Barely a month after sharing news about an Army polygraph instructor telling me he purchased a copy of THE CLAPPER MEMO, I received an inquiry from another polygraph loyalist, John J. Palmatier, Ph.D, whose name appears twice inside the pages of my second nonfiction book. What can I surmise from these communications? The book has generated interest among polygraph loyalists and, perhaps, has them worried.
Dr. Palmatier, who contacted me via my website Friday, wrote that he had just finished reading Chapter 22 of my book, THE CLAPPER MEMO, and needed my assistance in verifying some information presented in that chapter. Specifically, he pointed to details I shared about a research article by Professor Emeritus James L. Chapman and neuroscientist Marigo Stathis that appears on page 232 of the book. The article focuses on the duo’s research about the effectiveness of the Computer Voice Stress Analyzer® (CVSA®), a credibility assessment technology that competes head to head with the polygraph.
Though he included only an excerpt of a single paragraph from the chapter, I think it’s worthwhile to share all of CHAPTER TWENTY-TWO: The Capstone before proceeding. With only minor changes for stand-alone publication, the correction of a typo, the replacement of one footnote from the book with a hyperlink and the omission of another footnote to the study research paper in question, the text of that chapter appears below, accompanied by graphics that do not appear in the book:
Twenty-four-hundred miles east of Tucson, a man in Upstate New York spent more than four decades believing stress can be detected in the voice utterances of individuals facing real-life jeopardy.
Not your stereotypical academic, Professor Emeritus James L. Chapman began his adult life as a United States Marine and served in Southeast Asia. From there, he went on to become a New York cop and, later, a criminology justice and forensic crime professor at the State University of New York. And, oh yes, he’s the same professor I mentioned as having contacted Professor Kelly R. Damphousse and the other researchers involved in The Oklahoma Study.
Professor James Chapman
(7/04/1942 – 4/17/2012)
When he wasn’t teaching, Professor Chapman helped law enforcement professionals across the nation use CVSA® as an investigative tool.
During Professor Chapman’s four decades of work in the field of credibility assessment, he conducted more than 15,000 exams, ascended to the level of Master CVSA® examiner, served as director of Standards and Training of the National Association of Computer Voice Stress Analysts (NACVSA), and built a reputation as the world’s foremost authority on the application of CVSA® as an investigative tool.
After retiring from full-time teaching, Professor Chapman refocused his efforts on finding a more global way to disseminate the positive findings he had gathered on the use of CVSA®. Although he wrote a lengthy white paper detailing his data and results, his ultimate mission was to publish the latter in a peer-reviewed, criminology-based scientific journal.
In June 2010, Professor Chapman teamed up with Marigo Stathis, hoping she could help him reach the goal he had previously been unable to reach on his own.
Stathis, a well-respected neuroscientist and research analyst, brought to the equation her in-depth knowledge of the scientific method, statistics, and academic research experience that includes authorship of 27 published scientific articles.
Before working together, both Professor Chapman and Stathis discussed the obstacles each foresaw regarding their collaboration and the creation of a manuscript worthy of publication.
Professor Chapman warned Stathis of the turf war between polygraph loyalists and others advocating the use of other credibility-assessment technologies, including CVSA®. In turn, Stathis reassured the professor that the statistical analyses and numbers would hold their own and reveal the “truth” of the matter if CVSA® was truly as effective as he claimed.
Stathis also made it clear that, as a scientist, her goal was not to disprove other technologies (of which she had no prior knowledge or bias), but to illuminate and showcase Professor Chapman’s particular study of CVSA®. In addition, she reiterated that a scientific study worthy of publication should present hypotheses, methodologies, data, results, and conclusions that stand on their own merits.
After six months of intense collaboration which included consultations with world-class statisticians at The Johns Hopkins University, Professor Chapman and Stathis completed the first version of their paper showcasing the CVSA®-focused research study. After choosing a few reputable criminology-oriented, scientific journals in North America, they submitted the paper to the first, hopeful for a positive response. Soon thereafter, they received their first rejection letter. Still hopeful, however, they used the criticisms they received to rewrite the paper, and submitted it to their second choice journal. A few months later, they received the reply: another rejection letter.
Although disappointed, Professor Chapman was not surprised. He had, after all, witnessed the turf war firsthand.
Stathis was surprised, but undaunted, having amassed considerable experience when it came to editorial peer reviews and publications. She was familiar with rejection and understood that rewriting a paper based on reviewers’ comments and suggestions only makes for a stronger end product.
Despite the level of harshness, the types of editorial criticism to which she was accustomed were always grounded in validity and solid reasoning. Conversely, she was taken aback by some of what she perceived to be illogical, inaccurate, and even emotional feedback about the paper she and Professor Chapman had submitted.
One journal reviewer cited the studies completed by researchers at the universities in Florida and Oklahoma as reasons to discount the CVSA® study, apparently not realizing they had already been debunked.
Another cited the lack of laboratory studies to prove the technology works, despite the fact that Professor Chapman’s study was clearly cited as a “field-based” one, requiring real-world settings where jeopardy is present.
Still another reviewer, obviously not aware of the research being done in Arizona, claimed there was little or no scientific basis for claims that voice stress can be measured.
Due to their shared conviction in the validity and worth of the study, however, Professor Chapman and Stathis decided to remain persistent despite the opposition. After several months of further rewriting and chiseling, the duo finalized the last version of their manuscript in March 2012 and submitted it for peer review to editors at Criminalistics and Court Expertise, one of the world’s oldest and most-reputable scientific journals.
Although nervous about the fate of the paper, Professor Chapman and Stathis knew their final manuscript was scientifically sound. Every “i” was dotted, and every “t” was crossed. Several more months passed as the peer-review process took place. Then, finally, the pair’s efforts culminated in acceptance. Their paper, “A Long-Term, Retrospective Field Evaluation Of Voice Stress Analysis In A Criminal Justice Setting,” was published in the journal’s December 2012 edition. [Footnote appears in book.]
Unfortunately, Professor Chapman passed away unexpectedly at the age of 69, one month after the paper had been submitted for peer review, and never got to see it published. Others, however, will see the paper and learn from it, a capstone on the professor’s long and distinguished career.
The retrospective field study was, according to its authors, based upon actual CVSA® examinations Professor Chapman conducted in the United States and Canada during an 18-year period. It involved conducting four extremely thorough statistical analyses of data related to actual criminal cases, suspects and persons of interest. No white coats or lab rats were involved.
The total inventory of cases submitted was culled for ones that met several key criteria: A) A confession was a potential outcome; B) There was no involvement with veracity testing of previous statements; C) No employment clearance was involved; D) The case was not used as confirmation of prior witness testimony; and E) The facts of the case were such that responses could be verified by means of structured CVSA® follow-up questioning.
After excluding cases that did not meet all of the above criteria, the number of criminal cases remaining for study totaled 2,109. Those cases were then numbered in consecutive order, and 236 cases for study were randomly selected, by chance, from that number.
The cases selected involved 329 specific crimes, including — but not limited to — multiple homicides, corporate theft, organized crime, contract murders, sexual abuse of children and arson for hire.
Subjects examined represented a wide spectrum of people, male and female, and included individuals with no criminal history as well as others with arrest and/or conviction records.
The socio-economic strata ran the gamut from wealthy, well-educated professionals to illiterate indigents and included several subjects who had been professionally evaluated and found to be of below-normal intelligence.
The livelihoods of those within the study group ranged from elected public officials to professional criminals and even included some organized crime “hit men.”
Of the cases studied, according to the study’s final report, 91 percent represented criminal investigations in which legal authorities had reached investigative impasses. In other words, after following standard investigative protocols, investigators had been unable to reach firm conclusions as to the guilt or innocence of the subjects.
Each subject named in the “Confession Possibility List” was individually interviewed by the CVSA® examiner with two goals in mind: (1) To exonerate the innocent and identify the guilty, and (2) To obtain legally-valid and independently-verifiable confessions from those subjects who were unable to clear the CVSA® process.
Each interview, according to the final report, was conducted according to a standard protocol in which the wording of the interview, but not the methodology, was adapted on site to the specifics of each case.
In each case, the procedure used by the CVSA® researcher consisted of six steps:
Step 1 — The CVSA® examiner was briefed by the requesting authorities in order to become familiar with the circumstances of each case. This briefing included available physical evidence, witness and suspect statements collected to date, and any background information concerning the subject to be tested. A determination of focus areas for questioning was made at that time by the CVSA® examiner.
Step 2 — A recorded pre-test interview with the subject was conducted. Also, during this portion of the interview, sufficient time was allotted to: a) obtain the subject’s informed consent for the interview; b) review the topical areas of questioning with the subject; c) attempt to help the subject relax and feel at ease with the examiner; d) provide an opportunity to discuss the subject’s concerns and to clarify terms and issues; and e) allow the examiner to formulate precise questions to be asked of the subject during the CVSA® examination.
Step 3 — The initial test questions were then presented to the subject. Examinations contained from nine to 31 questions, consisting of relevant, irrelevant, and control questions for which the subject provided “yes” or “no” answers. The initial set of questions was designed to address specific issues concerning the crime(s). As the examination progressed, the questions became more specific about the crime issue(s). Included in the set of questions, interspersed among the relevant questions, were questions with known answers, such as, “Is your name Bob Smith?” To this, the subject was directed to answer correctly. Other questions with known answers were also included, but to which the subject was directed to answer incorrectly. The CVSA® graphs produced by these non-relevant questions provide the examiner with additional reference points for the interpretation of the CVSA® results.
Step 4 — The fourth step included processing the responses with the CVSA® instrument, after which the resulting CVSA® charts were analyzed and interpreted by the examiner.
Step 5 — If stress patterns associated with specific relevant questions were observed by the examiner, an opportunity was given for the subject to provide additional clarification regarding the stress. Prior to the re-examination, questions were reformulated by the examiner to evaluate the veracity of the explanations offered by the subject. This procedure was repeated until all necessary questions had received responses that included no displays of stress reaction, or until the remaining stress reactions could not be eliminated by the explanation or the re-questioning.
Step 6 — The final step of the process was to provide a conclusion regarding the outcome of the CVSA® examination. If the relevant questions produced a “No Stress” chart, the subject was “cleared” by the CVSA® procedure. This information was then turned over to the agency requesting the examination.
If a confession was made by the subject during the CVSA® examination, the examiner would ask the subject to support his/her confession by verifying details or by providing further details concerning the events under investigation. Further, if a confession occurred, the subject was asked to provide a written statement and to confirm evidence that had not been made public about the case. Another CVSA® examination would then be conducted to validate the accuracy of the written statement. If no confession occurred, the examiner reported the findings to the agency requesting the CVSA® examination, such that the information could guide further investigation.
Of the 329 confession possibilities in this study, 92.1 percent of the CVSA® examinations produced a “Stress Indicated” result and 89 percent of those culminated in actual confessions. Most notably, suspects made self-incriminating confessions during 96.4 percent of the interviews during which CVSA® indicated stress.
The results of this study, according to its authors, clearly demonstrate CVSA® is not only a useful tool in obtaining valid confessions, but also that the likelihood of obtaining valid confessions increases based upon whether or not stress is present for relevant crime issues.
In each of the 236 cases included in the study, inclusive of 329 confession possibilities, a trained and experienced CVSA® examiner (i.e., Professor Chapman) used well-established CVSA® protocols with the objectives of producing legally-admissible confessions and obtaining additional supporting evidence from suspects and/or persons of interest.
With current scientific research revealing that only 20 percent to 50 percent of police interviews/interrogations produce valid confessions, the study’s authors explained, a 96.4 percent-verified confession rate is considered phenomenally high.
Adding to the conclusions above, the study’s authors shared additional findings.
“More importantly,” they wrote, “this study also proves CVSA® to be a useful and predictive tool in separating the innocent from the guilty, by conclusively demonstrating the ability to discriminate stress from no-stress in the human voice. In one stand-alone case of Grand Larceny, 20 individuals were considered suspects. Of the 20 CVSA® examinations conducted for this case, 19 resulted in a finding of ‘No Stress Indicated,’ while only one produced a ‘Stress Indicated’ finding – resulting in a confession.
By use of the CVSA®, the accurate identification/ separation of the 19 innocent individuals from the one who was guilty, far surpassed a “chance” rate of accuracy. The probability of 20 successful evaluations would have been less than 1 in 1,000,000.”
In summary, the conclusions of the study were: (1) During criminal justice investigations, CVSA® can serve as a reliable decision support tool to help discriminate between “Deception” and “No Deception”; (2) Voice stress and confession rates are interdependent; (3) The CVSA® process can precisely and accurately discriminate stress from no stress in real-life crime situations involving consequence and jeopardy; and (4) the level of consequence and jeopardy associated with specific crimes can affect the confession rates obtained from guilty individuals under examination.
In January 2013, Stathis shared the results of the study with attendees at a worldwide conference of CVSA® examiners held in South Florida. In addition to highlighting the information outlined above, Stathis delivered a clear message to those who discount the technology proven to work by Professor Chapman.
“Just as you would not call an X-ray scan a prop if it provides accurate information that helps a medical doctor in formulating a diagnosis,” she said, “you would not call the CVSA chart a prop if it reveals accurate information that helps a trained detective in making an assessment. The CVSA is not a prop. It’s a tool.”
Why did I opt to share the entire text of Chapter 22? Because I found it odd how Dr. Palmatier seemed to demonstrate through the heart of his inquiry that he is more interested in discrediting the journal than he is in evaluating the research conducted by Professor Chapman and Stathis. Of course, I hope I’m wrong.
“I publish in peer reviewed journals and thought that this journal as you described it may be an excellent resource for substantive information,” he wrote. “But alas, I could not find the Journal through any of the databases that I have access to for research and so asked reference librarians at the Shepard Broad Law School at Nova Southeastern University, Michigan State University, and John Hopkins University, using their resources to try and find this journal to no avail.
“Before I write something, as a scientific researcher, due diligence is a requisite before placing something to print,” he continued. “Consequently, I am asking for your assistance in locating ‘one of the world’s oldest and most-reputable scientific journals’ that three reputable ‘RESEARCH’ universities have not been able to locate.”
Dr. Palmatier went on to describe himself as an adjunct professor at Nova Southeastern University who teaches “credibility assessment” in the national security affairs graduate program. He closed his inquiry by writing, “Your assistance in locating this reference will help in my efforts to teach future thinkers. Thank you in advance,… JP”
Now, regarding Dr. Palmatier’s request for assistance in obtaining a copy of the study, I offer a concise reply.
As I recall, Professor Chapman and Stathis advised me that, prior to it being published, the paper was submitted to the non-U.S. scientific journal in both English and the language of the country in which it is published. It was accompanied by a lot of legal documentation and certifications prior to the journal’s peer-review process taking place. So, am I going to offer you the contact information for the editor of the foreign journal that published the study? Ne. Nee. O. Non. Nein. Nincs. Nyet. Nej. No. Afraid not.
Why? Because I spent much of the past four years conducting my own due diligence and think you should find it on your own — that is, if you’re not too busy teaching polygraph courses in China and running your polygraph business in Miami.
FYI: Crossposted at http://BobMcCarty.com.
Bob McCarty is the author of Three Days In August and THE CLAPPER MEMO. To learn more about either book or to place an order, click on the graphic above.