony, it has been tradi*363tionally regarded as suspect and untrustworthy. People v Smallwood, 306 Mich 49, 58; 10 NW2d 303 (1943); People v Jimmerson, 30 Mich App 147, 149; 186 NW2d 37 (1971); People v Bersine, 48 Mich App 295, 299; 210 NW2d 501 (1973); 58 Am Jur 2d, New Trial, § 175, p 391. Generally, too, where the new evidence is useful only to impeach a witness, it is deemed merely cumulative. Kube v Neuenfeldt, 353 Mich 74, 82; 90 NW2d 642 (1958).
However, the discovery that testimony introduced at trial was perjured may be grounds for ordering a new trial. 58 Am Jur 2d, New Trial, § 175, p 391. The problem for defendant becomes, then, how he or she goes about demonstrating that perjury was committed when the newly discovered evidence consists of a new witness or of an old witness’s revised story. What defendant suggests in the case at bar is that the polygraph may assist the judge to determine whether the newly-discovered evidence is worthy of belief.2 Specifically, if the jury would believe Metropoulos, then that would not only impeach the credibility of Lazaros, it would affect the credibility of Delores Lazaros as well.
We therefore review the question of whether, in a post-conviction motion for a new trial, polygraph evidence may be admissible to help the trial judge determine the credibility of witnesses offering to testify at this new trial. This is particularly significant when, as in the instant case, the only evidence that an offense was ever committed was largely based on the testimony of individuals *364whose credibility might be put into question by these new witnesses.
Ill — Michigan Rule of Non-Admissibility of Polygraph
Michigan decisions have uniformly held against admissibility of the polygraph. The basic rationale has been our determination that the polygraph technique had not received the degree of standardization or acceptance among scientists which would warrant admissibility.
"[T]estimony [must be] offered which would indicate that there is at this time a general scientific recognition of such tests. Until it is established that reasonable certainty follows from such tests, it would be error to admit in evidence the result thereof.” People v Davis, 343 Mich 348, 370; 72 NW2d 269 (1955), quoting People v Becker, 300 Mich 562, 566; 2 NW2d 503 (1942).
This approach is similar to the Frye requirement that "the thing from which the deduction is made must be sufficiently established to have gained general acceptance in the particular field in which it belongs”. 54 US App DC 46, 47; 293 F 1013, 1014. The policy of exclusion applies equally to civil cases, and test results may not be admitted in either civil or criminal cases pursuant to stipulation. Stone v Earp, 331 Mich 606, 611; 50 NW2d 172 (1951).
We have applied this rule to other types of evidence. For example, in determining the admissibility of drunkometer results we used the rule because "both [polygraph and Drunkometers] involve tests by scientific devices, the accuracy of which can scarcely be determined by a jury on the basis of complicated, scientific testimony concern*365ing the theory and operation of the devices in the face of a difference of scientific opinion as to their accuracy”. People v Morse, 325 Mich 270, 274; 38 NW2d 322 (1949). Compare, People v Kenney, 354 Mich 191, 196; 92 NW2d 335 (1958), where evidence of a speedwatch was admitted when defendant offered no evidence that the instrument was mechanically and scientifically inaccurate.
We have not required infallibility, but only that "reasonable certainty” follow from "such tests”. 300 Mich 562, 566.
Therefore, before the polygraph may be admitted into evidence in Michigan, there must be a showing by the proponent of the evidence that the state of the art satisfies the Davis/Frye requirements.3
Both Davis and Frye emphasize that it is the burden of the party seeking admissibility to demonstrate acceptability of the proposed technique. We must therefore examine the record in the case at bar to determine if a showing has been made that polygraph test results meet the standards of the Davis/Frye test.
IV — Defendant’s Case for Admissibility of the Polygraph
To establish the reliability of polygraph testimony and to meet the Davis/Frye tests for admissibility of polygraph findings, the defendant in this case introduced Lynn Marcy, a well-known private polygraph examiner and the special record made outside the presence of the jury in the case of People v Lazaros (Oakland Circuit Court Docket No. CR-6237) before then Circuit Judge James P. *366Churchill. In that case, attorney F. Lee Bailey sought to lay a foundation for the introduction into evidence of the results of a polygraph examination.
In that hearing, Bailey presented testimony of Lynn Marcy and five other polygraph experts. William J. Yankee was Dean of Academic Affairs at Delta College and a teacher at Keeler Polygraph Institute in Chicago. He received a master’s degree in psychology from Michigan State University as well as a Ph.D. in higher education. Leonard H. Haroldson was President of Keeler Polygraph Institute, Chicago. John E. Reid was a polygraph operator, attorney and co-author with Professor Fred E. Inbau of a standard book on polygraphy. Richard North and Eugene Dinkel were Michigan State Police Officers and polygraph operators. All six witnesses had polygraphy as a principal occupation.4
The record made by the defense in the Lazaros case was, as acknowledged by the trial judge, outstanding. Even a summary of some of the highlights of the testimony is lengthy, but necessary to a proper determination of the case at bar.
Marcy, who had been Bailey’s witness on the polygraph in other trials, discussed the theory and operation of the device and the techniques of administering a test. He also indicated that several states require that polygraph operators be licensed, but have varying standards, although all *367require a formal training course approved by the licensing authority. There was testimony that there were five accredited polygraph schools. The accrediting agency, Bailey explained, is the American Polygraph Association.
Marcy testified to being able to verify, by "guesstimate,” 30 percent of the over 9,000 examinations he reported administering over a 14-year period. Of these he reported that "better than 95 percent” were consistent with later developments.5 Less than 3 percent of his results were inconclusive. He also reported that of the mistakes he could recall, almost all were screening cases involving inventory pilferage where he cleared individuals who were implicated by later investigations.
Marcy also testified concerning whether various devices might be effective in "beating” the machine. He stated that hypnosis would not work if the examiner were qualified and "aware of the various facets of hypnosis and how it might be used against him”. He asserted that there was no way somebody "through their own act” could deceive an examiner into diagnosing as true that which is not. Mistakes were due, he said, "[o]nly through human frailty”. The cause of such errors could, he said, be detected by retracing steps through the charts.
The polygraph, he said, could distinguish between a deceptive and a nervous subject. Individuals who would lie about anything and everything also pose no problem. He also said there are no illnesses or diseases, even those that require drugs, that cause him a problem.
He reported he would be able to read charts *368made by another examiner "[i]f we were within the same ballpark on technique and systems”. He could form an opinion on the charts if he knew something about the circumstances at the time. However, he admitted he would have to rely on the original examiner to report what question was asked when.
Marcy testified it was quite common to be asked to give an opinion based on another operator’s chart, having in addition to the charts "full faith and confidence in the individual who is asking”. Only infrequently do two polygraphers reviewing the same charts come up with opposing results, he said.
He acknowledged too, that opinions on polygraph results were subjective, based not only on the tracing, but on the operator’s interpretation of his or her reactions during the pretest period when the subject is prepared, and on the post-test period when results are discussed with the subject.
There were instances, he reported, where, after running the tests and coming to a conclusion, he changed his mind after sitting down and talking to the subjects.6
Police polygraphs, he said, were generally easier to interpret than those made by individuals paying for and offering their own examinations, perhaps because the fear of detection is less in the latter situation.
He testified in a general way to the use of the polygraph, usually as a screening device, by business and industry. He also expressed the opinion that the device has great value in the initial phase of the investigation before the warrant is issued. *369He reported what he characterized as the police or law enforcement position, that admissibility in evidence would cause the polygraph to lose its utility as an investigative aid, because individuals would refuse to take the test since there was no longer the safeguard that it could not be used in court.
It was his opinion that the polygraph technique has developed to the point where it should be admitted in court in criminal cases where the polygraph shows that the defendant is not deceptive. It was also his opinion that the properly instructed jury would not be overwhelmed by the polygraph and gave, as an example, a 1960 New Jersey case where he had testified the defendant was deceptive, and the jury still acquitted him. He also placed polygraph evidence "on a plane of reliability” with other kinds of scientific evidence such as fingerprints.
Dr.7 Yankee testified concerning the psychological aspects of polygraphy, the subject he taught at the Keeler Polygraph Institute.
Psychopaths and sociopaths do not present a problem to the polygraph operator, he said, because the device relies on the operation of the sympathetic, or involuntary, nervous system. Further, the charts of individuals such as sociopaths are so characteristic, he testified, there would be no chance of such an individual passing without detection.
The neurotic personality, however, he said, usually shows a very erratic slant which may be a little more difficult to interpret. Individuals with IQs below 70 may also present some difficulty. He *370would also choose not to examine an individual suffering a psychotic episode because he would not know whether the subject was responding to his questions or to his or her own delusions.
The test also only indicates that an individual is telling what he or she believes to be true or untrue, not what the truth actually is. Thus, two individuals with diametrically opposed stories may "pass” the test as long as each believes he or she is telling the truth. However, such a situation, he said, is not likely to happen.
He testified that the frequency of verifiable errors was rather small. He also predicted that the polygraph would not eliminate the jury because there is always the possibility of error on the part of the examiner.
. Leonard Haroldson, President of the Keeler Polygraph Institute, testified that approximately 25 Keeler graduates were working in Michigan. All potential operators are subjected to psychological, intelligence and polygraph tests before training.
He also testified to the common policy in many states that individuals who are cleared by the polygraph examiner are not prosecuted. He therefore believed that law enforcement officials rely highly on the polygraph.
It was also his opinion that different examiners would get the same result testing the same individual.
He also testified that since approximately 1956, the techniques and requirements have remained pretty well settled, and that there have not been any radical changes or breakthroughs. Some new instruments are being used, but,. Haroldson explained, they are only recording additional phenomena.
He said the device was useful to verify truth *371rather than to detect lies, and that, in fact, would be a more accurate description of its function. Further, test results are reliable when analyzing two conflicting stories even if only one of the individuals involved is tested.
John Reid, involved with the polygraph for over 30 years, has run his own company since 1947 and had trained, at the time of this record, approximately 50 examiners. Of the 35,000 tests he said he had run in his career, he knew of only six instances where a guilty person was classified as innocent. He knew of no case with which he was associated in which an innocent person was diagnosed as guilty.
He acknowledged that he and Professor Fred E. Inbau, co-authors of the standard book in the field, Reid & Inbau, Truth and Deception: The Polygraph ("Lie-Detector”) Technique (Baltimore: Williams & Wilkins Co, 1966), had been opposed to the use of polygraph evidence in the courtroom because of concern about the caliber of examiners and the non-uniformity of the technique. They were concerned about the lack of self-regulation, but feel this is being resolved by the polygraphers’ formation of the American Polygraph Association (APA) which is setting standards for the field.
As a consequence of the formation of the APA and the beginning of state licensing, he believed there is significant improvement in the standardization of the experts.
He qualifies polygraphy as a scientific discipline, comparable to other kinds of scientific evidence such as electroencephalograms, fingerprints and firearm identification. Just as the handwriting examiner compares the known exemplar with the unknown in order to determine if both were written by the same person, the polygraph examiner *372compares responses to the known irrelevant and control questions which are then compared with the questions to which the answer is not known. The significant thing is not what the subject says, but his or her reactions to the questions and their answers.
It was his opinion that the judicial process would not be impaired if the polygraph were used in evidence, and that there would be no difference in the ability of juries to reach an accurate result. He also felt the polygraph would be useful to assist the judge in sentencing and was at the time of this record preparing legislation to accomplish this in Illinois.
Richard Howard, a polygraph examiner, testified about the polygraph examination in the case for which the special F. Lee Bailey record was originally made. Of general interest was the statement that only the one examination selected by defendant could be considered by the jury, and not any other tests which might have been taken. Further, that the entire selected test would have to be used, and not just certain favorable portions.
Richard North, a member of the Michigan State Police and a polygraph examiner at the Second District Headquarters in Detroit, reported administering 825 tests since his graduation from the National Training Center of Lie Detection in New York City three years before. He could recall only one instance where his test conclusions had been proven wrong by extrinsic evidence. He was able to re-evaluate his charts and find his mistake, but reported that the case had been borderline, the kind which most examiners ordinarily characterize as inconclusive. The subject, who had been considered truthful by the examiner, confessed when he returned to the police station following the test.
*373It was his judgment that the amount of verified error was 1 percent.
His opinion was that almost 2 percent of his own tests were inconclusive, and 4 percent throughout the department were.
He also opined it would be unusual for two experts to have diametrically opposed opinions about a single chart.
He reported use of the polygraph as, for example, a way to avoid injustices resulting from mistaken identification, and as valuable in cases where there were sharply conflicting versions of an incident. The preferred technique is to test the victim first. The polygraph examiner is the sole judge of who is tested, but the polygraph is no substitute for investigation.
His experience was that the prosecutor and State Police relied heavily on the opinions of the polygraphers. His department of the State Police had 8 full time operators and a relief person. There were other examiners who had been promoted out of the polygraph unit. He estimated there were 50 examiners, including police and private examiners, who were members of the state polygraph association.
He also felt that the polygraph was no different from other kinds of expert testimony, and a bad test could be revealed as such on cross-examination.
He also testified that the use of polygraphs as courtroom evidence would not be detrimental to the use of polygraphs as an investigative device because, for example, he has told people taking tests of instances where tests have been admitted by stipulation and it did not seem to have much effect on those tested. This was his own opinion, however, and not that of his department.
*374He testified that some polygraph operators were not members of the professional association. Nothing would stop an amateur from calling himself or herself an expert, and the only way to determine if an individual was really an expert would be to look at that person’s qualifications and experience. Those in the profession, however, would be aware of the charlatans.
Eugene Dinkel, a State Policeman who was also a polygraph examiner, noted that it was the State Police policy to test any person under suspicion who requested a test.
Bailey himself testified as to his experience with the polygraph. He had no doubt, as a trial lawyer, that "the ordinary safeguards of judicial control” would be sufficient to effectively and safely handle polygraph evidence. He noted that the polygraph community itself has come forward to give attorneys information to impeach a charlatan on the witness stand. He noted that the military never brings a person to trial if the G-2 section performing polygraph examinations clears him or her. He said the military has a much higher conviction rate than civil courts and is more accurate in its results.
Bailey claimed he knew of no conflict between experts on a single chart. He maintained this even though, in the subject case, the defendant passed part of one test and failed another. Bailey said that there was no conflict even in this case, because there had been no effort to reconcile these two tests.8
*375He also felt an unethical lawyer could not use polygraphs to deceive the court.
Judge Churchill, observing that "the people came forth with nothing, absolutely nothing in the way of evidence * * * that attacks the scientific reliability of this device, except as it was developed on cross-examination”, concluded:
"I am convinced from the evidence that within a certain field of use, assuming there are proper circumstances, competent and credible witnesses as experts, that the polygraph is a scientifically reliable instrument to determine deception.”
He observed there were serious problems concerning its use, including the possibility that there would be a battle of experts, and that unless the courts control an expected proliferation of polygraphers "we are going to have offers of expert testimony that simply can be bought”.
However, despite his hope that "someday we have an enlightened use of the polygraph in court”, he felt obliged to deny admissibility of the *376proposed polygraph evidence until the Michigan Supreme Court decreed it would be admissible.
V — Defendant Fails to Meet Davis/Frye Test
This special record, which so impressed the trial judge seven years ago, also was the basis of the offer of proof in the instant case. The record is clearly superior advocacy.
However, the witnesses were not disinterested scientists. While one would not want an expert witness without experience or background in the technical field, one would want, where the task was to demonstrate general scientific acceptability, and acknowledgment of the value of the device and the techniques by disinterested scientists whose livelihood was not intimately connected with it. In addition, the Davis/Frye test requires acceptability by those in certain established scientific disciplines. This was not forthcoming.
Further, with the aid of excellently written and argued amicus presentations9 in the matter at bar, we find that the case for the polygraph has not in fact been clearly decided and that acceptance by scientists rather than polygraph operators has not been general and widespread.
It is significant to note that only the defendant and the American Polygraph Association argued for admissibility. Even those who might be generally expected to be the first to advocate devices which would favor defendants, such as the State Appellate Defender Office (SADO) and the Michigan Association of Criminal Defense Lawyers, op*377posed admissibility. The prosecutors, as represented by the Oakland County Prosecutor in the case at bar, and the Prosecuting Attorneys Appellate Service, argued on the same side of the issue as SADO and the Criminal Defense Lawyers.
Numbers by themselves are not necessarily persuasive, but, where we are looking for evidence of general acceptability, such a broad spectrum of opinions arguing against acceptability is certainly significant.
While the special record before us establishes that the polygraph is accepted as reliable by polygraphers, it does not establish that polygraph analysis is accepted as reliable by the scientific community. Credentials of the witnesses, although outstanding for polygraph technicians, are not those of scientists. Therefore, unless we depart from the standard Davis/Frye test for admissibility, defendant has failed to convince us that the polygraph should be admitted into evidence at trial in our state.
VI — Should Michigan Depart from the Davis/ Frye Rule?
A. Where Do Other Jurisdictions Stand?
a) Frye Majority Rule.
No appellate court opinion specifically overruling Frye has been called to our attention or discovered by our research. As late as 1974, it was said in United States v Skeens, 161 US App DC 131, 134; 494 F2d 1050, 1053 (1974):
"The leading case in this Circuit is Frye v United States, 54 US App DC 46; 293 F 1013 (1923), which holds such tests inadmissible. This case has been followed uniformly in this and other Circuits and there *378has never been any successful challenge to it in any federal court.”
See also the most recent ALR survey, 23 ALR2d 1306, § 2 (Later Case Service, 1976, Supp, pp 127-128), which cited cases holding polygraph tests inadmissible in the following states: Alabama, Alaska, Arizona, Florida, Illinois, Indiana, Iowa, Kansas, Kentucky, Louisiana, Maryland, Michigan, Missouri, Nebraska, Nevada, New Jersey, New York, North Carolina, Ohio, Oklahoma, Pennsylvania, South Dakota, Texas, Virginia and Washington. See further the excellent survey in 3A Wigmore, Evidence (Chadbourn rev), § 999, pp 946-949, fn 2.10
b) Admission by Stipulation in a Few States.
There has been some chipping away from Frye by those states admitting polygraph evidence by stipulation of the parties. See, especially, State v Valdez, 91 Ariz 274; 371 P2d 894 (1962).11 See also, *379State v Ross, 7 Wash App 62, 69; 497 P2d 1343, 1347 (1972); State v Stanislawski, 62 Wis 2d 730, 742-743; 216 NW2d 8, 14 (1974). See also State v McDavitt, 62 NJ 36, 46; 297 A2d 849, 854-855 (1972); People v Houser, 85 Cal App 2d 686, 695; 193 P2d 937, 942 (1948); State v Brown, 177 So 2d 532, 533 (Fla App, 1965); State v Freeland, 255 Iowa 1334, 1339; 125 NW2d 825, 828 (1964) (allowed admission by joint stipulation, but rejected ordering test even at defendant’s request); State v Galloway, 167 NW2d 89, 94-95 (Iowa, 1969).
However, consideration of polygraph evidence even with a stipulation of admissibility has not been unanimous. State v Trimble, 68 NM 406; 362 P2d 788 (1961); Pulakis v State, 476 P2d 474, 479 (Alas, 1970) ("A stipulation for admission does not increase the reliability of polygraph results and therefore should not lead to any deviation from the exclusionary policy.”); LeFevre v State, 242 Wis 416, 425; 8 NW2d 288, 292 (1943); People v Zazzetta, 27 Ill 2d 302, 309; 189 NE2d 260, 264 (1963) (”[W]e think it manifestly unfair to bind him [defendant] by a stipulation regarding the trustworthiness of scientific opinion far beyond his expected ken.”)
c) Some Academic Criticism.
The Frye standard12 has been criticized as stricter than that applied to other scientific evidence, and as in effect amounting to a test of judicial notice.13 See, e.g., McCormick, Evidence *380(1st ed), § 170, pp 363-364; Kaplan, The Lie Detector: An Analysis of Its Place in the Law of Evidence, 10 Wayne L Rev 381, 402-407 (1964). Thus, while other tests may be admitted in the face of serious scientific dispute concerning their credibility, in states following Frye, polygraph tests have not been.14
d) Five Cases Challenging Frye.
*381Certain jurisdictions, however, particularly several Federal district courts, have opted to admit polygraph evidence by substituting, at least implicitly, another standard for Frye.
For example, although the polygraph was not admitted in United States v Wainwright, 413 F2d 796, 803 (CA 10, 1969), it was precluded only because the expert did not testify
" 'that the proposed test is an accepted one in his profession and that it has a reasonable measure of precision in its indications.’ 3 Wigmore on Evidence (3d ed), § 990. The trial court properly excluded it even though in a proper case it may be admissible.” (Emphasis added.)
Thus, in this instance, by looking to acceptability by the profession, the tenth circuit appeared to substitute the polygraph profession for the professions of physiology and psychology as required by Frye.
A similar approach to modifying Frye was taken in an extremely thoughtful United States District Court opinion, United States v DeBetham, 348 F Supp 1377, 1381 (SD Cal, 1972). That Court reviewed testimony of physiologists and psychologists and applied it to the Frye standard. The Court concluded that "physiologists as a group have no real quarrel with the polygraph as an instrument for measuring and recording certain physiological responses of the human body”. Thus, *382the evidence apparently indicated that a properly designed, operated and maintained machine would accurately record certain specified physical changes. 348 F Supp 1381. A psychophysiologist, however, testified that while the device is "a useful tool for studying physiological changes”, "he expressed doubts as to whether any conclusions about deception could be derived therefrom”. 348 F Supp 1381 (emphasis added).
Psychologists, the Court indicated, also do not support the polygrapher’s claim of widespread acceptance. The Court noted that even defendant’s psychologist witness "could not state unequivocally that the polygraphic lie detection technique enjoyed the general recognition of his profession”, and had authored a paper conceding:
" 'The lie detection technique was developed pragmatically by the field users, with the theoretical underpinnings following later. Even now, there exists no generally accepted theory which adequately explains all of the phenomena.’ ” Quoting Raskin, "An Experimental Study of Field Techniques in 'Lie Detection’ ”, a paper presented at the Society for Psychophysiological Research, St. Louis, October 24, 1971. 348 F Supp 1381, fn 10.
The Court concluded that while psychologists and psychophysiologists may eventually scientifically verify "the underlying psychological hypothesis of the polygraph, no conclusive findings have been brought to the Court’s attention which would warrant a holding that the Frye test of admissibility, as interpreted by subsequent decisions, has finally been satisfied after almost half a century”. 348 F Supp 1381-1382.
However, the DeBetham opinion did not stop there, but instead interpreted Frye to indicate a *383concern with "the normal obligation of the trial judge to protect the jury from exposure to evidence which might mislead them”. 348 F Supp 1382. Thus, the court says that Frye itself did not intend "to require a specially rigorous standard of admissibility for such evidence”. 348 F Supp 1382. However, the Court acknowledges that nearly all subsequent decisions have applied Frye "to require more of the polygraph than of other types of scientific evidence”. 348 F Supp 1383.
To this the Court responded by noting the recommendations of McCormick, supra, fn 13, and of Kaplan,15 and agreeing that the standard for the admission of polygraph evidence is artifically high and that normal evidentiary requirements for admissibility should be established. This would weigh the probative value of the polygraph evidence against factors such as the possibility of jury prejudice or undue delay, but would apparently require the support of at least "a substantial body of scientific opinion”. 348 F Supp 1384.16 Under the DeBetham approach, the qualified expert would be permitted to present foundational evidence demonstrating the device’s "substantial reliability and acceptance, in an effort to establish its probative value.” 348 F Supp 1384.
Thus DeBetham looked to the testimony of poly*384graphers who testified to accuracy in excess of 90 percent, and of studies performed by polygraphers in which they experimentally achieved accuracy between 80 and 90 percent.17 348 F Supp 1385. The Court justified placing the opinions of these polygraphers along with those of the more critical physiologists and psychologists because of advances in the field of instrumental lie detection since Frye. Therefore, the Court reasoned, the search for scientific acceptance should not be limited "to those branches of science familiar to the courts in 1923”. 348 F Supp 1388.
The Court found particularly significant that the polygraph is extensively used in Federal and state law enforcement agencies,
"who apparently rely upon the technique in their day-to-day prosecutorial decision making. * * * Such widespread acceptance, though not of a strictly 'scientific’ nature, is considerably more important, in terms of its practical effect on society, than any academic approval by the psychological community could ever be; and the Court finds it to be an ample substitute, if necessary, for the requirement of substantial scientific acceptance.” 348 F Supp 1389.
Having found that the probative value of the polygraph was established, the Court also examined various policy issues which might militate against admission. However, the Court found that none18 of these objections had merit.
However, because this was a District Court decision, and the Circuit Court of Appeals had held *385that polygraph results are properly excluded from evidence, the Court did not adopt a finding of admissibility. The Court of Appeals affirmed that result. 470 F2d 1367 (CA 9, 1972).
Similar deference did not deter the United States District Court in United States v Ridling, 350 F Supp 90 (ED Mich, 1972). There, in a perjury case, the court, citing United States v Wainwright, supra, declared that the question of reliability of the polygraph is to be determined in each individual case, noting that evidence in the case at bar was that the techniques and equipment had improved markedly in the past ten years.
In order to determine what it considered to be the proper test of admissibility, the court first analyzed what it felt to be the "historical process of developing the admissibility of opinions interpreting scientific evidence”. 350 F Supp 94. First,
"Someone has an idea and a theory * * * developed] * * * until it has some acceptance.
"When opinions interpreting the results are first offered in Court, the underlying premises require a great deal of proof, as well as does the proper use of these premises, the necessary controls used in the specific cases and the appropriate qualifications of the expert. On proper proof, the evidence becomes admissible. The attention of the Courts at this point seems to be directed at the proper qualification of an expert witness, including testimony, establishing the underlying theory.
"Finally, the underlying principles and premises become so well established and known that the only real issues for determination in connection with the reception of evidence is the proper use of the principles, premises and theories and the use of adequate controls in the specific case to assure good results. In other words, at this stage the Courts judicially notice the basic theories and premises. They need no longer be proved.”
*386The Court then held:
"The use of expert opinions interpreting the results of lie detector tests in Court is still at one of the first two stages, and judicial opinions denying admissibility give way to the developments in the science itself or in the techniques used in its application or in the interpretation of the results. The record in this case indicates that the theory of the polygraph is sound and that it is directly relevant to this case (a perjury case), and that therefore the cases denying admissibility on these grounds are not controlling.” 350 F Supp 95.
Thus, declaring "the scientific psychological basis for the polygraph examination is well established”, 350 F Supp 93, Ridling rejected the Frye conclusion. The only hesitation the Court expressed about the technology of the device was that the machine seems "to have outstripped the general acceptance of polygraph personnel in development”. To surmount this obstacle, Ridling provided for the use of court-approved experts to test the defendant, with the opinions of both defendant’s polygrapher and that of the court to be admissible as long as the court’s expert was able to come to a conclusion, even if the two experts disagreed. 350 F Supp 96-97.
F. Lee Bailey appeared before the Court in United States v Zeiger, 350 F Supp 685 (DDC, 1972), rav'd 155 US App DC 11; 475 F2d 1280 (1972). That Court, similar to DeBetham but contrary to Ridling, recognized it was bound by Frye. The Court then recognized that "polygraphy has emerged from * * * [the] twilight zone into an established field of science and technology”. 350 F Supp 688. To so establish this status the Zeiger court, like DeBetham, gave consideration "to the opinions of experts who are neither physiologists nor psychologists”. Although Frye required *387" 'recognition among physiological and psychological authorities’ ”, the Court rejected this exclusive approach because although "polygraphy at one time may have been dependent on physiological and psychological authorities for certification of its reliability, it is no longer appropriate to confine consideration solely to those disciplines”. 350 F Supp 689.
Therefore, the Court concluded,
"The testimony of the experts and the studies appearing in the exhibits19 lead the Court to believe that the polygraph is an effective instrument for detecting deception. The failure of the Government to demonstrate significant disagreement with this basic proposition, the absence of statistical data pointing to any other conclusions, and the accepted and widespread absorption of the polygraph into the operations of many government agencies, all confirm the Court’s conclusion that the polygraph has been accepted by authorities in the field as being capable of producing highly probative evidence in the court of law when properly used by competent, experienced examiners.” 350 F Supp 690.
Zeiger was reversed by the appellate Court. 155 US App DC 11; 475 F2d 1280 (1972).
Commonwealth v A Juvenile (No 1), 365 Mass 421; 313 NE2d 120 (1974), is a somewhat unique case involving the trial court’s reluctant refusal to admit defendant’s polygraph test results because of the Massachusetts equivalent of the Frye rule. The defendant originally moved for consideration of polygraph tests already performed by an examiner of his own choice but subsequently requested a court-ordered polygraph examination as well.
*388The Massachusetts Supreme Judicial Court recognized the advances in the state of the art but stated "we nevertheless are unwilling to say at this time that the standard in the Fatalo [Frye] case [Commonwealth v Fatalo, 346 Mass 266; 191 NE2d 479 (1963)] has been met and that the polygraph test results should henceforth be subject to the same rules of evidence applicable to other forms of acceptable expert scientific evidence.” 313 NE2d 123-124. It reached this conclusion because it found, after considering the District Court decisions in DeBetham, Zeiger and Ridling and the relevant data:
"[D]espite very significant progress in recent years, the field of polygraphy is still challenged forcefully on theoretical grounds and has yet to achieve a predictable level of consistency among examiners. For these reasons we do not believe that at this time polygraph test results should be generally admissible in evidence in criminal trials. Commonwealth v Fatalo, supra.” 313 NE2d 125.
The Court, however, went on to hold
"that if a defendant agrees in advance to the admission of the results of a polygraph test regardless of their outcome the trial judge, after a close and searching inquiry, * * * may, in the proper exercise of his discretion admit the results, not as binding or conclusive evidence, but to be considered with all other evidence as to innocence or guilt.” 313 NE2d 124.
It further held "a defendant could not be compelled initially to take such an [polygraph] examination”. 313 NE2d 127.
This case is peculiar in that it purports to follow the Frye rule but nonetheless makes a special exception to permit the defendant to submit to *389polygraph testing and offer the tests in evidence, if he agrees "in advance that the results would be admissible irrespective of the results”, 313 NE2d 126-127.
e) Present Status of Frye.
No appellate court has specifically overruled Frye, and Frye remains the majority rule on the admissibility of polygraph evidence. Although several states admit such evidence on a limited basis upon stipulation of the parties, still other states specifically exclude this evidence, even on stipulation. There has been some academic criticism, noted in a few courts, that the Frye rule is more stringent than the rule ordinarily used to test the admissibility of scientific evidence. Four interesting Federal cases have called the Frye rule into question. In Wainwright, a Federal Circuit Court of Appeals sub silentio substituted a different rule for Frye. In DeBetham, a different Circuit Court of Appeals affirmed a District Court opinion which thoughtfully considered a rule different from Frye but retained Frye. In Zeiger, still a different circuit sharply reversed a District Court trying to overrule Frye. Finally, the District Court for the Eastern District of Michigan substituted, in Ridling, a more lenient test than Frye. A peculiar Massachusetts case, Commonwealth v Juvenile, supra, claims to follow the Frye rule, but makes a special exception where the defendant agrees in advance to take a polygraph test and submit the tests irrespective of results. Yet, by and large, these cases remain anomalies.
B. Why the Frye Rule Should Not Be Changed
The two reasons why the Frye requirement of acceptability by the scientific community should *390not be changed were summed up succinctly in Juvenile:
"[Djespite very significant progress in recent years, the field of polygraphy is still challenged forcefully on theoretical grounds and has yet to achieve a predictable level of consistency among examiners.”
Even the Federal Courts which modified Frye recognized that these problems existed, but they were not deemed dispositive because these courts also found that the polygraph had received general acceptance. Their new tests effectively substituted the judgment of polygraphers for those of scientists in order to demonstrate scientific acceptance.
However, in order to find such acceptance, these courts adopted what was in fact a logical fallacy.
Under the present state of the art the general acceptance of the polygraph among psychologists and physiologists cannot be demonstrated, because such acceptance does not exist. Therefore, these courts, in order to find general acceptance, found it amongst polygraphers. Once finding general acceptance, the courts then found they did not have to rely on scientific testimony, but were able to rely on the testimony of polygraphers to establish reliability of the device.
Such reasoning is circular and further reinforces our decision not to modify or otherwise change the Frye/Davis rule.
To further buttress this position, it is useful to review the factors behind the theoretical and practical problems still confronting the polygraph.
To begin with, serious scientific questions have been raised about the ability to evaluate the polygraph at all.
The basic question which challenges the funda*391mental theory underlying the polygraph is whether there is a measurable relationship between conscious deception and a person’s physiological state. Skolnick, Scientific Theory and Scientific Evidence: An Analysis of Lie-Detection, 70 Yale LJ 694, 700-702 (1961); People v Leone, 25 NY2d 511, 515; 255 NE2d 696, 698; 307 NYS2d 430, 433 (1969); and see Henderson v State, 94 Okla Crim 45, 51-52; 230 P2d 495, 501-502 (1951), cert den 342 US 898; 72 S Ct 234; 96 L Ed 673 (1951).
While there appears to be little question that a properly designed, maintained and operated machine will accurately record physiological phenomena, there is a great deal of debate concerning what precisely these recordings mean.
One famous survey of 1,200 psychologists is often pointed to as signifying that psychologists have indeed accepted the device. Cureton, A Consensus as to the Validity of Polygraph Procedures, 22 Tenn L Rev 728 (1953), cited in Richardson, Modern Scientific Evidence (2d ed), § 10.3, p 311. The results, however, do not appear to sustain that contention. For example, in response to the request for an opinion concerning validity of the polygraph "as an indicator of emotional reaction”, 25 percent of the psychologists polled indicated they thought it was .highly valid and useful, and 70 percent felt it was moderately valid and useful. The responses, however,
"by no means justify the conclusion that psychologists regard the polygraph as a valid instrument for the detection of lies. What the question asks, in essence, is whether the psychologist believes that blood-pressure can be measured with a sphygmomanometer and whether respiration can be measured with a pneumo*392graph.” Levitt, Scientific Evaluation of the "Lie Detector”, 40 Iowa L Rev 440, 456-457 (1955).
The question does not ask if the respondent believes the recorded emotional reaction indicates lying. More significant, therefore, is that 64 percent disagreed with the polygraphers’ basic statement of the theory underlying effectiveness of the device, viz.,
"The fear of being found out and/or conscious efforts to deceive are the main cause of significant reactions in polygraphic tests of deception.” Levitt, 40 Iowa L Rev 457.
Furthermore, the group tested included non-psychologists as well as psychologists, and the non-psychologists including polygraph examiners seemed to be twice as enthusiastic as the psychologists as to the high efficacy of the polygraph. Cureton, supra, 22 Tenn L Rev 734.
This challenge to the basic premise behind operation of the polygraph is long-standing. Both the literature and the record in this case indicate that experimentation began early to demonstrate that emotional changes created physiological changes which could be recorded by predecessors of the polygraph. However, it is still argued that while these changes can be recorded and noted,
"There seems to be little evidence that upholds the claim to a regular relationship between lying and emotion; there is even less to support the conclusion that precise inferences can be drawn * * * between emotional change and physiological response.” Skolnick, supra, 70 Yale LJ 727.
"[T]he autonomic response to the critical question will always be influenced by individual difference variables which are not a function of the subject’s guilt or *393innocence.” Lykken, Psychology and the Lie Detector Industry, 29 Am Psychologist 725, 733 (1974).
Further,
"[a] change in respiration, blood pressure, or GSR [skin response] beyond what would seem normal is considered important, but no speciGc physiological change is uniquely associated with lying. Instead, deception is diagnosed by comparing the subject’s response on relevant items with his response on items that should be equally arousing or anxiety-provoking to an individual who is not lying.” Greenfield & Sternbach, eds, Handbook of Psychophysiology (New York: Holt, Rinehart & Winston, Inc, 1972) ch 19, p 749. Hereafter Greenfield & Sternbach. (Emphasis added.)
Thus, the Warren Commission investigating the assassination of President Kennedy "concluded that physiological responses due to factors other than deception, such as fear, anxiety, nervous and other emotions, had to be given consideration in evaluating the polygraph”. Burkey, The Case Against the Polygraph, 51 ABA J 855, 856 (1965). See also, Dearman & Smith, Unconscious Motivation and the Polygraph Test, 119 Am J Psychiatry 1017 (1963), cited id.20
The theoretical opposition is apparently strengthened by another operational weakness of the test. Scoring of the physiological parameters is performed on an individual basis, with the scores which indicate deception determined by the judg*394ment of the individual examiner on each individual subject. Thus, the same score may indicate deception for one subject and not for another. The determination is ultimately made only by the examiner each time. Thus, there appear to be no standard "normal” ranges outside of which deception is indicated, and decisions appear to be highly judgmental and ad hoc.
"When one considers the use of the polygraph in a real life situation, it becomes clear why it is difficult to evaluate. The interrogation procedure is an art rather than a science — much depends upon the manner in which the individual interrogator formulates his questions, the tone of his voice, and the subtle cues he may provide in the interrogation, etc.. The experienced interrogator must intuitively deal with problems inherent in such issues as individual differences in reactivity, response specificity, and the overall state of the person being questioned. Interrogators who have worked with 'lie detection’ will readily admit that the psychological edge the machine provides during the pretest interview is, from their point of view, of major importance. Some go on to say that, even if the machine were out of order, it would still be a major asset. Scientific evaluation of the validity of the 'lie detection’ procedure is very difficult, since the test itself is only a small part of the total procedure * * * No fully satisfactory way is available at this time for evaluating the overall effectiveness of the technique.”21 Greenfield & Sternbach, pp 749-750, 751.
Further, we find there is difficulty in empirically *395verifying polygraphic findings, Skolnick, supra, 70 Yale LJ 699, some questions concerning the accuracy figures, 70 Yale LJ 714-721, Lykken, supra, 29 Am Psychologist 725, 734, 738, and an apparent lack of a rational scientific explanation for this phenomenon.22
The fact is, that the results of most polygraph examinations are not verified by extrinsic evidence,23 and there are no scientific studies performed by other than polygraphers which purport to demonstrate scientific means of verification. This factor alone has been responsible for much of the anti-polygraph sentiment outside the polygraph profession.
Reid and Inbau themselves recognize that there may never be a confession or other evidence developed to support the examiner’s conclusion. Inbau & Reid, The Lie-Detector Technique: A Reliable and Valuable Investigative Aid, 50 ABA J 470, 472 (1964). United States v Wilson, 361 F Supp 510, 513-514 (D Md, 1973). Most studies have been performed by polygraphers, with few psychological or other scientific studies available on the validity of polygraphic investigation, partly "because of the difficulty in checking the results against irrefutable extrinsic facts”, Burkey, The Case Against the Polygraph, 51 ABA J 855, 856 (1965), and perhaps, partly, because of "lack of funds rather than lack of interest”.24 Levitt, Scientiñc Evaluation of the *396"Lie Detector”, 40 Iowa L Rev 440 (1955). One well-known text, not by polygraphers but by psychophysiologists, forthrightly states, "No adequate evaluation of the validity of the polygraph * * * [as 'a scientific way to record lying’] is yet available”. Greenfield & Sternbach, Handbook of Psycho-physiology, p 750.
Thus, claimed accuracy figures in the polygraph must be considered with these limitations in mind:
(1) They are reported by polygraphers.25
*397(2) They are largely unverified by extrinsic evidence.26
(3) There appears to have been no significant scientific studies by other than polygraph operators purporting to demonstrate such accuracy.
(4) There appear to be no substantial techniques developed outside the polygraph profession which would permit scientific verification of these figures.
Further, even if we grant polygraphers that their accuracy claims are true, we must confront the question of what is, after all, being analyzed. The operator "cannot prove that he [the subject] is telling the truth, but only that he is telling what *398he believes to be the truth”. Streeter & Belli, The 'Fourth Degree’: The Lie Detector, 5 Vand L Rev 549, 557 (1952), cited by Note, Problems Remaining for the Generally Accepted” Polygraph, 53 B U L Rev 375, 385, fn 94 (1973).
Thus, even proponents of the device agree:
"[T]he polygraph will not detect the witness who makes honest misstatements of fact, and it may not detect the hardened individual who recognizes no responsibility to conform to moral codes or standards of the community.” Richardson, Modern Scientific Evidence (2d ed), § 10.3, p 310.
In addition, the polygraph, unlike many other devices, relies solely on examiner judgment for its value.
Without examiner opinions, results of "lie detector” tests are meaningless. While the results of the examination are recorded on graphs, or polygrams, the determination of veracity or falsehood is made as a result of the examiner’s interpreting both the polygraph and his or her evaluation of the behavior of the person tested in the light of the questions asked. Note, Pinocchio’s New Nose, 48 NY UL Rev 339, 341 (1973). Reid and Inbau themselves state that the single most important factor in assuring proper testing procedures and interpretation is the polygraph examiner. Reid & Inbau, Truth and Deception, p 235. See discussion, Henderson v State, 94 Okla Crim 45; 230 P2d 495 (1951);27 United States v Wilson, 361 F Supp 510, 512-513 (D Md, 1973).
*399Without a capable examiner, even a totally effective polygraph would be worthless.
"An untrained, inexperienced examiner is about as incapable of detecting deception with a polygraph as a layman is to diagnose a heart condition with a stethoscope or a cardiograph * * * .” Wicker, The Polygraphic Truth Test and the Law of Evidence, 22 Tenn L Rev 711, 712 (1953).
This extreme dependence on the operator is emphasized by studies indicating that successful tests vary directly with the experience of the examiner. Horvath & Reid, The Reliability of Polygraph Examiner Diagnosis of Truth and Deception, 62 J Crim L C & P S 276, 279 (1971).28 It is also suggested that a trained examiner can recognize abnormalities which might affect test results.29
There is a question whether one examiner will arrive at the same conclusion as another from the first examiner’s test records. Examiners testifying in the instant case said this was the usual result. But see Bersh, A Validation of Polygraph Examiner Judgments, 53 J Applied Psychology 399-403 *400(1969), cited in Lykken, supra, 29 Am Psychologist 735; Skolnick, supra, 70 Yale LJ 704; Note. The Polygraphic Technique: A Selective Analysis, 20 Drake L Rev 330, 335 (1971).
The polygraphers maintain, too, that no one can "beat” the machine. However, others dispute this.
Inbau and Reid themselves initially found that psychopathology in the suspect, especially neurotics and psychopaths, made detection difficult. Levitt, supra, 40 Iowa L Rev 451. While the examiners testifying in the instant case did not consider this a significant problem, the controversy remains whether certain physiological and mental abnormalities may materially affect test results, without being detected by the examiner. See, e.g., Note, supra, 20 Drake L Rev 333, fn 36. Henderson v State, 94 Okla Crim 45, 51; 230 P2d 501-502 (1951). This may include even factors well within the control of the subject, such as flexing the toes,30 or taking certain drugs. It has been demonstrated that habituation renders the technique ineffective, and it has been suggested that biofeedback training may make it possible to "beat” the machine. See Greenfield & Sternbach, pp 776-779.
The original Inbau publication which opposed the admission of the polygraph did so largely because of concern about the competence of examiners.31 The operator plays such a substantial part *401in the testing procedure, in the formulation of questions and the subjective evaluation of the subject, in addition to the reading of graphs, that this disquiet was obviously well-founded. The importance of the pretest interrogation and the lack of standardized procedures, even today, make such concern still significant. So despite real progress, it does not appear that the goal of uniform and high competence has yet been reached.
The "American Academy of Polygraph Examiners was formed with the avowed purposes of training experts and gaining more widespread recognition for the profession”. Comment, Evidence: Lie Detector Tests — Effect of Prior Stipulation on Admissibility of Results, 18 U Fla L Rev 527, 532 (1965). For AAPE statement of principles see Richardson, Modern Scientific Evidence (2d ed), § 10.17, pp 332-335. However, even where there are statutorily-imposed standards,32 the statutes apparently do not require the study of psychology and physiology, the two fields which seem to be most critical of the polygraph,33 and the fields described in Frye as those in which the polygraph must attain general acceptance before it is to be admitted into evidence.
Further, even some courts which have admitted the polygraph, recognize that "as a profession, polygraphy has yet to achieve a uniform minimum level of competence”. Commonwealth v A Juvenile *402(No 1), 365 Mass 421; 313 NE2d 120, 124-125 (1974) (admitting polygraph where defendant requests he be given a test and agrees to abide by whatever results ensue). Judge Joiner wrote, in Ridling,
"[W]ith the polygraph, the theory, the technique and the hardware seem to have outstripped the general acceptance of polygraph personnel in development. Although the profession of polygraph experts is becoming standardized and professionalized, it has not yet developed adequate ways and means to police itself. Although the profession has adopted a code of ethics and machinery exists in the professional organization to discipline the unqualified and unethical, the chance of serious impropriety on the part of polygraph examiners must be considered.” 350 F Supp 96.34
In view of the critical, even decisive, role of the polygraph examiner in the testing and evaluation process, we find such concerns to be disturbing and not easily overlooked.
We therefore find that the Frye./Davis test establishes the requisite standard for polygraph testimony and that there has been no clear consensus established on the part of the relevant scientific community which would demonstrate that the state of the art is such that the polygraph should be admissible at trial in our state.
Serious questions remain about the reliability of the technique and of its operators. We are also not convinced that the validity of the device has been satisfactorily established.
The record in this case, while full and well-documented, seems to demonstrate that polygra*403phy has made substantial strides. Perhaps eventually, the technique will have progressed to the point where we may again consider the question. As of the present, however, the technique is yet too questionable to satisfy the high evidentiary requirement of our state.
It is true that the polygraph has apparently become widely used as an investigative tool by government35 and by private industry. Note, supra, 20 Drake L Rev 343-344; Trovillo, Scientific Proof of Credibility, 22 Tenn L Rev 743 (1953). But the differences between use as an investigative and as an evidentiary device are great, and a technique accepted for one limited purpose may not yet be suitable for use in the other. Damage done by false results may be more easily correctible in an investigative situation. This was recognized by the prosecuting attorneys who opposed admissibility in the case at bar in the face of the documented use of the polygraph by prosecutors in the investigative phase.
*404We have also found that the Frye/Davis standard is the appropriate one to use to determine admissibility of polygraph testimony. Very basic questions concerning the theory, validity and reliability of the device have not been satisfactorily answered, and until it is demonstrated that they have been, it is unwise to follow the new tests of admissibility recommended by some of the federal district courts which have permitted use of polygraph testimony.
The fact is that the polygraph, while scientific evidence, is different from other scientific evidence. "[T]he quantity it attempts to measure — the truthfulness of a witness — is * * * directly related to the essence of the trial process”. Note, The Emergence of the Polygraph at Trial, 73 Colum L Rev 1120, 1141 (1973). This does not render the evidence less admissible than others, but it does serve to underscore at least some of the rationale underlying the judiciary’s concern for as much scientific consensus as possible concerning the polygraph’s value. Further concern is based on a fear that by use of the polygraph, we run dangerously close to substituting a trial by machine for a trial by jury.36
*405It therefore is best to adhere to a standard which in effect permits the experts who know most about a procedure to experiment and to study it. In effect, they form a kind of technical jury, which must first pass on the scientific status of a procedure before the lay jury utilizes it in making its findings of fact.37 In the case of the polygraph, such forbearance has proved appropriate, for the dispute concerning its value, which apparently persists in the scientific community, should not be prematurely decided in the courtroom. Note, Evolving Methods of Scientific Proof, 13 NY L Forum 679, 683, fn 18, 696 (1968).
It is argued that "the principal role of a trier of *406fact is the search for truth and any reasonable procedure or method to assist the court for truth should be employed”. Tarlow, Admissibility of Polygraph Evidence in 1975: An Aid in Determining Credibility in a Perjury-Plagued System, 26 Hastings LJ 917, 919-920, fn 21 (1975), citing People v Cutler, (Cal Super Ct, Los Angeles County, Docket No A176965, Nov 6, 1972), 12 Crim L Rep (BNA) 2133 (1972). But first, it must be ascertained that the procedure will indeed assist in finding the truth. To this end, the Davis/Frye test was developed, and to this end our Michigan courts shall retain it.
VII — Some Other Questions About the Polygraph
Even if the Frye criteria had been met, or if another test were adopted which polygraph testimony could satisfy, the amici have raised some philosophical and constitutional questions concerning the wisdom of admitting this evidence. While it is not necessary to, and we do not resolve these today, it is wise to at least recognize that these questions exist.
Thus, for example, the United States Supreme Court intimated in Schmerber v California, 384 US 757, 764; 86 S Ct 1826; 16 L Ed 2d 908 (1966), that forcing a defendant to take a polygraph test would violate the fifth amendment right against self-incrimination.38 It has been argued that it is *407impossible to waive this right where a polygraph examination is involved, because, since such tests reveal the unconscious, the examinee never knows in advance the precise scope of his or her waiver.39 Silving, Testing of the Unconscious in Criminal Cases, 69 Harv L Rev 683, 692 (1956).
Silving also brilliantly discusses a German case,40 which reversed a conviction based on results of a polygraph test to which the defendant had consented. While we express no opinion on whether, if polygraph evidence is scientifically *408valid it may be admitted with the consent of defendant, we observe that the court, according to Silving, reasoned that
"application of the polygraph test, in the course of which the unconscious mind of the accused answers involuntarily, violated both the procedural requirement of freedom of the accused to conduct his defense and his ethical status as an independent moral 'personality.’ For moral personality implies 'an essential and inalienable autonomous area of the mind, which must remain inviolate also in criminal procedure.’ ” 69 Harv L Rev 689 (footnotes omitted).
The accused’s " 'freedom of will with regard to decision and action’ ” thus makes any waiver in regard to the polygraph untenable because
"freedom precludes deliberate suspension of his consciousness for the purpose of exploring his unconscious. The accused cannot waive this standing and this freedom, for they are not privileges granted to him, but fundamental principles of criminal procedure, based upon law, which must be observed in accordance with the idea of government of laws.” 69 Harv L Rev 688-689.
Thus, the Court accepted defendant^ contention, based on a principle of the Constitution of the Bonn Republic: " 'Man’s dignity is inviolate. All state authority has the duty to respect and protect it.’ ” 69 Harv L Rev 688.
In our own country, considerations of fairness have also been significant in considering admissibility of the polygraph. For example, one Court rejected polygraph evidence after noting, among other reasons, that only those defendants who "pass” examinations want them admitted, thereby making it difficult for the indigent defendant to be *409similarly advantaged, primarily because he or she could not afford to take an examination in confidence knowing that the results need not be revealed until and unless they were favorable. This is a significant advantage compared with the indigent defendant who can take an examination only with the government’s financing and knowledge, and who must stick with the results of that one attempt. Further, the Court observed that the sense of security exhibited by the defendant who knows he or she can continually take examinations "diminishes the fear of discovered deception, upon which an effective examination depends.” United States v Wilson, 361 F Supp 510, 514 (D Md, 1973).
For an excellent marshalling of arguments against use of the polygraph as substantive evidence at trial, see statement of Henry S. Dogin, Deputy Assistant Attorney General, Criminal Division, the Department of Justice, in testimony before the Foreign Operations and Government Information Subcommittee of the House Committee on Government Operations, 16 Crim L Rep (BNA) 2306 (1975).41
*410The amici and the authorities on which they rely suggest there may also be a possibility of bogging down trials with collateral matters, perhaps resulting in a trial of the polygraph,42 or a battle of experts.43 They note there may be a sixth amendment problem involving the right to confrontation because of the impossibility of cross-examining the machine.44 There is also the possibility of a hearsay problem if polygraph evidence is used to prove the truth of the statement made by the subject during the test. Note, The Polygraph: Scientiñc v Judicial Acceptance, 27 U Miami L Rev 254, 259-260 (1972). Contra, United States v Ridling, 350 F Supp 99.
Such arguments are of varying merit, but there is neither the need to decide these today nor the record before us to help us do so. Indeed, perhaps it is most useful to observe that such serious questions serve, if nothing else, to reinforce the propriety of the high standard of admissibility required of the polygraph.
*411VIII — Use of Polygraph at New Trial Motion
Our conclusion thus far is that the state of the art does not warrant the introduction of poly-graphic tests into evidence in a trial. But the precise question before us is different. We are asked whether a judge may consider polygraph tests in a motion for a new trial on the basis of newly discovered evidence. More specifically, the issue in this case is whether polygraph tests could be used to bolster the testimony of a new witness (the defendant’s own test is immaterial since he had already testified at trial and the jury rejected his testimony).
In short, is the character of the proceedings on a motion for new trial because of newly discovered evidence so significantly different from the proceedings at trial as to permit the use of polygraph tests in the one and deny its use in the other?
The answer is that the procedure at trial and at a post-conviction hearing for new trial is different and significantly so. The procedures are significantly different because their purposes are significantly different. The purpose of a trial is to determine the guilt or innocence of the defendant. The purpose of a post-conviction hearing for a new trial is, as its name suggests, an action to determine whether there should be such a trial. It is a preliminary, not a final procedure.
The implication of this difference is basic. Since the defendant’s guilt or innocence is not at issue, some procedures are permissible which would not be acceptable at trial. For example, the motion may be argued on the basis of affidavits, which, of course, would not be possible at trial. It does not determine the defendant’s guilt or innocence.
The question arises therefore whether, under *412these post-trial conditions preliminary to actual trial, the judge may exercise his or her sound discretion as to whether to be guided in the exercise of that discretion by the results of a polygraph test. In other words, may the judge at this stage prior to trial exercise a discretion similar to that which the prosecutor often exercises before going to trial, namely consider the results of a voluntarily taken polygraph test?
In our opinion the judge in a post-conviction hearing on a motion for new trial based on newly found evidence may in his or her discretion consider the results of a polygraph examination. If the judge chooses to consider such results it must be under certain hereinafter enumerated conditions.45
(1) The results of the polygraph tests are offered on behalf of the defendant.
(2) The polygraph test was taken voluntarily.
(3) The professional qualifications of the polygraph examiner are approved.
*413(4) The quality of the polygraph equipment is approved.
(5) The procedures employed are approved.
(6) Either the prosecutor or the court may ask the subject of the polygraph examination to be examined by a polygraph operator of the court’s choice or such operator may be asked to review the offered data with the original operator, or both.
(7) The test results shall be considered only with regard to the general credibility of the examinee not as to the truth or falsehood of any particular statement.
(8) The affidavits or testimony of the polygraph operator shall be a separate record and shall not be used in any way at any subsequent trial.
(9) A judge granting a new trial on the basis of polygraph tests shall not thereafter act as a trier of fact in that case but may preside with a jury. A substitute judge as trier of fact shall not be privy to the polygraph examination or results, or to the fact that a polygraph examination was taken or was in any way involved.
By permitting such limited use of the polygraph, we are operating well within the limits prescribed by the state of the art. Further, we avoid prematurely considering those policy questions . which inevitably accompany use of such evidence at trial.
The advantage of permitting the polygraph to be used for such limited purposes is two-fold:
(1) It gives us an opportunity to use this potentially useful device under strictly controlled conditions, thereby permitting it to establish a "track record” in our state and help provide some of the missing data.
(2) All the polygraph does in this context is, in *414effect, help the judge decide whether the evidence is good enough to go to the jury. Thus it provides individuals some assistance in arguing their case, and it gives a judge some assistance in making a decision, without casting the polygraph in a decisive role.
IX — Conclusion
The attraction of the polygraph is obvious. The issue of credibility is central to any legal dispute. A device which purports to determine once and for all who is telling the truth and who is lying would be of tremendous service to our system of justice.
The fact is, we don’t know the extent of the polygraph’s usefulness. If we totally foreclose admissibility, we may never know. On the other hand, experimentation would be unwise when an individual’s life or freedom might hinge directly on the effect at trial of a questionable device.
Admission of polygraph evidence for purposes of assisting the judge to determine a post-conviction motion is a different matter. The orderly trial process has already been concluded, and the finder of fact has rendered its judgment within a constitutionally structured adversarial system.
Post-conviction hearings are handled within different constraints from trials. The judge, for example, may use certain data which would be inadmissible at trial to assist in rendering decisions. Yet precisely because the adversarial process has been completed and has done its job, the post-conviction stage of judicial proceedings can be conducted under slightly different rules.
Therefore, admission of the polygraph at a post-conviction hearing for a new trial would not set a pattern for admissibility at trial, but it would *415provide an opportunity to test effectiveness of the polygraph while at the same time adding another tool to assist the judge.
The need for such assistance is particularly evident in the type of situation presented by the instant case. Traditionally, the testimony of recanting or suddenly discovered witnesses has been highly suspect, largely because it is impossible to determine when the truth is being told. The polygraph won’t do this either; not even its most ardent proponents would so contend. But it might help.
Further, the decision of the judge would be followed by a trial at which the finder of fact would have an opportunity to render a decision totally independent of any influence of the polygraph test.
Thus, in a post-conviction motion for a new trial, the polygraph may be observed in as close to an experimental status as is possible within the actual judicial process, the only way apparently it will be possible to truly gauge its worth. We hold, therefore, that polygraph results may be considered, within the discretion of the judge, to enable a decision to be reached at a post-conviction hearing for a new trial.
Certain safeguards must, of course, be present.
Validity of the machine, the test and testing condition, and competence of the examiner must be established.
The polygraph tests must be presented on behalf of thé defendant so that a constitutionally satisfactory waiver is established. As a new trial may be awarded only upon presentation of genuinely new evidence meeting certain strict standards, supra, the polygraph results must be used to support the credibility of new witnesses or to otherwise sup*416port new evidence which the finder of fact in the previous trial has not already reviewed. Thus, a polygraph test purporting to demonstrate that the defendant whose story the finder of fact has already rejected is actually telling the truth would not satisfy the conditions. This is as it should be, for the polygraph is not a truth machine, and the collective wisdom of the jury has already passed on defendant’s story.
The trial judge may also appoint the court’s own expert, or the prosecutor may request an expert to conduct another examination, or to merely review the results of the test already taken.
Results of the polygraph itself may not be admitted at the new trial and the judge so granting a new trial may not become a finder of fact at that trial.
In this case, the trial judge assumed he had no discretion to consider the polygraphic evidence, even if he was satisfied with the reliability and proficiency of the polygraph machine and the polygraph operator. This was error.
As we have just indicated, we hold that polygraph evidence properly qualified may be considered, within the discretion of the judge, to enable a decision to be reached at a post-conviction hearing for a new trial.
We therefore remand this matter to the trial court to permit the judge to exercise his discretion to consider whether to admit testimony about polygraph results in a post-conviction hearing on a motion for new trial, provided that he finds a proper foundation has been laid for such testimony as outlined in this opinion.
Kavanagh, C. J., and Levin, J., concurred with Williams, J. *417Fitzgerald, Ryan, and Blair Moody, Jr., JJ., took no part in the decision of this case.Appendix A
Relevant portions of the testimony of Lynn Marcy concerning the polygraph examinations given in the instant case are summarized below to illustrate appropriate procedures for utilizing polygraph evidence in a motion for a new trial:
The polygraph examiner testified to the questions asked, answers given, and his opinion as to truthfulness or deception on each. The procedure used was a controlled question technique using three types of questions.
One type is the irrelevant question to which the answer is obvious and which the examiner knows in advance. It is used for purposes of comparison. An example is, "Are you now in Michigan?”
The second type, a question of a broad general nature, is the control question, or the "known lie”. The subject is either presumed to be lying or it could be presumed from the nature of the question, that he could not be sure he was answering it truthfully. Marcy explained its purpose is for diagnostic comparison basis and said it also helped the examiner achieve a psychological set during the examination procedure. It enables the examiner to see what a reading of deception looks like for a particular person. An example is, "Did you ever steal anything of value in your life?”
The third type is the relevant or issue question, to which Marcy testified the attention of the individual attempting deception will be directed because he or she finds that area of greatest consequence. The person who believes he or she is answering the relevant or issue questions correctly will direct his or her attention and psychological *418set toward the control or lie questions. The polygraph operator's diagnosis is based on whether there is greater emotional stimulation on the control, or on the relevant issue questions.
The instrument used in these procedures was a Keeler Polygraph, a four-stylus machine recording upper and lower respiratory movement changes, galvanic skin responses, and cardiovascular response changes.
In order to prepare the instant examinations, Marcy testified to spending time reviewing the facts of the case, and then had a pretest interview with Barbara in which he took a history and discussed with Barbara his background and physical condition. In addition, Marcy attempted to ensure, during the pretest conference, that Barbara and he mutually understood the particular wordings of questions. Another purpose of the review was to remove the element of surprise which might distort results. The tests of Barbara occurred on two successive days, one involving primarily questioning on the rape charge, the other, that of extortion.
Marcy testified that the facts and circumstances of the instant case would be the proper subject of the polygraph examination. He further specified that such an examination is a test of whether, on the specific questions asked and reported, the subject is telling what he or she believes to be the truth. He also testified:
"With the conditions, as outlined in the record that has been made a part of this proceeding, and a qualified examiner with a suitable subject, proper equipment and the use of proper techniques, the accuracy that has been found by examiners in the field, including myself, would be something on the plus side of 99 percent.”
*419Marcy concluded that Barbara answered "each of the listed relevant questions truthfully * * * to the best of his knowledge and belief’.
Marcy reviewed the results of Barbara’s polygraph examination with two polygraph examiners for the Michigan State Police, and did the same for that of Metropoulos. The only change in procedure with Metropoulos was an interview with the interpreter, briefing him on the polygraph procedures which would facilitate working with Marcy, and working with him on question structure so that they could be clearly conveyed in Greek. Marcy also attempted to assure the court that it was not unusual to conduct polygraph examinations through an interpreter. He concluded that Metropoulos told the truth on the relevant questions in the examination.
Appendix B
How the Polygraph Works
The polygraph has matured since the systolic blood measurement device in Frye was first used. Today’s machines measure several changes, pal-mar skin resistance or galvanic skin response (GSR), Lykken, Psychology and the Lie Detector Industry, 29 Am Psychologist 725, 729 (1974), abdominal respiration, thoracic respiration, and blood pressure-pulse rate. Horvath & Reid, The Reliability of Polygraph Examiner Diagnosis of Truth and Deception, 62 J Crim L C & P S 276, 277 (1971). Others may add a Psychological Stress Evaluator that records the presence of a low-frequency component which diminishes in speed for sounds under stress. Lykken, supra, 29 Am Psychologist 729. Most have a device for recording muscular activity, added when it was discovered *420the examinee could affect the accuracy of the diagnosis by tensing or relaxing muscles unobserved by the operator. Note, The Polygraphic Technique: A Selective Analysis, 20 Drake L Rev 330, 331 (1971).
The theory is that a suspect’s awareness that he or she is lying, combined with concern over being caught in a lie, create emotional disturbances which are transformed into physiological changes which can be detected by the polygraph. Reid & Inbau, Truth and Deception: The Polygraph ("LieDetector”) Technique, pp 50, 196.1
To risk oversimplification, following a pre-test interview, the examiner generally administers a specific response test, including relevant, irrelevant, and control questions.2 A peak-of-tension test, using some critical detail of the alleged offense which the suspect does not know has been discovered, is rarely used. Note, supra, 20 Drake L Rev 331.
Inbau & Reid maintain the accuracy rate is 95 percent, with a 4 percent margin of inconclusive results and a 1 percent margin of possible error. They also maintain that the errors tend to be a failure to detect the lies of guilty subjects, rather than a finding of false-positives. Note, supra, 20 Drake L Rev 334, citing Inbau & Reid, Lie Detection and Criminal Interrogation (Baltimore: Williams & Wilkins Co, 3d ed, 1953), pp 110-111, and *421Inbau & Reid, The Lie Detector Technique: A Reliable and Valuable Investigative Aid, 50 ABA J 470, 472 (1964). In the case at bar, the examiner testified his test was 99 percent accurate. This is consistent with the results reported on the special record. Such results, however, have not been universally accepted. See, eg., United States v Wilson, 361 F Supp 510, 512-513 (D Md, 1973), reporting the accuracy range from 70 percent to 95 percent.
See Appendix A for summary.
The polygraph test of defendant Barbara, because it serves only to in effect duplicate evidence the jury has already heard, i.e., Barbara’s denial of his guilt, is merely cumulative, and does not properly fall within the purview of new evidence. The same reasoning would apply had Metropoulos testified. It is not clear whether Spadafore’s testimony is new or had been given at another unrelated trial in Federal Court.
This, of course, does not preclude questions raised concerning admissibility of an otherwise accepted technique in a particular trial, such as adequacy of the given test and of the particular examiner.
Bailey also took the stand himself, basing his expertise on his experience as an investigator in the Marine Corps G-2 section, as a private investigator who referred cases to polygraph examiners, and as an attorney who uses the polygraph in investigative work, who lectures extensively about the polygraph, and who has done research with the polygraph, especially investigations in the field of hypnosis with Marcy. Bailey is a former associate member of the American Polygraph Association, and has lectured at the Keeler Polygraph Institute and the National Training Center of Lie Detection (now the Baxter School of Lie Detection).
There was no showing of the nature of this follow-up, i.e., if there were convictions or corroboration by extrinsic evidence, for example.
He explained these circumstances as occurring when a subject confessed after he had determined the test showed deception. This, he maintained, was not inconsistent with the conclusion reached.
Although a licensed psychological examiner, Dr. Yankee’s Ph.D. is in higher education. His bachelor’s and master’s degrees are in psychology.
Since the problem of potential conflict between two experts is so central to the ultimate question, Bailey’s observations are useful:
"Mr. Goussy [Assistant Attorney General]: That report indicates that Mr. Zimmerman passed Peter Lazaros in relation to the affair that occurred in the Diplomat Towers in Florida.
"Mr. Bailey: Right. In part he passed it.
"Mr. Goussy: In part.
*375"Mr. Bailey: Yes.
"Mr. Goussy: I have personal knowledge of a polygraph of a highly respected person that it failed Peter Lazaros.
"Mr. Bailey: I am aware of that.
"Mr. Goussy: It seems to me this very situation in this case indicates that there can be serious conflicts among polygraph experts as to a given incident.
"Mr. Bailey: No one made an eifort to reconcile these two tests. If the examiners were to sit down and go over their questions together, the information they had at the time they conducted the examination and compared their charts, I am quite satisfied that you would not wind up with a conflict.
"On two examiners who had not had any contact with each other, run the same subject and it is very possible Mr. Lazaros withheld information from the examiner, which didn’t mean he was lying but which he would more readily disclose to counsel, which is a critical part of a clearing test that could well explain it. Mr. Lazaros did show responses which merely fortify the result because it is a control, it shows responses to one aspect of that transaction.”
Amicus briefs were submitted by the American Polygraph Association, the State Appellate Defenders Office (also adopted by the Michigan Association of Criminal Defense Lawyers), the American Civil Liberties Union of Michigan, and the Prosecuting Attorneys Appellate Service.
The most significant departure until recently from this solid wall of non-admissibility was People v Kenny, 167 NY Misc 51; 3 NYS2d 348 (1938), where the results of a pathometer test administered to defendant were admitted over the prosecutor’s objection. However, Kenny was probably overruled sub silentio the same year by People v Forte, 279 NY 204; 18 NE2d 31 (1938), when the New York Court of Appeals affirmed the trial judge’s exclusion of lie-detector test results. The only other substantial situation when polygraph evidence has been admitted in some states has been when the parties have stipulated to the admission of the testimony. See discussion, infra.
Wigmore cites cases prohibiting admission from federal courts including the tenth circuit, Marks v United States, 260 F2d 377 (CA 10, 1958); eighth circuit, McCroskey v United States, 339 F2d 895 (CA 8, 1965), and from states including Arizona, California, Florida, Hawaii, Illinois, Iowa, Kansas, Kentucky, Maine, Minnesota, Missouri, Montana, Nebraska, New Hampshire, New Mexico, New York, North Carolina, North Dakota, Ohio, Oklahoma, Pennsylvania, South Carolina, Tennessee, Texas, Virginia and Wisconsin.
The complete Valdez requirements are:
(1) That all parties sign a written stipulation;
(2) The judge may still refuse to admit the test, if, for example, the examiner is not qualified or if the test conditions were faulty;
(3) The examiner may be cross-examined as to his or her qualifica*379tions and training, the test conditions, limitations of the polygraph technique, and any other matter deemed relevant by the trial judge;
(4) The jury shall be instructed as to limited use of polygraph evidence to indicate defendant’s veracity at the time of the examination, and not to prove any element of the crime charged, and that corroborative weight and effect of the testimony is for them to determine. 91 Ariz 283-284; 371 P2d 900-901.
This includes, necessarily, the Davis standard and those of other states which derive from Frye.
" 'General scientific acceptance’ is a proper condition for taking *380judicial notice of scientific facts, but not a criterion for the admissibility of scientific evidence. Any relevant conclusions which are supported by a qualified expert witness should be received unless there are other reasons for exclusion.” McCormick, Evidence (2d ed), § 203, p 491.
"Even if judicial notice is not taken of the validity of a general proposition of science, that validity may still be established by an appropriate evidentiary showing by expert testimony.” Strong, Questions Affecting the Admissibility of Scientific Evidence, 1970 U Ill L Forum 1, 9.
McCormick would apply the same standard for all scientific evidence, that it should be admissible if it would "render the desired inference more probable than it would be without the evidence”. McCormick, Evidence (1st ed), § 152, p 318.
Polygraph adherents maintain that scientific evidence such as psychiatry, which they maintain is even less verifiable than that of the polygraph, is readily admissible without such a test. Reid & Inbau, Truth and Deception: The Polygraph ("Lie-Detector") Technique (Baltimore: Williams & Wilkins Co, 1966), pp 255, 257.
It is suggested that scientific disagreement should affect only the weight and not the admissibility of evidence. See, e.g., how courts handled disputed status of Breathalyzer. State v Olivas, 77 Ariz 118; 267 P2d 893 (1954); People v Bobczyk, 343 Ill App 504; 99 NE2d 567 (1951).
One Federal district court recommended still another approach:
"[A] polygraph examiner, having satisfied a particular court that he is qualified as an expert in his field, should be permitted to present foundational evidence to that court demonstrative of the polygraph’s substantial reliability and acceptance, in an effort to establish its probative value.” United States v DeBetham, 348 F Supp 1377, 1384 (SD Cal, 1972) (emphasis added).
See Coppolino v State, 223 So 2d 68, 70-71 (Fla App, 1968), cert den, 399 US 927; 90 S Ct 2242; 26 L Ed 2d 794 (1970), where a test showing death resulted from succinylcholine chloride was admitted, even though the test was designed specially for that case and therefore was not generally accepted. The court held the lack of general acceptance affected only the weight and not the admissibility of the evidence. The court further supported admissibility as within the wide discretion exercised by the trial judge, noting that the prosecution witnesses stated they held their conclusions "with reasonable medical centainty”, that unlike lie detector and intoxication tests, *381"there is a dearth of literature and specific case law to guide the trial and appellate courts”, and that defendant did not show the court abused its discretion. Compare Hodo v Superior Court of Riverside County, 30 Cal App 3d 778; 106 Cal Rptr 547, 553 (1973), applying Frye standard to voiceprints, and Worley v State, 263 So 2d 613 (Fla App, 1972), which does not, cited in Tarlow, Admissibility of Polygraph Evidence in 1975: An Aid in Determining Credibility in a Perjury-Plagued System, 26 Hastings LJ 917, 939, fn 107 (1975).
"[W]here judicial notice is not appropriate, due to substantial scientific disagreement, the courts [should] apply the more normal standard of admissibility, which simply weighs the probative value of the proferred evidence against the traditional policy factors, such as the possibilities of misleading or prejudicing the jury, or of unduly consuming the court’s time.” 348 F Supp 1383-1384, citing Kaplan, The Lie Detector: An Analysis of Its Place in the Law of Evidence, 10 Wayne L Rev 381, 402-407 (1964).
"This is not to say that the polygraph may escape all requirements of demonstrating its reliability, since that is an integral part of whatever probative value the technique may possess. Nor is this to say that in applying the normal standard of admissibility of scientific evidence the polygraph need not demonstrate the support of at least 'a substantial body of scientific opinion.’ ” 348 F Supp 1384.
There was no indication whether the lone psychologist testifying was also involved in the polygraph field, but he presented a validation test reporting a percentage of accuracy better than 80 percent.
Issues discussed included the effect of the polygraph on self-incrimination, possible violation of the right to privacy, and danger of displacing the jury. 348 F Supp 1389-1391. These are referred to in somewhat greater detail in part VII of this opinion, infra.
The record of exhibits apparently included a bibliography of scientific studies, 350 F Supp 688, fn 15, and studies performed by David Raskin, a psychologist who performed research in the areas of psychophysiology.
As acknowledged in the special record, polygraph techniques and devices have remained largely the same since 1956, and we have been presented with no indication by proponents of the original special record that matters have changed since then. Thus, it does not appear that conclusions and studies of recent years need be relied upon.
Apparently, based on John E. Reid’s testimony in the special record, Reid and Inbau’s celebrated change in position to urging admissibility of the polygraph stems from formation of the professional association and its attempts at self-policing, and not from any new developments in the area or improvements in technique or instrumentation.
The text also reports that "the preponderance of evidence seems to indicate * * * the technique is effective — at least more effective than alternative methods of ascertaining truth — it still remains to be determined how accurately and under what conditions.” Greenfield & Sternbach, p 751.
In a previous publication, one of the text’s authors concluded, " * * * there exists no public body of knowledge to support the enthusiastic claim of operators”. Sternbach, Gustafson & Colier, Don’t Trust the Lie Detector, 40 Harvard Bus Rev 127, 130 (1962), quoted in Burkey, The Case Against the Polygraph, 51 ABA J 855, 856 (1965).
See Appendix B for brief description of how the polygraph works.
The special record in this case indicated that most cases are never verified, and the record did not clarify the method used on those which were verified.
One possibility might well be a Law Enforcement Aid Administration-funded study which would be designed, not just to compile statistics, but to verify field operation of the polygraph. Mere statistical studies are not sufficient, for their weakness is again usually lack of verification.
There may, however, be some extrinsic verification from police *396investigation. For example, the fifth column in the chart below, "Reported Guilty and Confessed” provides some measure of extrinsic support:
PERIODIC REPORTS FROM POLICE RE: POLYGRAPH EXAMINATIONS (Total Subjects: 7622)
Trovillo, Scientíñc Proof of Credibility, 22 Tenn L Rev 743, 758 (1953). (Data was computed by writing letters to about 50 police departments in principal United States cities.) Trovillo, 22 Tenn L Rev 759. The author acknowledges that "some of the most important information as to validity or reliability is missing”. However, this chart is an interesting demonstration of the usefulness of the polygraph as an investigative tool. Kg., the Michigan State Police reported 50 percent "Reported Guilty and Confessed” out of a reported 41.7 percent guilty.
The examiner in this case reported he was able to verify only approximately 30 percent of the examinations he has made. Without giving more details, he reported that of these, "something better than 95 percent” were consistent with later developments.
Such claims, made in the context of the difficulties of evaluating the polygraph, must also be weighed against such competing conten*397tions that "the degree of success is close to 70 percent”, Burkey, supra, 51 ABA J 856, and that "a leading polygraph firm claiming 95% accuracy could verify only 18.9% of its judgments”, Note, Evolving Methods of Scientific Proof, 13 NY L Forum 679, 690, fn 12 (1968), citing Hearing Before the Subcommittee of the Committee on Government Operations of the House of Representatives [on the Use of the Polygraph] by the Federal Government, 88th Cong, 2d Sess, pt 1, p 32 (1964).
Reid & Inbau, however, report that the percentage of known errors with the technique in the laboratories of John E. Reid & Associates is
"less than 1 percent. Of the remainder no diagnosis at all is attempted in about 5 percent of the cases, because of such factors as the physiological or psychological impairment of the subjects.” Reid & Inbau, Truth and Deception, p 234.
Other proponents report accuracy ranging from 75 percent to 96 percent. 13 NY L Forum, 690, fn 12.
One problem involved in evaluating the statements of polygraphers is that these would involve laboratory studies. The difficulties of comparing field and laboratory studies are closely bound up with the problem of motivation: in the field, the suspect’s concern about the test and the consequences of a positive evaluation are obvious. Greenfield & Sternbach, p 752. While many will not completely subscribe to the judgment that "[laboratory studies cannot provide adequate validity estimates”, Lykken, Psychology and the Lie Detector Industry, 29 Am Psychologist 725, 734 (1974) (emphasis in original), the obvious problems with comparability make the validation question even more urgent.
Obviously, in addition to the difficulties of developing extrinsic evidence to support the polygraph conclusion, other problems of verification in the field include the probability that "false-negatives” will not challenge the results, and that a jury verdict, although legally conclusive, is not a scientifically controlled means of verification.
E.g., "The process of reading the charts involves considerable interpretation, and the possibility that the examiner’s biases may affect his conclusions cannot be discarded.” United States v Urquidez, 356 F Supp 1363, 1367 (CD Cal, 1973). Fred Inbau, one of the most active proponents of the polygraph, reported in 1964 that only 20 percent of those calling themselves polygraph operators at the time were actually qualified. Note, The Emergence of the Polygraph at Trial, 73 Colum L Rev 1120, 1124, fn 23 (1973). A classic article by an *399early polygraph pioneer, Leonarde Keeler, concluded by emphasizing "the importance of having only well trained persons as operators must be stressed. This approach to criminal investigation can prove dangerous in the hands of the untrained.” Keeler, A Method for Detecting Deception, 1 Am J Police Sci 38, 51 (1930), quoted in 3A Wigmore, Evidence (Chadbourn rev), § 999, p 953.
In a study based only on comparison of graphs, experienced examiners were correct 91.4 percent of the time. The inexperienced succeeded in only 79.1 percent of the cases.
E.g., a "false positive” may be obtained due to undetected irregular blood pressure, heart disease, respiratory disorders, extreme nervousness, anxiety, anger, or nervous tension. Highleyman, The Deceptive Certainty of the "Lie Detector,” IQ Hastings LJ 47, 60-61 (1958). It is maintained however, that examinations tend to err more on the side of finding a guilty subject innocent, than finding an innocent subject guilty of lying. Horvath & Reid, The Reliability of Polygraph Examiner Diagnosis of Truth and Deception, 62 J Crim L C & P S 276, 279 (1971).
Some polygraphs have devices which are designed to detect such muscular exertions.
In 1964, Professor Inbau stated that "eighty percent of people who administer tests do not measure up to standards we feel are required.” Note, supra, 13 NY L Forum 692. Two years later, Reid & Inbau suggested, in Truth and Deception, p 257:
"(1) That the examiner possess a college degree. (2) That he has received at least six months of internship training under an experienced, competent examiner or examiners with a sufficient volume of case work to afford frequent supervised testing in actual case situations. (3) That the witness have at least five years’ experience as a specialist in the field of polygraph examinations. (4) That the examin*401er’s testimony be based upon polygraph records that he produces in court and which are available for cross-examination purposes.”
See, e.g., Forensic Polygraph Examiners Act, MCLA 338.1701 et seq.; MSA 18.186(1). Note, however, that the Attorney General has ruled that licensing is not required where a psychological stress evaluator is used for purposes other than to detect deception or verify truthfulness. OAG, 1975-1976, No 4855 (April 23, 1975).
See discussion, Trovillo, supra, 22 Tenn L Rev 763-765, suggesting minimum academic training desirable but not now required.
He attempted to assure the appointment of qualified examiners by utilizing the Federal Court’s power to appoint experts to verify defendant’s examination. 350 F Supp 96-97.
According to a recent law review article, prosecutors in several jurisdictions engage in pretrial agreements to withdraw prosecutions if defendant "passes” the polygraph. If defendant "fails,” the agreement usually includes a pledge to either plead to a lesser charge or to not object to the admission of the test at trial. Note, The Emergence of the Polygraph at Trial, 73 Colum L Rev 1120, 1127, fns 41-42 (1973). See People v Reagan, 395 Mich 306, 308-311; 235 NW2d 581 (1975); Butler v State, 228 So 2d 421, 422-424 (Fla App, 1969), requiring such bargains to be kept.
However, the Oakland County prosecutor supplemented the record in the instant case by advising this Court that as of January 1975, the office did not retain a polygrapher on a regular basis; polygraph results were used at the initial complaint and warrant-issuing stage, and not in the information-issuing process; however, at those initial stages, polygraph results are used in substantially less than one per cent of the cases. The tests are administered to both complainant and respondent.
See also, Hearing Before the Subcommittee of the House Committee on Government Operations on Use of Polygraphs as "Lie Detectors” by the Federal Government, 88th Cong, 2d Sess (1964), cited in Note, The Emergence of the Polygraph at Trial, 73 Colum L Rev 1120, fn 3 (1973); Richardson, Modern Scientific Evidence (2d ed), § 10.2, p 309, fn 6.
The concern persists that the possibility of the jury treating polygraph evidence as conclusive proof of guilt or innocence may give such evidence an overbearing prejudicial effect. People v Davis, 343 Mich 348; 72 NW2d 269 (1955). Judge Joiner in Ridling resolved the problem by maintaining faith in the intelligence of the modern jury. Judge Kaufman, on the other hand, in United States v Stromberg, 179 F Supp 278, 280 (SD NY, 1959), felt admissibility of such evidence would be tantamount to overturning the jury system.
"The most important function served by a jury is in bringing its accumulated experience to bear upon witnesses testifying before it, in order to distinguish truth from falsity. * * * I am not prepared to rule that the jury system is as yet outmoded.”
See also, State v Smith, 113 Ohio App 461, 463-464; 178 NE2d 605, 607 (1960). McCormick wrote: "The exclusion seems to rest more upon a judicial estimate of the weight that the trier of fact will give to the opinion, and a demand that the opinion be almost infallible because the trier will think it so.” Evidence (2d ed), § 207, p 507. However, he also said, "We cannot in our hearts be so confident of the reliability of *405the present system of resolving conflicts in testimony by impeachment, cross-examination and inferences from demeanor, that we can afford to reject scientific aid in the task.” McCormick (1st ed), § 174, pp 369-370.
Part of the reason for this problem no doubt lies with the myth of infallibility, considered essential by polygraph examiners to create the appropriate mental set in the tested subject.
The myth persists despite the fact that proponents have never claimed that a detector will ring bells or set off alarms automatically when a subject lies. Skolnick, Scientifíc Theory and Scientifíc Evidence: An Analysis of Lie-Detection, 70 Yale LJ 694, 704-705 (1961).
Examples are offered, moreover, of the jury disregarding the polygraph, the classic example occurring in Commonwealth v George O Edgerly (Docket No 59459) (Mass, Middlesex Superior Court, 1961), where the defendant was acquitted in spite of the admission of adverse polygraph test results. Note, Problems Remaining for the "Generally Accepted" Polygraph, 53 B U L Rev 375, 387 (1973).
The jurors in People v Kenny, 167 NY Misc 51; 3 NYS2d 348 (1938), in which polygraph evidence was admitted, found the defendant innocent. In response to the question, "was the 'lie detector’ testimony, in your opinion, conclusive proof of the innocence or guilt of Kenny?” six jurors responded affirmatively. Four said no, and two did not respond. Forkosch, The Lie Detector and the Courts, 16 NY U L Q Rev 202, 228-229 (1939), quoted in People v Schiers, 160 Cal App 2d 364, 379, 380, fn 2; 324 P2d 981, 329 P2d 1, 2, fn 2 (1958) (dissent).
"[T]he jury of twelve men is designed to evaluate * * * testimony based on common sense and ordinary experience. This historic function should not be lightly exposed to drastic revision. Evaluation of the testimony of polygraph experts is indeed beyond the realm of ordinary experience.” United States v Wilson, 361 F Supp 510, 513 (D Md, 1973).
Ridling, however, suggested that defendant’s offer to testify waives this right, and further, that polygraph results may not be testimonial. 350 F Supp 97. Contra, Bowen v Eyman, 324 F Supp 339, 341 (D Ariz, 1970). It has been eloquently argued that use of polygraph evidence violates these policy foundations of the privilege against self-incrimination:
"1. To stimulate law enforcement agencies to discover independent evidence to uncover offenders rather than resort to the inquisitorial examination of many persons.
*407"2. To keep individuals free from police investigation and meddling unless there is probable cause to believe one has broken the law.
"3. To preclude inquisitions into a person’s private life by local 'tyrants’ for the purpose of 'ensnarement.’ * * *
"4. Dependence upon confession at trials.
"5. Assuming the detector becomes 100 percent accurate and its results are admissible, the jury will become vestigial and the defendant will no longer be a 'party’ since he cannot make a realistic defense. The real case will be decided not in open court but in the laboratory.” People v Schiers, 160 Cal App 2d 364, 379, 383; 324 P2d 981, 329 P2d 1, 4 (1958) (dissent) (citations omitted).
While polygraphers unanimously agree that a coerced test is invalid, the question may not be academic because of subtle means of coercion, or because of the development of new techniques. Note, Pinocchio’s New Nose, 48 NY U L Rev 339, 351 (1973). See, e.g. Pinter v State, 203 Miss 344, 347; 34 So 2d 723, 724 (1948), admitting confession secured by a threat to require a defendant to submit to a polygraph test because, "It would seem that his fear was not of the machine but of its capacity to elicit truth. * * * A desire to anticipate by voluntary disclosure, the supposed revelations of a Tie detector’ has its origin in the mind and conscience of the defendant, and is not an 'undue influence’ ”. Approved, Richardson, Modern Scientific Evidence (2d ed), § 10.8, p 318.
Although referring to the possibility of a false positive, Justice Carter, writing for the dissenters of the District Court of Appeal of California, observed, "An innocent man, convinced of the device’s accuracy, would anticipate a favorable result, not an involuntary admission of guilt. He cannot be held to have waived his privilege against an untruth.” People v Schiers, supra, 160 Cal App 2d 364, 384; 329 P2d 1, 4 (dissent).
"Judgment of Bundesgerichtshof (I. Strafsenat), Feb. 16, 1954, 5 Entscheidungen des Bundesgerichtshofes in Strafsachen [hereinafter B.G.H. St.] 332. The Bundesgerichtshof is the highest court of the Bonn Republic in civil and criminal matters.” Silving, supra, 69 Harv L Rev 688, fn 22.
Mr. Dogin noted that the Criminal Division finds such examinations useful in investigative work in only a few limited situations, particularly in "screening members of 'closed’ groups with access to property that was stolen or embezzled, as an informant screening device to prevent unnecessary investigative efforts and to test informant credibility.”
Additional policy considerations raised against admission of polygraph examination results include these: "(1) that the jury will put inordinate weight on the testimony of the examiner; (2) that, although the testimony of a ballistics or fingerprint expert is usually significant, 'a demonstrated lie or truth to the question "Did you kill XT’ would invariably be perfectly conclusive’; (3) that examination results offered unilaterally by the defendant would always be favorable to him and his opponent would not be able to examine him under the lie-detector, nor to cross-examine him should he refuse to take the stand; (4) that, even where the defendant consents to take an examination, while incarcerated the conditions under which the consent is obtained and the examination is given may be a factor in producing *410results unfavorable to the defendant; (5) that the presumption of innocence may be threatened by the inordinate weight likely to be given to failing to take a polygraph examination once such examination results are generally admitted * * * ”, Note, Problems Remaining for the "Generally Accepted” Polygraph, 53 B U L Rev 375, 377 (1973).
The prediction that admission of the polygraph would lead to a confusing battle of the experts was dealt with by one of the earliest courts admitting polygraph evidence. Noting that "the deductions of handwriting experts and of psychiatrists are not at all uniform * * * their testimony is received in evidence, and it is left to a jury to determine which, if either, expert or experts they are going to believe and accept”. People v Kenny, 167 Misc 51; 3 NYS2d 348, 351 (1938).
This was a concern of the trial judge who sat during the making of the special record.
This objection completely misconstrues the nature of polygraph testimony which, as we have seen, is really the opinion of the polygraph expert. The expert can, of course, be cross-examined. Note, supra, 48 NY U L Rev 350. However, our own Court has noted there is a possibility that the credibility of the examiners rather than the credibility of witnesses would become the issue at trial. People v Davis, 343 Mich 348, 372; 72 NW2d 269 (1955).
Some other states which have considered the issue have done so only in the context of whether the polygraph evidence itself would be admissible at a new trial. However, our decision today is based precisely on the difference between a post-conviction motion for a new trial and the trial itself. If the only new evidence is the polygraph examination, this of course would be inadmissible at trial and would not be a sufficient basis for granting a new trial. Thus, in State v Brown, 177 So 2d 532 (Fla App, 1965), the court was held to have appropriately exercised its discretion to grant a new trial on the basis of stipulated-to polygraph tests given to defendant to confirm his alibi defense. Defendant passed two out of three tests. 177 So 2d 532. The court, however, relied on Florida’s general rule of admitting polygraph tests when both parties so stipulate. 177 So 2d 533.
In contrast, in United States v Stromberg, 179 F Supp 278, 280 (SD NY, 1959), the Court rejected the use of polygraph evidence on a motion for a new trial, because the evidence would not have been admissible at trial. The results of the polygraph simply supported defendant’s story that he had not been involved in the conspiracy of which he had been convicted, and did not relate to new witnesses or other new evidence. To have made a difference in the outcome of a trial, the polygraph test would itself have to be admitted into evidence. This, the trial court was not prepared to do. Accord, State v Scott, 210 Kan 426, 434; 502 P2d 753, 760 (1972).
The main features of the apparatus and the theory behind it are well expressed in 3A Wigmore, Evidence (Chadbourn rev), § 999, pp 949-953.
A sample test might include this pattern of questions:
"(1) Are you called John? (2) Have you had a smoke since you entered this room? (3) Do you know who killed Frank Kieson last night? (4) Have you eaten anything today? (5) Did you kill Frank Kieson last night? (6) Do you know where the killer went after he left Kieson’s house? (7) Do you know what implement was used to kill Kieson? (8) Have you lied to any of my questions?” Trovillo, Scientific Proof of Credibility, 22 Tenn L Rev 743, 753 (1953).