A Critical Examination of Converus Inc.’s Pseudoscientific “EyeDetect” Lie Detector

An EyeDetect station

In “An Eye-Scanning Lie Detector Is Forging a Dystopian Future,” Wired science writer Mark Harris investigates “EyeDetect,” a new scientifically unproven lie detector being hawked as a replacement for the old scientifically unproven polygraph.

Converus Inc. of Lehi, Utah, which markets EyeDetect, claims without proof that it can detect lies with 86% accuracy. Mark Harris takes a critical look at this claim. Excerpt:

Sitting in front of a Converus EyeDetect station, it’s impossible not to think of Blade Runner. In the 1982 sci-fi classic, Harrison Ford’s rumpled detective identifies artificial humans using a steam-punk Voight-Kampff device that watches their eyes while they answer surreal questions. EyeDetect’s questions are less philosophical, and the penalty for failure is less fatal (Ford’s character would whip out a gun and shoot). But the basic idea is the same: By capturing imperceptible changes in a participant’s eyes—measuring things like pupil dilation and reaction time—the device aims to sort deceptive humanoids from genuine ones.

It claims to be, in short, a next-generation lie detector. Polygraph tests are a $2 billion industry in the US and, despite their inaccuracy, are widely used to screen candidates for government jobs. Released in 2014 by Converus, a Mark Cuban–funded startup, EyeDetect is pitched by its makers as a faster, cheaper, and more accurate alternative to the notoriously unreliable polygraph. By many measures, EyeDetect appears to be the future of lie detection—and it’s already being used by local and federal agencies to screen job applicants….

Read the entire article here, and if you missed it, be sure to also see Harris’ earlier article, “The Lie Generator: Inside the Black Mirror World of Polygraph Job Screenings.”

Snake Oil Salesman Alex Martin and AC Global Risk’s “Remote Risk Assessment”

Alexander Martin (LinkedIn profile)

In “The Dangerous Junk Science of Vocal Risk Assessment,” Ava Kofman of The Intercept examines “Remote Risk Assessment,” a voice stress analysis application being marketed by AC Global Risk, a San Francisco-based company led by U.S. Marine Corps veteran Alexander Martin. The company characterizes Remote Risk Assessment as “a sophisticated system that uses unique voice analytics processes to evaluate and quantify distinct characteristics of the human voice related to risk.” Excerpt:

AC Global Risk declined to respond to repeated requests for comment for this article. The company also did not respond to a list of detailed questions about how the technology works. In public appearances, however, Martin has claimed that the company’s proprietary analytical processes can determine someone’s risk level with greater than 97 percent accuracy…. Several leading audiovisual experts who reviewed AC Global Risk’s publicly available materials for The Intercept used the word “bullshit” or “bogus” to describe the company’s claims. “From an ethical point of view, it’s very dubious and shady to give the impression that recognizing deception from only the voice can be done with any accuracy,” said Björn Schuller, a professor at the University of Augsburg who has led the field’s major academic challenge event to advance the state of the art in vocal emotion detection. “Anyone who says they can do this should themselves be seen as a risk.”

NSA Blocks AntiPolygraph.org on Twitter

The NSA (@NSAgov) has blocked AntiPolygraph.org (@ap_org) from following and viewing the NSA’s tweets:

Screen shot of the NSA’s Twitter profile as viewed from AntiPolygraph.org’s Twitter account, 3 November 2018

Curious about what may have prompted this action, we used Google to search twitter.com for matches including both “ap_org” and “NSAGov.” We found two replies that we posted to tweets the NSA made on 9 July 2018. Here is the first reply:

And here is the second reply:

We have asked the NSA to explain why they blocked us and will share their reply with an update here when received:

Update (12 November 2018): The NSA has not responded to our inquiry and continues to block AntiPolygraph.org on Twitter.

Mark Harris on Law Enforcement Pre-Employment Polygraph Screening

Mark Harris (Twitter profile)

Wired has published a major investigative article on law enforcement’s use of pre-employment polygraph screening. In The Lie Generator: Inside the Black Mirror World of Polygraph Job Screenings, science/technology writer Mark Harris (@meharris on Twitter) reports based on, among other sources, data gleaned from numerous public record access law requests filed with police and sheriff’s departments across the United States. Excerpt:

Data obtained by WIRED showed vast differences in the outcomes of polygraph tests depending on the examiner each candidate faced. Consider another law enforcement agency that uses polygraphs in its employment process: the Washington State Patrol (WSP). Between late October 2011 and the end of April 2017, the WSP conducted 5,746 polygraph tests on potential recruits. This was the largest data set WIRED received, including copious data on both applicants and examiners. While one examiner failed less than 20 percent of candidates, others failed more than half the applicants they screened. And while two examiners disqualified just four people in more than 1,000 applicants for supposedly having sex with animals, one of their colleagues failed more than 10 times as many for bestiality—around one in 20 of all job seekers. The same examiner was also twice as likely as the rest of his peers to fail applicants on the grounds of child pornography.

There were no further hearings trials for these supposed crimes, and no jury to convince or judge to adjudicate, just scores of otherwise qualified applicants who would now not become Washington state troopers.

“We don’t know which, if any, of the examiners are accurate, but the disparity between them suggests the test is not being used in a way that is at all reliable,” says John Allen, a professor of psychology at the University of Arizona. And tests that are not reliable, Allen says, cannot be valid.

Harris’ article is too important for any summary posted here to do it justice. Go read it in its entirety.

The AntiPolygraph.org Podcast: Episode 1 – Polygraphs and the Kavanaugh Nomination

In this episode, podcast host George Maschke discusses Professor Christine Blasey Ford’s polygraph examination in connection with her allegations against U.S. Supreme Court nominee Judge Brett Kavanaugh, critical journalism on polygraphy inspired by this development, a proposed polygraph dragnet at the White House, a recent interview of polygraph critic Doug Williams on the Lions of Liberty podcast, and AntiPolygraph.org’s 18th anniversary online. (Direct download, 7.9 MB MP3)

Show notes:

On Professor Christine Blasey Ford’s allegations against Judge Brett Kavanaugh:

“California professor, writer of confidential Brett Kavanaugh letter, speaks out about her allegation of sexual assault,” by Emma Brown. The Washington Post, 16 September 2018.

“Brett Kavanaugh’s Confirmation at Risk After Accusation by Christine Blasey Ford,” by Sheryl Gay Stolberg. The New York Times, 16 September 2018.

House Minority Leader Nancy Pelosi calls on Brett Kavanaugh to take a lie detector “test.”

General reporting on polygraphy:

“Polygraph tests don’t work as lie detectors and they never have: It’s time to stop pinning our hopes on pseudoscience” by Eleanor Cummins in Popular Science.

“Why Do We Still Believe in Polygraphs?” by Alice B. Lloyd in the Weekly Standard.

“The Truth About Polygraph Tests: They’re junk science, inadmissible in court, and about as reliable as a pack of Tarot cards,” by Claire Berlinski in City Journal.

Senator Rand Paul Suggests Lie Detector “Tests” to Identify Author of New York Times Op-Ed

John Odermatt’s interview of polygraph critic Doug Williams on the Lions of Liberty podcast show, Felony Friday.

How to Contact AntiPolygraph.org.

 

DHS OIG Whitewashes CBP Polygraph Complaints (or, CBP Dismisses 5 out of 6 Valid Complaints Against Its Polygraph Operators)

On 26 July 2018, the Department of Homeland Security Office of the Inspector General published report OIG-18-68, titled “Most Complaints About CBP’s Polygraph Program Are Ambiguous or Unfounded” (PDF).

The 17-page report’s findings are summarized on page 5:

Review of CBP’s Polygraph Complaints
We reviewed 157 complaints to determine whether CBP had an effective process and whether the complaints were true. The complaints fell into three categories — those missing information or otherwise too vague to review; those which were not true (the allegation was not substantiated by an audio review); and those which were true (the allegation was substantiated by an audio review). Of the 157 complaints, we determined that:

  • 130 (83 percent) were either not specific or did not have enough information to review;
  • 21 (13 percent) were not true based on the allegation; and
  • 6 (4 percent) were true.

We determined CBP did not adequately address five of the six substantiated complaints. For the complaint it addressed adequately, CBP conducted an audio review and allowed the applicant to retest.

What stands out is that 83% of complaints were judged as either “not specific” or “did not have enough information to review.” No breakdown is provided between these two categories, and no examples of such complaints are provided. However, it seems likely that DHS OIG would have placed in this category any complaint by any CBP applicant who alleged she told the truth yet was accused of lying by her CBP polygraph operator , simply because one cannot prove the negative in such a situation. That is, there is no way any CBP applicant falsely accused of lying can prove that she did not lie. It appears DHS OIG therefore dismisses such complaints out of hand. Yet complaints of being falsely branded as a liar by CBP polygraph operators are by far the most common ones heard.

The above citation also shows that “CBP did not adequately address five of the six substantiated complaints.” A more informative title for this DHS OIG report would be “CBP Dismisses 5 out of 6 Valid Complaints Against Its Polygraph Operators.”

NITV Hires Disgraced Ex-Cop Jerry W. Crotty II as Director of Law Enforcement Operations

The National Institute of Truth Verification, which markets a voice-based “lie detector” called the Computer Voice Stress Analyzer (CVSA), has hired disgraced ex-cop Jerry W. Crotty II to serve as its “Director of Law Enforcement Operations”:

NEW DIRECTOR OF LAW ENFORCEMENT OPERATIONS

NITV Federal Services (NFS) is pleased to welcome Detective Jerry Crotty as its Director of Law Enforcement Operations. Jerry will oversee all aspects of law enforcement operations for NFS and provide direct advice and assistance to law enforcement agencies worldwide, as well as oversee our Technical Services Division.

Jerry joins the NSF team from the Manatee County Sheriff’s Office in Florida where he served for 20 years in various positions, including supervisory positions in such specialized disciplines as Crimes Against Children, Domestic Violence and Internet Crimes Against Children (ICAC).  While assigned to the State and Federal ICAC Task Force, Jerry developed ground-breaking techniques for using the CVSA® to identify and bring to justice serial child predators, which are now taught nationally.  These techniques were so significant and effective that in 2015 he received the National Association of Computer Voice Stress Analysts (NACVSA) annual Professor James L. Chapman Award for Excellence.  Jerry is considered an expert in interviewing and interrogation and has a confession rate well above 95%.  He is also an expert utilizing the CVSA and is highly experienced in conducting specialized covert and structured examinations.  Jerry holds a Master’s Degree in Science from the University of Central Florida.

Charles Humble, Founder of NITV Federal Services and the developer of the CVSA, states “We are very excited to have an individual with the qualifications and strong moral character of Jerry Crotty joining our team.  With his law enforcement background, especially in the ICAC arena, Jerry will bring a new dimension to an organization that is already recognized as the Gold Standard for our industry.”

In February 2018, Jerry Crotty, then a supervisor in the Manatee County, Florida Sheriff’s Office Child Protection Investigative Division, “retired to avoid a demotion and discipline,” as Jessica de Leon reported in April for the Bradenton Herald. Excerpt:

Former supervisor of Manatee sheriff’s child protection division opts to retire after demotion and facing suspension

April 04, 2018 06:08 PM
Updated April 05, 2018 09:46 AM

A former supervisor in the Manatee County Sheriff’s Office Child Protection Investigative Division has retired to avoid a demotion and discipline after an internal affairs investigation concluded he had been inappropriate and demeaning toward child protective workers. His supervisor was suspended and reassigned after a related internal affairs investigation concluded he did nothing to stop the behavior and participated in the berating of others.

Former Sgt. Jerry Crotty announced he would be retiring in a letter dated Feb. 4 to Capt. Brian Schnering in the Professional Standards Section. According to an internal affairs investigation concluded on Feb. 20, allegations that Crotty violated three general orders — harassment on the basis of disability, conduct unbecoming and failing to adhere to general orders — were sustained against him.

“Over the last year, my physical and mental health has been pushed beyond its limits and this time off of work has given the clarity that I need moving forward,” Crotty wrote. “When I started my career in 1997, I told my family that the day it was no longer fun and felt like work, I would resign. That day has come now.”

Crotty had been on medical leave but anticipated being cleared by his doctors on Feb. 19, he said. His resignation as a result was effective Feb. 20, just two days after his demotion to deputy in the Crime Against Children Section became effective, according to his personnel file. Crotty also said in his resignation letter that he wished to discuss the investigation with Schnering so that it could be closed properly and because he did not wish to resign while he was under investigation.

After a related investigation, Crotty’s supervisor, Lt. Barry Overstreet, was suspended for six days without pay and reassigned to the patrol division.

Crotty and Overstreet both supervised the Crimes Against Children section of the Child Protection Investigative Division, which handles criminal investigations, often working investigations in tandem with the Child Protection Section. The Child Protection Section handles all child welfare investigations in Manatee County for the Florida Department of Children and Families.

“Based upon a review of all documentation, evidence and interviews, it is apparent Sgt. Crotty demonstrated harassment based on a disability when he purposely spoke in a diminished tone in an effort to force Deputy Director Connie Keehner, a hearing impaired employee, to ask him to repeat himself,” Sgt. Paul Davis stated in the internal affairs report. “Furthermore, it is evident Sgt. Crotty utilized a pattern of influential comments, actions and/or omissions, as well as deliberately phrased electronic communications to forge a mindset of separation and contempt between his subordinates and the employees of the Child Protection Section.”

Read the rest of the story here.
It may seem curious that NITV founder Charles Humble would consider Jerry Crotty to be an individual of “strong moral character.” But then again, Humble doesn’t think there is anything wrong with his passing himself off as a Ph.D. based upon a “diploma” received following a 6-hour course of study at a strip mall Bible school.
It should also be noted that the National Institute for Truth Verification has admitted in federal court that the Computerized Voice Stress Analyzer “is not capable of lie detection.”

“Nailing the Pretest Interview”: A Presentation by Skip Webb

AntiPolygraph.org has obtained a copy of a 2005 PowerPoint presentation prepared by former U.S. Army Criminal Investigation Command (CID) Supervisory Agent and American Polygraph Association past president and past chairman of the board of directors Milton O. “Skip” Webb, Jr.

Titled “Nailing the Pre-Test Interview: The key to reducing no opinion tests,” this presentation has been a favorite at polygraph conferences for years.

With a focus on specific issue polygraph interrogations conducted in criminal investigations, the opening slides offer a candid acknowledgement of the sort of examiner bias and error that can beset polygraph “tests.” This documentation may be useful to any attorney who needs to challenge the reliability of a client’s polygraph “test.” The following slides speak for themselves:

The entire 44-slide presentation may be downloaded as a 6.5 MB PDF file or as a 269 kb PPT file.

Polygraph Discussion on Witwatersrand University Radio Academy Program “The Science Inside”

Among the topics addressed in the 26 June 2018 episode of the Witswatersrand (Wits) University radio program and podcast “The Science Inside” was polygraph “testing.” Those interviewed for this segment by program editor Elna Schütz include Niel Visser, a retailer who has been subjected to workplace polygraph “testing,” Clifton Coetzee, a practitioner of a variety of pseudosciences including polygraphy, voice stress analysis, statement analysis, and physiognomy, and AntiPolygraph.org co-founder George Maschke. The polygraph segment begins about 36 minutes into the program.

Google De-lists AntiPolygraph.org on Key Search Terms

Search DuckDuckGo for “polygraph” and you’ll find AntiPolygraph.org, which hosts more documentation on polygraphs than any other site on the Internet, in the top 10 results. The same is true for searches on “polygraph” with Microsoft’s Bing search engine (in the United States), with Russia’s Yandex search engine, and with the open-source searX search engine.

But search Google for “polygraph” and you very likely will not be shown any links to any pages on AntiPolygraph.org.

This has not always been true. For years since going online in 2000, AntiPolygraph.org was Google’s top search result for “polygraph.” Eventually, the Wikipedia article titled “Polygraph” took the number 1 spot, but AntiPolygraph.org long remained among the top 10.

Then, around the time the federal government launched Operation Lie Busters, which targeted the owners of websites that provided information on how to pass (or beat) a polygraph “test” for entrapment and criminal prosecution, AntiPolygraph.org’s Google ranking for “polygraph” slipped, often appearing on the second, third, or later pages of results. AntiPolygraph.org has reason to believe that co-founder George Maschke was targeted in Operation Lie Busters, and that visitors to AntiPolygraph.org have been the target of electronic eavesdropping.

AntiPolygraph.org provides documentation on polygraphy that the U.S. government’s intelligence and law enforcement agencies do not want the public — especially those who face polygraph “testing” — to know, including the precise questions asked in various polygraph techniques (PDF), how to pass or beat a polygraph “test” (PDF), and the federal government’s unscientific and ineffective methodology for attempting to detect polygraph countermeasures.

In 2018, it appears that Google has effectively de-listed AntiPolygraph.org for such relevant keywords as “polygraph” and “lie detector.” We first noticed this early in the year; the de-listing has persisted for at least four months.

Google Search Analytics shows that from 29 March 2018 to 26 June 2018, there were no clicks via Google searches on “polygraph” to any page on AntiPolygraph.org:

There were only 7 clicks via Google searches on “lie detector” to any page on AntiPolygraph.org:

Similarly, there were zero clicks via Google searches on “polygraphs” or “lie detectors.”

For comparison purposes, during the same period, there were 165 clicks via Google searches on the exact phrase “where to find underage porn,” even though the linked page does not remotely address this question:

It seems that something is not kosher here. AntiPolygraph.org is a non-profit, public interest website. We host no advertising and we don’t engage in spam, link-farming, or other abusive practices.

Reader thoughts are welcome.

Update (30 June 2018): The situation appears to be worse than we imagined, as illustrated in this screenshot from Google Search Console Beta:

The chart shows that over the past 16 months, Google reports 146,576 impressions for the search word “polygraph,” with an average position of 8.4, but only 381 total clicks, for an average click through rate of just 0.3%!