P.S. Docket No. 7/141
May 21, 1980
In the Matter of the Complaint Against
TELESTAR, INC.,
TELESTAR INCORPORATED,
TELSTAR INCORPORATED
200 S. Front Street at
Wormleysburg, PA 17043
P.S. Docket No. 7/141;
Duvall, William A.
APPEARANCE FOR COMPLAINANT:
H. R ichard Hefner, Esq.
Law Department
United States Postal Service
Washington, D.C. 20260
APPEARANCE FOR RESPONDENT:
Jeffrey A. Rabin, Esq.
Jacob R. Evseroff, Esq.
186 Joralemon Street
Brooklyn, NY 11201
INITIAL DECISION 1/
This proceeding was initiated on January 28, 1980, when the Consumer Protection Division, Law Department United States Postal Service, filed a Complaint in which it is alleged that the
Respondent, under the names Telestar, Inc., Telestar, Incorporated, and Telstar, Incorporated, is engaged in conducting a scheme or device for obtaining money through the mails by means of false representations in violation of 39 U. S. Code, Section 3005.
Although the Complaint sets forth the name of the Respondent as Telstar, Incorporated, no exhibits on which that name appears in that form were received in evidence. I am advised that such an exhibit is available and can be, if necessary, offered into evidence, but in view of the provisions of § 952.10(b) of the Rules of Practice, it is not necessary that that name appear on an exhibit received in evidence. The cited rule provides that in the event there is an allegation of fact in a complaint, which fact is not specifically admitted or denied in the complaint, it may be deemed to be admitted, and no further proof with respect to that matter need be offered or received in evidence. The Respondent, not having denied that it uses the name of Telstar, Incorporated, that allegation is deemed to be admitted.
The Complainant alleges that public attention is attracted to the Respondent's scheme by means of advertisements which are calculated and intended to induce readers thereof to remit money through the mail to Respondent.
It is further alleged that certain materials attached to the Complaint as Exhibits A and B (Appendices A and B hereto) are copies of the advertisements referred to in the foregoing paragraph.
Complainant alleges in Paragraph III of the Complaint that by means of the aforementioned advertisements, Respondent expressly or impliedly represents to the public, in substance and effect, that:
(a) The TRUTH MACHINE is a certain means for the detection of oral falsity.
(b) The TRUTH MACHINE will enable the user to obtain correct answers in the face of conscious falsehood.
(c) "THE TRUTH MACHINE is a new generation voice stress analyzer that flawlessly pinpoints deception."
(d) The TRUTH MACHINE may be employed successfully as a certain means for the detection of falsity without the knowledge of the person whose veracity allegedly is being monitored.
(e) The TRUTH MACHINE may be relied upon to supply correct information upon which the user may safely rely in the conduct of his daily affairs.
(f) The TRUTH MACHINE is an effective "voice stress analyzer."
(g) The TRUTH MACHINE is an accurate means of detecting and registering voice "'microtremors'".
(h) A "few hours" or a "weekend" is sufficient time to enable a purchaser to acquire the experience and knowledge necessary to employ the Truth Machine as a certain means of detecting falsity.
(i) "...if anything goes wrong with the Truth Machine Microtronics will repair it free of charge... ."
Finally, Complainant alleges that the aforesaid representations are materially false as a matter of fact.
In the Answer to the Complaint, Respondent:
1. Denies that any scheme was entered into to obtain remittances of money through the mails by means of false representations.
2. Admits that Exhibits A and B are its advertisements, but denies that such are part of a scheme containing false repre- sentations.
3. Denies that the representations set forth in Paragraph III of the Complaint are materially false as a matter of fact.
Use of the Mails
Appendices A and B hereto present order blanks by which remittances of money through the mail are sought for a "Truth Machine" (Appendix A) and a "Microtronics 'Truth Machine'" (Appendix B).
Exhibits C-1 through C-4 are copies of a circular describing the device in language similar to that used in the advertisements offering for sale for $129 a "Truth Machine"; an order form for a "Microtronics 'Truth Machine'", which form was filled out by Postal Inspector W. F. Powers; a copy of a postal money order in the amount of $129 payable to Telestar, Incorporated; and a business reply envelope addressed to Respondent at Wromleysburg, Pennsylvania.
It is found that Respondent uses the United States mails in the conduct of the business of selling certain devices called either "Truth Machine" or "Microtronics 'Truth Machine'".
I find as a fact that in the conduct of said business Respondent makes the representations set forth in Paragraph III of the Complaint.
In support of said finding the following portions of the advertisements are cited: 2/
III (a) and (c) -
Appendix B, left-hand column, first paragraph:
"The Truth Machine is a new generation voice stress analyzer that flawlessly pinpoints deception."
III (b) -
Appendix B, left-hand column, fourth paragraph:
"When you ask a direct question you deserve a straight answer. And that's the beauty of the Truth Machine; it will give you a straight answer... even if someone else doesn't. It's your best possible defense against doubt, risk and deception".
III (d) -
Appendix A, left-hand column, sixth paragraph:
"Like many technological discoveries, voice stress analyzers grew out of military research during the Vietnam war. Army intelligence needed something better than the standard polygraph to interrogate prisoners. A simple method that could be used without the subject's knowledge. The voice stress analyzer was the result]"
III (e) -
Appendix A, middle column, third paragraph:
"There is no better way to get at the truth.. and remove the risk and uncertainty from those important decisions that face you every day]"
III (f) -
Appendix B, right-hand column, first paragraph:
"The Truth Machine from Microtronics is the ultimate voice stress analyzer."
III (g) -
Appendix B, left-hand column, paragraph starting at bottom of the page:
"The principle is remarkably simple. Scientists already knew lying produced unconscious and uncontrollable stress that could be recorded by a polygraph. Researchers soon discovered that this stress also affected the muscles controlling the vocal cords, and caused an inaudible 'microtremor' in the voice. All that was needed was a device sensitive enough to pick up and record these inaudible vibrations. And that was a relatively easy accomplishment considering the state of modern electronic technology."
III (h) -
Appendix A, middle column, last paragraph:
"With our easy, step-by-step instruction manual you can easily master the Truth Machine with only a few hours of practice."
Appendix B. middle column, last paragraph:
"You can easily master the Truth Machine after only one weekend of practice."
III (i) -
Appendix B, right-hand column, first complete paragraph:
"The Truth Machine is designed and built to guarantee you years of dependable use. It should never need servicing. But if anything ever does go wrong, Microtronics will repair it free of charge through their service-by-mail center and return it to you in a matter of days."
In addition to the foregoing positive evidence, it is noted that Respondent in its Answer did not deny making the representations in Paragraph III of the Complaint. Again, under 39 C.F.R. 952.10(b), when allegations of fact in a Complaint are not denied or expressly admitted, such facts may be considered as proved, and no further evidence regarding these facts need be adduced at the hearing.
Truth or Falsity of Representations
Dr. William Apple, Assistant Professor of Psychology, Psychology Department, Columbia University, was called as Complainant's first witness. Dr. Apple received his Bachelor's degree in Psychology from the City College of New York, his Master's and Ph.D degrees in Psychology from Columbia University. Dr. Apple's training has been primarily in experimental social psychology. His areas of interest include non-verbal behavior and non-verbal communication and, in particular, the effects of strong emotion and deception on the voice. He collaborated in two published papers dealing with voice changes during emotion. 3/ (Tr. 12)
In August, 1979, Dr. Apple saw a full-page advertisement of the Microtronics Voice Stress Analyzer Series IV in the New York Times Business Section. He was interested in the product because at $150 it was remarkably cheap compared to other units costing perhaps $4,500.
He contacted the New York City Office of Consumer Affairs and offered his service, without charge to the city, in connection with the advertised device. After about two months he was contacted by the City of New York and authorized to do some tests on a device obtained by subpoena by the New York Consumer Affairs Office from Larry Leeds, the name appearing on the amended answer to the Complaint as being an officer of the Respondent. (Tr. 14-15)
The witness stated that the unit he tested was similar to Complainant's Exhibit D-1, except that the tested unit did not have a built-in microphone, whereas Exhibit D-1 did have such a microphone. 4/
The witness stated that aside from that one feature the labeling, switches and phone jacks on Exhibit D-1 are identical with the device he tested.
On the face of the Exhibit D-1 are the words "MICROTRONICS SERIES IV Voice Stress Analyzer." Dr. Apple said that the same words are on the apparatus that he tested. (Tr. 23)
The report of Dr. Apple's tests of the Microtronics Voice Stress Analyzer, Series IV, was received in evidence as Complainant's Exhibit F. (Tr. 35) He testified that this report accurately reflects the substance of his testimony. (Tr. 34)
Some pertinent portions of his report are as follows:
"Abstract. Twelve subjects were interviewed using a standard deception task requiring them to lie about a numerical digit they had previously selected. Speech material was analyzed by the Microtronics Voice Stress Analyzer, Series IV. No differences in stress readings were found between subjects' truthful and untruthful interview responses, and the machine proved no better than random guessing in detecting deception.
"Subjects and interrogation procedure. Twelve men (mainly graduate student volunteers) were individually interviewed using a standard laboratory deception task (adapted from Friedhoff, Alpert, & Kurtzberg, 1964); the test, in fact, was a variation of a procedure suggested in the Microtronics Owner's Manual (pp. 45-46). (Ex. F, p. 1)
"The interview was presented as an 'attempt to validate a new voice lie-detector,' and subjects were told to try to be 'as convincing as possible in attempting to deceive the machine on their untruthful answers.' Subjects were then asked to choose a digit between 1 and 9, which would be 'their' number throughout the session, and to tell their choice to the interviewer. The interview itself consisted of a series of inquiries about different digits, e.g., 'Is ___ the number you chose?' The subject's task was simply to reply 'No' in response to each such query. The interviewer pointed out that for most questions, the subject's negative response would be truthful; however, when it came to the interviewer's asking about the 'critical' digit (i.e., the one the subject had chosen as 'his' number), the subject's reply would be a lie.
"All subjects indicated understanding the instructions -- simply to repond No to each question. There were five interrogation series: a warm-up series of 15 questions (none of which asked about the critical digit) followed by four 'stress' series. Each of these latter groups of questions contained roughly ten or twelve inquiries with a query about the subject's critical digit somewhere in the center (e.g., the first of these 'stress' series consisted of twelve questions, and Question 8 asked about the critical digit).
"All spoken responses were tape-recorded on high-quality audio equipment for subsequent voice-
stress analysis with the Microtronics machine.
"Results. We analyzed 28 spoken replies for each subject: the critical question in each stress series as well as the three (nonstressful) replies immediately preceding and immediately following it. The operator calibrated the stress analyzer anew for each stress series, adjusting the calibration knob as suggested in the manual so the first three (truthful) replies gave an average reading in the 020-040 range. Two readings were then taken on each of the subject's 28 replies, and the results for each syllable were averaged.
"The Owner's Manual asserts that deceptive responses should give higher stress readings than truthful responses. However, our data show that, on average, true and lie responses do not differ appreciably in the number of stress units the machine displays." (Ex. F, p. 2)
* * *
"These results indicate the Truth Machine cannot be used reliably to detect deception. An interviewer using a table of random numbers to determine each reply's 'stress' would be just as accurate detecting deception as one using the Truth Machine's digital readout." (Ex. F, p. 3)
* * *
"Ease of operation. The operator must first calibrate the machine (i.e., adjust a knob so that the stress reading is in the 020-040 range for a one-syllable truthful response). Following calibration, the operator must press a 'clear' button prior to each spoken response (to reset the digital readout) and then quickly depress (and continue holding) the 'analyze' button for the response's duration (usually one syllable); this button must be released immediately following the spoken response.
"Such a sequence of operations seems to require an extraordinary amount of dexterity and coordination -- so much so that I cannot imagine anyone seriously using the stress analyzer in a 'live' situation and getting usuable results.*** (Ex. F, p. 4)
* * *
"What does this machine count? Although I have not systematically tested this it was my impression that loud responses gave the highest digital readouts. Every speaker will ordinarily vary in his amplitude, and I assume this will appreciably affect his stress readings. Similarly, any physical movement on the subject's part that changes his distance from the microphone, including changes in the orientation of his head, will also add to the amplitude-related fluctuation in stress readings.
"Further, I observed that long responses likewise gave the highest stress readings. Thus a sustained 'No o o' gets a higher readout than a clipped 'No,' all other things equal." (Ex. F. pp. 4-5)
* * *
"However, my results show that the machine is of little value in detecting deception using a standard interrogation task which has proven useful in other detection situations (Friedhoff et al., 1964)." (Ex. F, p. 6)
In addition to the foregoing statements as to the results of the scientific tests, Dr. Apple answered in the negative, and gave supporting statements for his answers, when asked whether the device, (a) may be operated, and (b) will produce for the user thereof the results claimed in the representations heretofore found to have been made by the Respondent in its advertising matter. (Tr. 40-44).
On cross-examination it was further developed, that while Dr. Apple followed approved procedures that constituted a variation of the procedures set forth in the operator's manual accompanying the device, he used equipment that is superior to that recommended by Respondent. (Tr. 46; see also Tr. 36, 66, 73)
In the course of cross-examination, Respondent's counsel asked for, and was granted, opportunity to analyze the tapes and work papers developed in Dr. Apple's study. Whether such an examination was ever carried out is not known, but, if it was, the results were not disclosed at the hearing.
The next witness called by Complainant was Harvie Hammond Branscomb, A.B. Magna Cum Laude, Cornell University, 1974, and M.S. in Electrical Engineering and Computer Sciences, Massachusetts Institute of Technology, 1978. 5/
Mr. Branscomb played tapes and recorded voices two or more times into the device under the same conditions in terms of length of speech, amplitude and background noise, but got a number of different results. (Tr. 101, 102)
Some of these variations range from 0 to 040 on the LED readout device. The same input into the device should have produced the same readout on the indicator, but this was not the case. (Tr. 103)
Mr. Branscomb stated that while there are instruments which purport to measure stress by the sound of the voice, there is none that is recognized as being scientific. (Tr. 107)
Some sources of inaccuracy or imperfection in such devices are: (1) phonetic characteristics of speech; (2) room acoustics; (3) limited evidence for the hypothesis that the microtremor can be measured by this or any other device; and (4) lack of persons who are aware of, or fully trained in, the use of scientific methods to maintain the best possible accuracy to achieve an accurate measurement of stress. The last point was characterized by the witness as a "suspicion." (Tr. 108-110) The device will neither operate as indicated, nor produce the results held out in Respondent's representations. (Tr. 110-112)
On cross-examination, the witness agreed that the owner's manual contains the statement that the device is only an aid or guide in determining deception. (Tr. 117) Granting the truth of the foregoing statement, it is not applicable to the letter or the spirit of Respondent's advertisement since no such idea is suggested in the advertisement.
The last witness called by Complainant was Dr. Malcolm Brenner, B.A., Magna Cum Laude, Boston University, 1967; M.A., Stanford University, 1968; Ph.D., University of Michigan, 1976. 6/ Dr. Brenner is a psychologist, who is a Research Associate at the University of Oregon, and, also, a contractor for the National Aviation and Space Agency, doing research in the lIfe Science Division.
Dr. Brenner collaborated with Mr. Branscomb in tests described in the testimony of the latter, and Dr. Brenner performed the analysis of the results. (Tr. 137)
Using excellent equipment under high fidelity conditions, tests were run to determine the effect of three factors: degrees of loudness of speech, the speed of speech, and the reliability of the readout results on repeated playing of the tapes. Dr. Brenner described the different phases of the tests. (Tr. 138-142)
Dr. Brenner's impressions, based on his tests, are as follows: the machine does not respond well to deception; loudness made a difference in the LED readout results, and could have been scored on the machine as deception; in some cases the machine kept registering after the subject stopped talking, or, in other cases, before the subject started talking; and the machine started to register in some cases when a car passed by or when the microphone was shaken. (Tr. 144-148)
Dr. Brenner does not concede that the device is easy to use or that it flawlessly pinpoints deception. His view is that there is no instrument that measures deception, let alone accurately pinpoints it. (Tr. 150, 151) The device being sold by Respondent will not operate or produce the results as claimed in the advertising material published by Respondent. (Tr. 159, 160)
It was brought out again, through cross-examination of Dr. Brenner, that the owner's manual indicates that the device is only an aid or a guide to the determination of deception. On this point Dr. Brenner stated: "The manual says that. The ad does not. The manual has extensive qualifications." (Tr. 162) Dr. Brenner's statement is accurate.
The manual refers to a five-second time interval between question, answer, next question and so forth. A question on this point was raised by Respondent's counsel. Dr. Apple previously had stated that he did not believe this interval is of any real significance. (Tr. 77) Dr. Brenner's testimony indicated that he agreed with Dr. Apple's assessment. (Tr. 163)
The polygraph machine, with which Dr. Brenner is familiar, enables one to tell whether or not someone is under stress. For the purposes of this case, however, one has to determine whether the stress is the product of deception or some other cause. (Tr. 167-168)
Present experiments indicate that voice stress analyzer machines are not as reliable as the polygraph. Stress is not very well understood and different measures tend not to agree very well. Psychological stress may result from a number of things. The fact that a person suffers stress, which is detectable on a particular machine, does not necessarily mean that the indication of increased stress is an indication that the subject has lied. There is a good deal of art as well as science involved in determining deception. (Tr. 172-173)
Dr. Brenner, who has professional and technical knowledge and experience in the precise field in which this hearing centers, concluded his testimony with a statement concerning voice measures as a means of detecting deception, so pertinent that it merits direct quotation, in part, as follows:
"I think that using stress analysis to infer deception is one of the most demanding applications of stress analysis and would require an extremely fine measure, a hypothetically excellent measure, before you could make that kind of judgment. I think it may be one of the last applications that can ever be done of stress. At this point the technology developed I do not think is at a point that it gets even close to being used in deception situations." (Tr. 172-3)
Complainant rested its case.
The hearing in this proceeding began on March 4, 1980. The hearing proceeded with the presentation of Complainant's case on the basis of the representation that Respondent would present its witnesses at a later date (Respondent's letter of February 29, docketed on March 3, 1980).
At the conclusion of the Complainant's case in chief, the hearing was recessed until April 8, 1980, to permit Respondent to further cross-examine Complainant's witnesses based on Respondent's requested study and analysis of all working papers and tapes of Complainant's witnesses. At the resumed hearing on April 8th, Respondent was also to present its own witnesses.
On April 8, 1980, Respondent advised that witnesses who Respondent's counsel believed would be present to testify, had not appeared. These witnesses were contacted by telephone on April 8th. One such witness, a California resident, indicated that other commitments prevented his appearance. Two other witnesses, both psychologists located in Burlington, Vermont, were represented to Respondent's counsel as being available on April 8th. These witnesses were called, and they advised that they had not been contacted for a period of about three months, and one of them said that he had never been asked to appear as a witness for Respondent. (Tr. 184, 190)
A mail detention order, pursuant to 39 U. S. COde 3007, with respect to Respondent's mail was in effect. Partly on the basis of this fact the hearing was again recessed, this time to April 29, 1980. Respondent was advised that all work papers and tapes developed by Complainant's witnesses would again be available for copying and analyzing, if Respondent so desired. (Order of April 8, 1980)
Respondent's counsel stated that he had no intention of asking Dr. Brenner and Dr. Apple any further questions, and he indicated that they need not be required to attend a later session of the hearing. (Tr. 192)
There was introduced into evidence in this proceeding evidence involving the solicitation and sale through the mail of a device called a "Truth Machine." Dr. Apple testified with respect to a device called a "Truth Machine", which had been furnished by one Larry Leeds to the Consumer Affairs office of the City of New York. This device came into Dr. Apple's hands from a responsible representative of that office. The other device concerning which Dr. Brenner and Mr. Branscomb testified, was obtained through the mails by means of a test purchase from the Respondent. This second device is called a "Microtronics 'Truth Machine'". (Tr. 155) Thus, regardless of the name of the device, both came from the same source, namely, the Respondent in this case.
Appendix A advertises a "Truth Machine", and Appendix B advertises a "Microtronics 'Truth Machine'". In both adver- tisements interested persons place their orders with one or another variation of Respondent's name as shown in the caption.
There is expert testimony in this record about the specific devices heretofore indicated, and there is expert testimony about voice stress analyzers in general as they may be used or useful for the detection of deception.
While this decision was being dictated into the record, Respondent's counsel came into the hearing room at approximately 10:44. Respondent's counsel indicated that he had been detained because of the heavy traffic in the airlines in New York, and that this fact was responsible for his late appearance at the hearing.
The dictation of the decision was interrupted in order that Respondent's counsel might present oral argument. I listened with care to the presentation made by Respondent's counsel, and I find that nothing that was urged by counsel in his oral argument persuaded me that there is any reason to change any of the statements contained in this decision prior to the time of the arrival of Respondent's counsel, and nothing that has caused me to change the remarks that I had proposed to make in the remainder of the decision.
I would point out that in his argument Respondent's counsel referred to photographs appearing in the advertisements, which are Appendices A and B to this decision. While I do not purport to be an expert in such matters, it appears to me that the illustrations on these two appendices look more like drawings or artist's presentations than actual photographs of an existing device.
This may or may not be the case, but that is my impression. Even if that impression is erroneous, it does not cause me to change the conclusions that I will soon reach in this matter.
The facts are that we have had testimony about the two Truth Machine devices, both of which came from the Respondent, and neither one of which will live up to its billing, as that billing is presented in the Respondent's advertising material.
Conclusions of Law
1. Respondent is engaged in the sale of a device or devices called the "Truth Machine" or "Microtronics 'Truth Machine'".
2. In the conduct of this business, Respondent solicits remittances of money and makes delivery by means of the United States mail.
3. Applying the criteria set forth in Donaldson v. Read Magazine, 333 U.S. 178, that the effect of advertising material is to be measured by its impact on the average reader, the Respondent makes the representations set forth in Paragraph III of the Complaint.
4. There is no evidence in this case that supports a conclusion that the representation in Paragraph III (i) is false, except as hereinafter set forth in Conclusion of Law number 5.
5. Representations in Paragraph III (a) through (h) of the Complaint are false based on the testimony with respect to the specific devices tested and based on the expert's testimony that the art and science of deception detection by means of voice stress analyzers has not developed to the point that such devices are reliable.
6. The representations found to have been made by Respondent are material representations since they are of the kind and character that would cause readers of Respondent's advertising matter to make remittances of money to Respondent to purchase the product offered for sale.
7. There is no evidence in this record other than the evidence adduced by Complainant despite the fact that Respondent has been given repeated opportunities to call witnesses on his own behalf.
8. Respondent, in the conduct of the business aforesaid, is conducting a scheme or device for obtaining money or property through the mails by means of false representation within the meaning of Section 3005 of Title 39, United States Code.
Accordingly, a mail stop order as provided for in 39 U. S. Code Section 3005, substantially in the form attached, should be issued against the Respondent in the names appearing in the caption hereof.
1/ This decision was rendered orally at the close of the hearing. It has been edited and transcribed for formal issuance.
2/ The citations shown do not purport to include all portions of Respondent's advertising material which support this finding of fact.
3/ Dr. Apple's Curriculum Vitae was received in evidence as Exhibit C-5.
4/ This statement, made during a cursory examination of the device, is in error, since Exhibit D-1 does not have a built-in microphone.