EyeDetect False Positives: Can an Innocent Person Fail?

Defining False Positives and False Negatives

A false positive occurs when a truthful person — someone answering honestly — is incorrectly classified as "deceptive" by the testing system. The machine says the person is lying when they are telling the truth. This is the mirror image of a false negative, where a deceptive person is incorrectly classified as "truthful" and passes the test [1]Verified The Validity of Polygraph Examination of an Innocent Person and a Perpetrator Administered Several Years after a Murder
Confirms polygraph validity findings in real-world case with innocent and guilty subjects examined years after a crime.

Both types of errors are inherent in every credibility assessment technology ever developed, including polygraph, voice stress analysis, fMRI-based lie detection, and EyeDetect. Understanding these error types is essential for properly interpreting any test result.

Why False Positives Matter Most to Examinees

While false negatives are concerning from a security standpoint (a deceptive person passes), false positives cause the most direct harm to individuals. A false positive on an EyeDetect test can lead to loss of employment or failure to be hired, denial or revocation of a security clearance, damage to personal relationships when used in infidelity testing, heightened suspicion in criminal investigations, and severe psychological distress.

Research has shown that failed polygraph results can lead to false confessions in some circumstances [2]Verified Effects of Failed Polygraph Results on True and False Confessions
Demonstrates that failed lie detection results can lead to false confessions, relevant to understanding false positive consequences, and the same psychological pressure can apply when any credibility assessment technology produces a deceptive result for a truthful person. Under FBI polygraph policies, for example, 83% of actually innocent people are subjected to interrogation due to deliberate policy choices regarding test threshold settings [3]Verified Current FBI Polygraph/Interrogation Practices Puts the Actually Innocent at Risk of False Confession
Confirms 83% of actually innocent people are subjected to interrogation under FBI polygraph policies.

The Base Rate Problem in Screening

One of the most frequently overlooked aspects of false positive analysis is the base rate problem. In screening scenarios where the vast majority of people tested are truthful — such as pre-employment screening where perhaps 95% of applicants are honest — even a modest false positive rate produces a situation where a large proportion of all "deceptive" results are actually false positives.

For example, if 1,000 people are tested and 950 are truthful, a 12% overall error rate means a significant number of truthful people may be incorrectly flagged. The 2003 National Academy of Sciences report on polygraph testing emphasized this same mathematical reality, noting that polygraph screening "yields an unacceptable choice" between too many loyal employees falsely judged deceptive and too many security threats left undetected [4]Verified The Polygraph and Lie Detection
NAS 2003 landmark report finding polygraph accuracy insufficient for employee security screening, with wide variability in false positive rates. EyeDetect's chief scientist John Kircher has himself acknowledged this limitation, writing that "even if a test is 90 percent accurate, about 10 percent of the tested population would fail it, and the vast majority of those individuals who fail the test would be innocent" [5]Verified EyeDetect Marketer Converus Spends Hundreds of Thousands of Dollars Lobbying Federal Agencies
Reports on EyeDetect's accuracy limitations, unpublished Colombia field trial, and Kircher's acknowledgment that most people who fail a 90%-accurate test would be innocent.

Converus's Published Accuracy Claims

Converus, the company that manufactures EyeDetect, reports accuracy rates that vary by test type. According to Converus's official product page, scientific studies show EyeDetect has an accuracy range of 86 to 88%, depending on the type of test administered [6]Verified EyeDetect Product Page
Official Converus page reporting 86-88% accuracy for EyeDetect screening and diagnostic tests. These figures represent overall classification accuracy — the percentage of all examinees correctly classified.

Specifically, the Relevant Comparison Test (RCT) used for screening has a published mean accuracy of 86%, as reported by Kircher and Raskin in the European Polygraph journal in December 2016 [7]Verified Laboratory and Field Research on the Ocular-motor Deception Test
Kircher and Raskin's 2016 peer-reviewed publication reporting ODT accuracy rates from 78% to 86% and the RCT screening test mean accuracy of 86%. The Directed Lie Comparison (DLC) test for single-issue investigations shows 90% accuracy in field data [8]Verified EyeDetect Can Now Be Used for Single Issue, Diagnostic Tests for Investigations
Reports DLC investigative test accuracy at 90% based on field data, and EyeDetect overall accuracy range of 86-90%. The Multi-Issue Comparison Test (MCT) showed 88% mean accuracy in initial lab studies [9]Verified Credibility Assessment for Public Safety Agencies (EyeDetect Whitepaper)
Converus whitepaper reporting RCT accuracy at 86% (TN.89, TP.83), DLC accuracy at 90%, and MCT accuracy at 88%. For the upgraded EyeDetect+ automated polygraph, which combines ocular-motor data with traditional polygraph sensors, Converus reports 89–91% accuracy depending on test type [10]Verified Tech Co. Says its Advanced Lie Detector Will Save U.S. Gov't $29B Annually
Reports EyeDetect+ accuracy at 89-91%, over 1,000 customers worldwide, 50,000+ State Department-facilitated tests, and 12 peer-reviewed validation studies.

Understanding What Accuracy Figures Mean for Errors

When studies report an overall accuracy of 86–88%, this means approximately 12–14% of all test results are incorrect [11]Verified Eye Scan EyeDetect Lie Detector (British Polygraph Society Analysis)
Independent analysis confirming EyeDetect error rate of 12-14% based on 86-88% accuracy, compared to 5-7% for single-issue polygraph. This error rate is split between false positives (truthful people called deceptive) and false negatives (deceptive people called truthful). The distribution between these two error types is not always equal in published studies.

Converus has published some specific breakdowns. For the RCT screening test, one whitepaper noted true negative (correct identification of truthful people) at 89% and true positive (correct identification of deceptive people) at 83% [9]Verified Credibility Assessment for Public Safety Agencies (EyeDetect Whitepaper)
Converus whitepaper reporting RCT accuracy at 86% (TN.89, TP.83), DLC accuracy at 90%, and MCT accuracy at 88%. This means the false positive rate for truthful examinees in this protocol was approximately 11%, while the false negative rate for deceptive examinees was approximately 17%. These figures suggest the test is tuned to slightly favor identifying truthful people correctly.

The Foundational Research

The core scientific research behind EyeDetect was published by Anne E. Cook, Douglas J. Hacker, and colleagues in their landmark 2012 paper "Lyin' Eyes: Ocular-motor Measures of Reading Reveal Deception" in the Journal of Experimental Psychology: Applied [12]Verified Lyin' Eyes: Ocular-motor Measures of Reading Reveal Deception
Foundational peer-reviewed study reporting 85% accuracy in classifying guilty and innocent participants using ocular-motor measures. That study found ocular-motor measures "correctly classified 85% of participants as either guilty or innocent" [13]Verified New Study Shows Accuracy of EyeDetect Remains Unchanged Between English and Spanish-Speaking Participants
Reports ODT accuracy at 85% for both English and Spanish speakers, confirming cross-cultural generalizability. This foundational research was later expanded through additional studies conducted at the University of Utah.

The technology was further validated by Kircher and Raskin's 2016 publication in the European Polygraph journal, which reviewed laboratory and field research on the ODT and reported accuracy rates ranging from 78% to 86% based on logistic regression analysis [7]Verified Laboratory and Field Research on the Ocular-motor Deception Test
Kircher and Raskin's 2016 peer-reviewed publication reporting ODT accuracy rates from 78% to 86% and the RCT screening test mean accuracy of 86%. A 2016 cross-cultural study with Mexican participants confirmed the 85% accuracy rate generalized across English and Spanish speakers [13]Verified New Study Shows Accuracy of EyeDetect Remains Unchanged Between English and Spanish-Speaking Participants
Reports ODT accuracy at 85% for both English and Spanish speakers, confirming cross-cultural generalizability. Most recently, a 2024 chapter in "Modern Pupillometry" brought the total to 12 peer-reviewed publications validating the ODT protocol [14]Verified 12th Peer-Reviewed Publication Further Validates Eye Behavior as an Accurate Deception Detection Method
Confirms 12 peer-reviewed publications validating ODT protocol, 86-88% accuracy, 60 measurements per second, and 600+ customers as of June 2024.

Independent vs. Company-Affiliated Research

A critical consideration is the source of accuracy data. The vast majority of published research on EyeDetect originates from a closely linked academic-commercial group — primarily John C. Kircher, Anne Cook, Douglas Hacker, and David Raskin — who have disclosed financial interests in Converus [15]Verified Lyin' Eyes: Ocular-motor Measures of Reading Reveal Deception (ResearchGate)
Confirms authors' disclosed financial interests in Credibility Assessment Technologies, the predecessor company to Converus. While this does not automatically invalidate the research, it is an important factor. Truly independent, large-scale validation studies remain limited [5]Verified EyeDetect Marketer Converus Spends Hundreds of Thousands of Dollars Lobbying Federal Agencies
Reports on EyeDetect's accuracy limitations, unpublished Colombia field trial, and Kircher's acknowledgment that most people who fail a 90%-accurate test would be innocent.

To understand EyeDetect's place in the broader landscape of credibility assessment technologies, it is important to note that the National Center for Credibility Assessment, the federal agency that oversees polygraph exams in federal agencies, has not yet given EyeDetect a formal endorsement [16]Verified The eyes expose our lies. Now AI is noticing (CNN)
Reports that the National Center for Credibility Assessment had not yet endorsed EyeDetect as of 2017, and notes EPPA restrictions. Professor Ewout Meijer of Maastricht University has noted that while ocular-motor methods show promise, "there are boundary issues that mean this correlation between cognitive load and lying doesn't always apply" [17]Verified Lie Detector Firm Lobbies CIA, DOD on Automated Eye-Scanning Tech
Reports expert skepticism about cognitive load as a reliable proxy for deception, and EyeDetect+ claim of 90% accuracy.

How EyeDetect Measures Deception

EyeDetect operates on the principle that deception increases cognitive load — the amount of mental effort required to process and respond to questions. This increased cognitive load produces measurable changes in eye behavior that an infrared camera can detect [6]Verified EyeDetect Product Page
Official Converus page reporting 86-88% accuracy for EyeDetect screening and diagnostic tests. The system's high-definition infrared eye-tracking camera takes approximately 60 measurements per second of each eye [14]Verified 12th Peer-Reviewed Publication Further Validates Eye Behavior as an Accurate Deception Detection Method
Confirms 12 peer-reviewed publications validating ODT protocol, 86-88% accuracy, 60 measurements per second, and 600+ customers as of June 2024, collecting over 180,000 eye measurements during a single test [18]Verified EyeDetect – The First Business Intelligence Tool Providing Credibility Assessment Test Data
Confirms over 180,000 eye measurements collected per test and uploaded for cloud-based analysis.

The variables measured include pupil dilation, question response accuracy, question response time, gaze fixation, blink rates, and reading behavior [19]Verified FAQs for EyeDetect
Converus official FAQ confirming variables measured, scoring methodology, EyeDetect+ accuracy milestones, and combined testing confidence levels. The algorithm uses a logistic regression equation to analyze these eye behaviors and reading data, producing a Converus Credibility Score between 1 and 99 [19]Verified FAQs for EyeDetect
Converus official FAQ confirming variables measured, scoring methodology, EyeDetect+ accuracy milestones, and combined testing confidence levels. Scores from 50 to 99 indicate credible, while scores from 1 to 49 indicate deceptive [20]Verified EyeDetect Frequently Asked Questions (Global Data Fusion)
Confirms Converus Credibility Score ranges (1-49 deceptive, 50-99 credible) and scoring methodology details.

Why Truthful People Can Show Deceptive Patterns

The fundamental challenge is that increased cognitive load is not unique to deception. Many non-deceptive factors produce elevated cognitive load and corresponding changes in ocular-motor behavior. This is what creates false positives.

Test anxiety is a primary driver. A truthful person who is extremely anxious about the testing situation may show pupillary responses that mirror those associated with deception. The fear of being falsely accused creates genuine cognitive load unrelated to lying. Emotional reactivity to question content is another factor — questions about sensitive topics such as theft, infidelity, or drug use can produce strong emotional responses in truthful people, particularly those who have been falsely accused before. Research on innocent examinees in polygraph testing has demonstrated this phenomenon, with informed innocent participants showing larger electrodermal responses to critical items than uninformed innocents [21]Verified Leakage of relevant information to innocent examinees in the GKT
Demonstrates informed innocent examinees show larger physiological responses to critical items than uninformed innocents, relevant to false positive mechanisms.

Cognitive processing differences also play a role. People with ADHD, dyslexia, or other processing variations naturally show different reading patterns and fixation durations. Since EyeDetect requires reading text on a screen, any condition that alters reading behavior may affect the algorithm's measurements [5]Verified EyeDetect Marketer Converus Spends Hundreds of Thousands of Dollars Lobbying Federal Agencies
Reports on EyeDetect's accuracy limitations, unpublished Colombia field trial, and Kircher's acknowledgment that most people who fail a 90%-accurate test would be innocent. The algorithm is trained on population averages, and individuals who deviate significantly may be misclassified.

The Paradox in the Reading Data

Interestingly, Cook et al.'s original research revealed a paradoxical finding: guilty participants actually responded faster, made fewer fixations, and spent less time reading crime-related statements than truthful statements [12]Verified Lyin' Eyes: Ocular-motor Measures of Reading Reveal Deception
Foundational peer-reviewed study reporting 85% accuracy in classifying guilty and innocent participants using ocular-motor measures. This was contrary to the prediction that deception would produce slower, more effortful reading. The researchers proposed that guilty participants made a concerted effort to spend as little time on incriminating statements as possible to avoid detection [13]Verified New Study Shows Accuracy of EyeDetect Remains Unchanged Between English and Spanish-Speaking Participants
Reports ODT accuracy at 85% for both English and Spanish speakers, confirming cross-cultural generalizability. This paradox highlights unresolved theoretical tensions in how the algorithm interprets reading behavior, which may contribute to misclassification in some cases.

Psychological and Emotional Risk Factors

Several psychological conditions can elevate false positive risk. High test anxiety is among the most significant — people with generalized anxiety disorder show elevated autonomic arousal that can affect ocular-motor measurements. A history of false accusations may cause heightened emotional reactivity to test questions, even when answering truthfully.

PTSD and trauma-related hyperarousal affect the autonomic nervous system broadly, including pupillary response and attention patterns. Severe depression can alter cognitive processing speed and pupil responsiveness. Research on innocent persons in polygraph examinations demonstrates that emotional and physiological states can influence test outcomes even years after the relevant event [22]Verified The Validity of Polygraph Examination of an Innocent Person and a Perpetrator
Real-world case demonstrating polygraph validity with innocent and guilty subjects years after crime.

Cognitive and Neurological Risk Factors

ADHD affects fixation patterns, reading speed, and the ability to maintain consistent cognitive engagement with on-screen stimuli. Dyslexia and other reading disorders are particularly relevant because EyeDetect's reading-dependent format measures reading behavior as a key variable [19]Verified FAQs for EyeDetect
Converus official FAQ confirming variables measured, scoring methodology, EyeDetect+ accuracy milestones, and combined testing confidence levels. Autism spectrum conditions may produce atypical eye movement patterns that do not align with the neurotypical baseline the algorithm was trained on. Even mild traumatic brain injury can alter pupillary response, processing speed, and eye movement patterns.

Converus's unpublished Colombia field trial reportedly showed the test "was ineffective when we tested less well-educated applicants for security companies" — Kircher speculated reading problems may have been the cause [5]Verified EyeDetect Marketer Converus Spends Hundreds of Thousands of Dollars Lobbying Federal Agencies
Reports on EyeDetect's accuracy limitations, unpublished Colombia field trial, and Kircher's acknowledgment that most people who fail a 90%-accurate test would be innocent. This underscores the importance of reading proficiency to accurate results.

Physical and Environmental Risk Factors

Sleep deprivation significantly affects pupil dilation patterns, blink rate, and cognitive processing — all core EyeDetect measurements. Caffeine and stimulant consumption affects pupil dilation and can alter baseline autonomic arousal.

Eye conditions such as dry eye syndrome, glaucoma, cataracts, recent eye surgery, or certain contact lenses can affect the infrared camera's ability to accurately track eye behavior. Excessive mascara, eye liner, or false eyelashes may also be difficult to calibrate with the eye tracker, though removal usually resolves the issue [23]Verified Credibility Assessment for Public Safety Agencies Whitepaper
Details eye/vision conditions that may impact testing, suitability screening guidance, and countermeasure detection capabilities. Converus released EyeDetect 5.0 in August 2025 with two new pre-test suitability tests — a Pupil Light Reflex (PLR) test and a Horizontal Gaze Nystagmus (HGN) test — specifically designed to flag potential physical impairments and intoxication before testing begins [24]Verified Converus Releases EyeDetect 5.0 with New Suitability Tests to Boost Reliability
Confirms EyeDetect 5.0 includes PLR and HGN pre-test suitability tests, 1,000+ organizations worldwide, and 100+ U.S. law enforcement agencies.

For tips on preparing for any lie detector test, including managing controllable risk factors, consult our preparation guide.

Medications That Affect Pupil Dilation

Numerous medications have documented effects on the physiological parameters EyeDetect measures. Converus's own FAQ acknowledges that "any substance or medication that affects reaction time, eye movement or pupil dilation in general will affect comprehension, eye behavior and cognitive load" and that test results will be compromised in such situations [25]Verified EyeDetect Frequently Asked Questions – Medication Effects
Confirms any substance affecting reaction time, eye movement, or pupil dilation will compromise EyeDetect test results.

SSRIs and other antidepressants often cause pupil dilation through their effects on serotonin levels [26]Verified Dilated Pupils Drugs: Prescription, Nonprescription, and Recreational
Confirms SSRIs, tricyclic antidepressants, ADHD medications, and anticholinergics cause pupil dilation as documented side effects. Tricyclic antidepressants have anticholinergic effects that directly affect pupil dilation [26]Verified Dilated Pupils Drugs: Prescription, Nonprescription, and Recreational
Confirms SSRIs, tricyclic antidepressants, ADHD medications, and anticholinergics cause pupil dilation as documented side effects. ADHD medications such as methylphenidate and amphetamine significantly alter pupil dilation and attention patterns [27]Verified Dilated Pupils: Drugs, Side Effects, & Cautions
Confirms stimulants, hallucinogens, prescription medications including Adderall and antidepressants cause pupil dilation. Anti-anxiety medications like benzodiazepines can reduce autonomic arousal, potentially affecting the contrast between baseline and stimulus responses. Anticholinergic medications such as atropine and scopolamine directly cause mydriasis (pupil dilation) [28]Verified Pharmacologic Dilation of Pupil
Medical reference confirming atropine, scopolamine, amphetamine and other agents cause mydriasis through anticholinergic and sympathomimetic mechanisms. Opioid medications cause miosis (pupil constriction) that may interfere with dilation measurements.

Eye-Specific Medications and Conditions

Glaucoma drops such as timolol and pilocarpine directly alter pupil size and responsiveness. Cycloplegic drops paralyze the ciliary muscle and prevent accommodation, affecting pupil response. Post-LASIK or post-cataract surgery patients may have altered corneal properties that affect infrared tracking accuracy.

Converus's 2021 whitepaper for public safety agencies lists specific eye and vision-related conditions that may impact testing, including guidance on potential remedies [23]Verified Credibility Assessment for Public Safety Agencies Whitepaper
Details eye/vision conditions that may impact testing, suitability screening guidance, and countermeasure detection capabilities. However, the extent to which the algorithm can reliably adjust for all pharmacological effects is not fully documented in published research.

What to Disclose Before Testing

Before any EyeDetect examination, you should disclose all medications you are currently taking — including over-the-counter medications, supplements, and eye drops. You should also disclose any diagnosed eye conditions, recent eye surgery, neurological conditions, and psychiatric conditions.

Converus states that EyeDetect does not measure heart rate, respiration, blood pressure, or skin conductance, and conditions that affect those physiological measures do not generally impact EyeDetect test results [23]Verified Credibility Assessment for Public Safety Agencies Whitepaper
Details eye/vision conditions that may impact testing, suitability screening guidance, and countermeasure detection capabilities. However, any condition that significantly impacts mental acuity, physical functionality, or eye behavior is relevant. If the testing organization does not ask about medications and medical conditions before testing, this itself may be a concern about their testing protocols. Responsible EyeDetect operators should always conduct a thorough pre-test assessment, and the new EyeDetect 5.0 PLR and HGN suitability tests provide an additional safeguard [24]Verified Converus Releases EyeDetect 5.0 with New Suitability Tests to Boost Reliability
Confirms EyeDetect 5.0 includes PLR and HGN pre-test suitability tests, 1,000+ organizations worldwide, and 100+ U.S. law enforcement agencies.

Comparing Error Sources

Understanding how EyeDetect compares to traditional polygraph testing requires examining where each technology's errors originate.

With EyeDetect, error sources are primarily algorithmic — the machine learning model's inability to perfectly distinguish deception-related cognitive load from other sources of cognitive load. The automated scoring system eliminates examiner interpretation bias but cannot adapt to individual circumstances [6]Verified EyeDetect Product Page
Official Converus page reporting 86-88% accuracy for EyeDetect screening and diagnostic tests. The same algorithm is applied to every examinee regardless of their unique physiological profile, and reliability is 100% in the sense that the same data will always produce the same score [19]Verified FAQs for EyeDetect
Converus official FAQ confirming variables measured, scoring methodology, EyeDetect+ accuracy milestones, and combined testing confidence levels.

With polygraph, errors come from both physiological ambiguity and examiner interpretation. The 2003 NAS report found that the false positive rate "varied across interpreters by almost 0.50 — from 0.27 for the most conservative interpreter to 0.76 for the most liberal" [4]Verified The Polygraph and Lie Detection
NAS 2003 landmark report finding polygraph accuracy insufficient for employee security screening, with wide variability in false positive rates. A skilled polygraph examiner using APA-validated techniques can adapt questioning to individual circumstances, but this same flexibility introduces examiner bias as a variable. The APA's meta-analysis found an accuracy rate of 89% for validated polygraph techniques [29]Verified Do 'lie detectors' work? What psychological science says about polygraphs
APA report confirming NAS 2003 finding of 70% CQT accuracy and unknown false positive rate; APA meta-analysis found 89% accuracy for validated techniques, while the NAS found CQT polygraph could identify lies about 70% of the time [29]Verified Do 'lie detectors' work? What psychological science says about polygraphs
APA report confirming NAS 2003 finding of 70% CQT accuracy and unknown false positive rate; APA meta-analysis found 89% accuracy for validated techniques.

EyeDetect Advantages for Reducing False Positives

EyeDetect offers several structural advantages that can help reduce certain types of errors. No examiner interpretation bias means consistent scoring — the same data always yields the same result [19]Verified FAQs for EyeDetect
Converus official FAQ confirming variables measured, scoring methodology, EyeDetect+ accuracy milestones, and combined testing confidence levels. The less intimidating testing environment reduces some sources of anxiety-related false positives compared to a traditional polygraph exam with physical sensors [6]Verified EyeDetect Product Page
Official Converus page reporting 86-88% accuracy for EyeDetect screening and diagnostic tests. No physical attachment of sensors to the body makes the process less stressful for many examinees. Shorter test duration (15–30 minutes vs. 90+ minutes for polygraph) may reduce fatigue-related errors [6]Verified EyeDetect Product Page
Official Converus page reporting 86-88% accuracy for EyeDetect screening and diagnostic tests.

However, EyeDetect cannot adapt to individual medical or psychological conditions the way a skilled polygraph examiner can. Its reading-dependent format disadvantages those with reading difficulties [5]Verified EyeDetect Marketer Converus Spends Hundreds of Thousands of Dollars Lobbying Federal Agencies
Reports on EyeDetect's accuracy limitations, unpublished Colombia field trial, and Kircher's acknowledgment that most people who fail a 90%-accurate test would be innocent. And critically, the algorithm is trained on population averages, which means individuals who fall outside normal parameters may face higher error risk.

Using Both Technologies Together

For situations where false positive risk is a primary concern, many organizations now use both technologies together. Converus states that when EyeDetect and polygraph are used in combination and both tests agree, outcome confidence reaches approximately 97–99% [30]Verified FAQs for EyeDetect – Combined Testing Confidence
Confirms EyeDetect + polygraph combined accuracy of 97-99% when both tests agree, with mathematical derivation of joint probability. The mathematical basis: if the probability of a false negative error is 17% for EyeDetect and 20% for polygraph, the joint probability that a deceptive applicant would pass both tests is approximately 3.4% [30]Verified FAQs for EyeDetect – Combined Testing Confidence
Confirms EyeDetect + polygraph combined accuracy of 97-99% when both tests agree, with mathematical derivation of joint probability.

When the two tests disagree, however, additional investigation or retesting is warranted. Some law enforcement agencies use EyeDetect as a rapid first-stage screen and then refer flagged examinees for a full polygraph examination — a process that the VAST automated screening system has also been designed to facilitate.

Consequences of a False Positive Result

In employment screening contexts, a false positive can result in an applicant not being hired — often without being told the specific reason. Many employers treat a "deceptive" EyeDetect result as disqualifying, even though the technology's published accuracy shows a meaningful error rate. In criminal investigations, a false positive may increase suspicion and lead to more invasive follow-up investigations.

In relationship testing contexts, a false positive on an infidelity-focused EyeDetect test can cause severe damage to personal relationships. In sex offender monitoring, where PCSOT maintenance polygraph testing is standard practice, an incorrect deceptive result can trigger serious consequences including revocation of supervised release. Research confirms that understanding false confessions is critical in these high-stakes contexts, as confessions are often judged as honest regardless of their accuracy [31]Verified Understanding False Confessions
Demonstrates that confessions are judged as honest more frequently than denials, and observers fail to identify false confessions.

Steps to Take If You Believe Your Result Was Wrong

If you believe your EyeDetect result was a false positive, several options are available:

1. Request a copy of the test report, including your Converus Credibility Score. Scores close to the 50-point threshold (e.g., 40–49) carry more uncertainty than extreme scores.

2. Document any factors that may have affected the result — medications taken, sleep deprivation, anxiety levels, eye conditions, or reading difficulties.

3. Request a retest. Converus recommends a one-hour recess between tests. A diagnostic single-issue test following a screening test can help isolate specific concerns [20]Verified EyeDetect Frequently Asked Questions (Global Data Fusion)
Confirms Converus Credibility Score ranges (1-49 deceptive, 50-99 credible) and scoring methodology details.

4. Seek a traditional polygraph as a second-opinion assessment. When EyeDetect and polygraph disagree, this provides important additional data for decision-makers.

5. Consult with an attorney if the result has legal or employment implications. EyeDetect results are not admissible in most courts, and employers may have obligations depending on jurisdiction.

6. Request a post-test interview. Converus recommends post-test investigative interviewing for examinees who fail, noting this approach has been shown scientifically to improve information gain [20]Verified EyeDetect Frequently Asked Questions (Global Data Fusion)
Confirms Converus Credibility Score ranges (1-49 deceptive, 50-99 credible) and scoring methodology details.

Best Practices for Organizational Use

Organizations using EyeDetect should treat results as one data point in a broader evaluation, never the sole basis for employment or legal decisions. Conduct thorough pre-test screening for medications, medical conditions, and reading proficiency. Implement the new EyeDetect 5.0 suitability tests (PLR and HGN) to identify examinees who may not be suitable for testing at that time [24]Verified Converus Releases EyeDetect 5.0 with New Suitability Tests to Boost Reliability
Confirms EyeDetect 5.0 includes PLR and HGN pre-test suitability tests, 1,000+ organizations worldwide, and 100+ U.S. law enforcement agencies.

Consider using EyeDetect in combination with polygraph for higher confidence. Converus itself notes that EyeDetect is often used as an initial screening tool, with individuals flagged by EyeDetect then undergoing polygraph exams for deeper investigation [10]Verified Tech Co. Says its Advanced Lie Detector Will Save U.S. Gov't $29B Annually
Reports EyeDetect+ accuracy at 89-91%, over 1,000 customers worldwide, 50,000+ State Department-facilitated tests, and 12 peer-reviewed validation studies. This multi-stage approach helps minimize the impact of false positives from any single technology.

Organizations should also ensure their EyeDetect deployment aligns with best practices, including proper proctor training, standardized testing environments, and clear protocols for handling disputed results. In the United States, the Employee Polygraph Protection Act prohibits using lie detectors for employment-related testing within private companies, though federal, state, and municipal government employees or contractors are permitted [16]Verified The eyes expose our lies. Now AI is noticing (CNN)
Reports that the National Center for Credibility Assessment had not yet endorsed EyeDetect as of 2017, and notes EPPA restrictions.

The Importance of Context and Multiple Data Points

The history of credibility assessment demonstrates that no single technology provides perfect accuracy. The evolution of lie detection from Lombroso's plethysmograph through Mackenzie's ink polygraph to Reid's control question technique and now to ocular-motor testing reflects continuous improvement but never perfection.

The British Psychological Society's 1986 report on polygraph concluded that no single credibility assessment technique had sufficient scientific validity for standalone use [32]Verified Report of the working group on the use of the polygraph in criminal investigation and personnel screening
British Psychological Society concluded polygraph techniques lacked sufficient scientific validity for standalone use. This principle applies equally to EyeDetect. Research on choice blindness demonstrates that even explicit warnings may not fully eliminate cognitive biases in interpreting test feedback [33]Verified Choice blindness without deception: failures to notice false-feedbacks persist in explicit detection tasks
Demonstrates choice blindness persists even with explicit detection instructions, relevant to how test feedback is processed. Decision-makers should always consider the full context — including the base rate of deception in the population being tested, the specific test protocol used, and the individual examinee's circumstances.