Schrödinger’s Liar: Why Body Language Training Can’t Tell You What’s in the Box
The Three-Box Problem: An Investigative Thought Experiment
In 1935, the physicist Erwin Schrödinger imagined a cat sealed inside a box. Physicists still argue about what his thought experiment says about quantum mechanics. Here is the layperson version. From outside a sealed box, nobody gets to announce what’s inside. Someone has to open it.
Every interview you conduct is a sealed box. The statement a person gives you is true, deliberately false, or honestly mistaken (memory). You can’t tell which by looking at it through a body language lens. Body language training says you can. It teaches you to listen or stare at the lid: eye movement, posture shifts, a twitch across the face, and how it differs from what you asked someone before. It promises a read on what’s inside before anyone opens the box.
It doesn’t work, and the professional cost can be higher than most agencies and organizations would consider acceptable risk. This article explains why, and it presents to you a three-box sorting test you can run on any investigation and shows what opens the box instead.
What my wife saw that body language training experts missed
My wife watches police body-camera videos in the morning while she gets ready. I hate it.
She has never sat through an interview and interrogation class. She has no baseline to deviate from, no neurolinguistic programming cheat sheet, and no microexpression chart.
One morning, in one of her algorithm-delivered videos, an Officer was working on a non-injury collision between two drivers. One driver said the white car hit him and drove off. The woman driving the white car, who stopped some distance away, denied any involvement. But there was new paint transfer matching the colors of the two vehicles. She had been in the area she said. She said the damage on her car was old. It looked new to the officer in his experience.
My wife was frustrated. “Why isn’t he arresting her? She’s lying. Her story doesn’t match her own car. Look at the evidence! She said two completely opposite things.” I explained that with no injuries, the outcome in many jurisdictions is probably a citation or a summons with a court date. That wasn’t the interesting part.
The interesting part was what she noticed without any training. She never mentioned the woman’s eyes, her posture, stances, or the way she said something. She pointed at contradictions: what the woman said during the short interview versus what the physical evidence corroborated (statement-evidence contradiction). The woman is contradicting her own statement (statement in statement or within-statement contradictions). Those are the strongest indicators that a person isn’t telling you the truth. The suspect's statement on no collision and a freshly damaged bumper can’t both be true at the same time. Physical evidence, as they say, doesn't lie. It just is.
My wife was right...don't tell her I said that. The woman was lying. Witnesses, physical evidence, and the other driver’s account confirmed that the woman caused the crash and left. The physical evidence that was present corroborated the witness accounts and backed up getting to the truth of what happened.
The same statement worked in the other direction. The other driver said the woman hit him and drove off. The paint transfer, the damage, and the witnesses all backed him up. His statement was truthful, and it was confirmed without anyone reading his face, eyeballs, or his posture. Every detecting deception or body language course is built around catching lies. But testing a statement against the record doesn’t care who’s talking. It exposed the run driver's account as untruthful and verified the victim driver's statement, and nobody had to play psychic as to which was which. And any driver who has just been hit and left behind could easily look shaken or be nervous given the context. To someone trained to read nerves, that could look like deception; it doesn't discriminate. The real record cleared him anyway.
Now let's take a look from the officer’s perspective. He’s working next to live commuting traffic, keeping an eye on the run driver who might leave, collecting statements from everyone, and watching out for his safety. Nobody can read demeanor in that setting, and he didn’t need to. The woman’s statement is on his body camera. A supervisor, a prosecutor, a judge, or a jury can watch the video, listen to the statements, look at the bumper, and decide for themselves, just like my body-worn camera footage-addicted wife. Nobody has to trust his reading of anything based on demeanor.
Swap the traffic scene for a conference room. The bumper becomes a badge log, an email timestamp, a security video, or an outcry witness. The officer becomes an HR investigator, a compliance officer, a corporate security manager, or a loss prevention investigator. The structure is identical. The statement goes on the record. The record either supports it or it doesn’t.
The three-box problem
Every investigative interview produces output. A statement or more precisely, statement evidence that goes on the record.
Box 1 is a statement. This is what any person says, on the record: what they were capable of when they had the opportunity, who was there, what actions were taken, locations, timelines/times, and things that were said. Statement evidence is the answer to your questioning objectives built during interview planning. When an investigative interview is done well and done using science-based methods, you will get a lot of answers and a lot more cooperation and information compared to accusatorial methods.
Box 2 is demeanor judgments. This is reading body language and demeanor: nervousness, posture, eye contact, fidgeting, tone of voice, and how someone sits or what they do with their arms/hands. It also covers the named pseudoscientific methods built on some of those reads. Microexpressions. Baselining and “deviation from baseline.” Behavior analysis interviews. Behavior-provoking questions and “shock” questions designed to make someone react so you can read the reaction. Neuro-linguistic programming cues. Body language training lives here, whatever it’s called on the brochure and no matter who sells it.
Box 3 is where the decisions are made and where the risk lives. It’s what you can swear to and what you act on. In law enforcement, that’s the search warrant affidavit, the charging decision, and the arrest. In the private sector, it’s the termination, the suspension, the written warning, the demotion, the decision to clear someone, and the finding in a workplace investigation report. It’s also the arbitration brief, your testimony, and the deposition. Box 3 is where someone’s job, freedom, or reputation is on the line, and where your organization carries legal and financial exposure if you get it wrong. Everything in Box 3 has to survive someone else looking at it.
And Box 3 is just as much about clearing people as it is about holding them accountable. An investigation isn’t only there to prove someone guilty. It’s there to find out what happened. That means exonerating the innocent and the wrongfully accused, whether the accusation was malicious or just a misunderstanding. A finding that clears someone has to be as solid as one that charges or fires them, because that person’s job, freedom, or reputation is on the line too. Your organization carries legal and financial exposure either way if you get it wrong. Again, everything in Box 3 has to survive someone else looking at it closely.
Here’s the operational problem. Box 1 feeds Box 3. A statement tested against the record produces facts you can use: this includes a statement that contradicts a document, a video, a witness, or the person’s own earlier account. Box 2 doesn’t feed anything. Nothing in it can or should cross into Box 3 alone. Although interview and interrogation trainings are loaded with body language folklore that edges out important Box 1 skills that move investigation forward.
I use a simple, albeit non-scientific, test in classes and at conferences. Would you testify that a microexpression proved someone was lying? Would you put it in the affidavit? Would you say it to your boss and defend it in a deposition? How about a cluster of them? BTW, WTH is a cluster, for the record, of course? For the last three years I’ve asked rooms full of investigators. I’m still waiting for the first hand to go up. That’s a very informal survey for sure but I can guarantee your body language guru/trainer will not be there to support you.
The Three Box math
Box 1 is the statement. It can be tested and corroborated against the record: the questions, the listening, the checking, and the facts. It’s the foundation of investigations, police work, and communication.
Box 1 = Box 3
Box 1 + Box 2 = Box 3
Box 2 ≠ Box 3
Box 1 can carry a decision by itself. Box 2 can sit on top of it and may add a little context, but it doesn’t change the answer and shouldn't. On its own, it can’t carry a Box 3 decision at all. Nor should it; that will get you on Netflix. Box 2 is mostly noise, from an investigative perspective, yet it gets a large share of the training dollars and conference time in this field because it seems scientific.
Here is what that looks like in a report. These are invented case examples to illustrate my point.
Box 1 = Box 3
Finding: The respondent said he was at his desk from 2:00 to 4:00 p.m. on March 3. Badge records show him leaving through the east door at 2:14 p.m. and returning at 3:51 p.m. Parking garage video shows his vehicle leaving at 2:19 p.m. When asked about the gap, he said he “was sure he was at his desk.” A coworker confirmed he was not at his desk.
His own statement, tested against the badge log, the video, and a witness, supports the finding. Nothing else is needed.
Box 1 + Box 2 = Box 3
Finding: [same facts as above] Investigator’s note: The respondent appeared nervous, crossed his arms when answering questions, and avoided eye contact during the interview.
Delete the note and the finding doesn’t change. Box 2 added nothing the case needed.
Box 2 ≠ Box 3
Finding: The respondent was visibly nervous, shifted in his seat, and gave short answers. In my experience, these are signs of deception I was told.
There is nothing in Box 1 to test. If this memo ends up in arbitration or a lawsuit, the decision has nothing under it, and the risk falls on whomever signed off on it.
The same math on a traffic stop
Basis for extending the stop: The driver appeared nervous. The car was registered to a third party. His stated travel plans did not match the rental agreement or statements. He had a prior drug arrest for trafficking.
Take out “appeared nervous," and the other factors still carry the decision. Take out the other factors and nervousness can’t carry it alone. The factors that held up came from asking questions, investigative experience, and checking the answers. That’s Box 1.
Why investigators still see the pattern
We’re investigators, and finding patterns is the job. Humans are good at it, and that’s why we’re good at this work. But sometimes the patterns are wrong. The same skill that spots a real connection can find one that isn’t there.
There’s a name for the second kind: illusory correlation. People see a link between two or more things that aren’t related, especially when the link matches what they already believe. The classic work comes from Chapman and Chapman in the late 1960s. They found that beginning clinicians reported links between signs and traits that came from their own expectations, not from the data in front of them. That was for both students and for practicing clinicians alike.
Here’s how it plays out in an interview room. An investigator who believes liars look away during lies will notice and remember the times someone looked away and supposedly lied. Those instances are remembered. The times someone looked away and told the truth don’t register. Neither did the liars who held eye contact, because even liars know pop psychology, nor did the truthful people who did either. I have yet to run across any investigator, including myself, who keeps track of their actual batting average when using these lie detection techniques. They remember the hits only, not the strikeouts. This is a situation that is ripe for bolstering biases and cultivating false investigative information.
This is how a lie detection cue becomes “proof” in your head when nothing has been tested. Years of interviews seem to confirm it, and the confirmation comes from the same memory that was doing the sorting. It’s also why experience alone doesn’t fix it. More interviews give you more hits to remember, and the pattern feels stronger, confirming a false belief in lie detection myths with every new investigation.
Why Box 2 can’t and shouldn't cross into Box 3 alone
It’s close to a coin flip
Across more than 200 studies, people correctly sorted truths from lies about 54% of the time, where guessing gets you 50% (Bond & DePaulo, 2006). Professionals such as police investigators did no better than students. When the U.S. Government Accountability Office audited the evidence behind the behavioral indicators the TSA used to screen travelers, it found that 175 of the 178 sources cited did not provide valid evidence for the indicators they were supposed to support (GAO, 2017). Those 175 sources are what the lie detection trainers cite as evidence for their own special proprietary systems.
So why does Box 2 get so much attention?
Because of how the story gets told afterward. Box 1 does the work: the questions, the portion of the investigation that got you there to begin with, the listening, the checking, and the facts. Then someone remembers that the driver looked nervous or the employee wouldn’t make eye contact, or you think you saw a microexpression, and that cue gets linked up with the facts of the case. In hindsight, the cue looks like it was part of the answer all along. The case came together, so the "tell" must have mattered. Box 2 takes credit for the heavy lifting that Box 1 had already done.
Here is my argument. Remove anything from Box 2 and look at what’s left. Your traffic stop, your investigation, and your findings from Box 1 still carry the day. The facts, the record, and the statements that didn’t match are all still there. Box 2 still gets far more credit than it earned, and far more than your context, your experience, your intuition, and the work you put into the interview. That’s where the myths about body language come from. They aren’t built on studies. They’re built on stories where a "cue" showed up next to the facts and got the credit.
I watched this play out at a conference. A speaker described how a body language expert had caught someone in a notorious case, and it came down to one tell that gave it all away. The room was amazed. One thing, and that was how the person was caught. Period.
I had a question: why were you looking at this person in the first place? Something put that individual in front of the "expert." A tip, a witness, a record, an investigation, and an interview that went somewhere. That’s Box 1, and the story skipped it. I leaned over and asked the man next to me, and he said, “Exactly.” There was zero context, and we were supposed to believe that this guru was the one who broke it all open with a single observation.
I can’t tell you the story was false. I can tell you it was missing everything that mattered. It started after Box 1 had done its work, and it ended with a body language cue taking the credit. I mean, this cue was observed with a surveillance team already in place. Doesn't that tell you they already had some information on who or what to surveil? That’s the whole pattern of this body language training charade in one story. And somebody in that room walked out ready to buy the next lie detection course instead of investing in better communication, investigations, and interviewing.
The next time you hear a mythical-sounding lie detection story, ask the same question: why was that person there in the first place?
Run the removal test on your own investigative cases.
There’s a name for this error: post hoc ergo propter hoc, “after this, therefore because of this.” On a break in one of my classes, a drug interdiction officer asked me about it. His stop had gone the way these stories usually go: a quick driver turnaround in a border city, masking smells, car registered to a person unknown to the driver, a lived-in look, and afterward, “he was nervous, and there were drugs in the car.” The nervousness was highlighted, and then the drugs turned up, so the nervousness gets the credit as a cue.
So we ran the test. Take the nervousness out. What was left? A fast turnaround, a lived-in look to the car, masking smells, and a car registered to someone the driver couldn’t account for. The case was still there, and it was most of the case. His case came from the facts and the roadside interview, a great stop! Every one of those facts came from looking at the car, experience, asking questions, and checking the answers. That’s Box 1.
Now try the same thing in reverse. Leave the nervousness and take the other four out. It is a lot shakier ground to extend a stop or make an arrest on “he seemed nervous." My kid had an officer get behind him on the way home from work, and he said he was "really nervous" because he didn't want a ticket. Take Box 2 out of your last case and ask whether it still stands without Box 1.
And if you saw nervousness, put it in your report. Write down what you observed, with all the other facts. It’s an observation; it is what you do, and a court will let you say it. Just don’t think it is going to carry the case, and don’t let it take credit for the facts and what your interview and investigation already did.
Box 3 without Box 2 = Box 3 (investigation still stands)
Box 3 without Box 1 = nothing (investigations fall apart)
If removing something doesn’t change the outcome, it didn’t cause the outcome. Box 2 fails that test. Box 1 passes it. Tacitus described the pattern nearly two thousand years ago: when things go well, everyone claims the credit, and when they go badly, the blame falls on one. In an investigation, Box 2 is the one who shows up to claim the win. Box 1 did the work.
If you want more on what the research says about workplace interviews specifically, start with Can You Tell If an Employee Is Lying? and Body Language and Lie Detection Myths.
Some courts and regulators don’t treat it as proof
The Kansas Supreme Court reversed the convictions in State v. Elnicki (2005) after jurors watched a videotaped interrogation in which a detective told the defendant, “When somebody lies, their eyes shift.” The court held that jurors should not hear that kind of statement, from a witness stand or from a videotape, and sent the case back for a new trial (opinion).
A federal appeals court took apart a nervousness read in United States v. Beck (1998). The officer himself testified that in roughly a quarter of his traffic stops, the driver was at least as nervous as this one. He had also never met Beck before, so he had nothing to compare Beck’s behavior against (context). The court found the read could not justify extending the stop and reversed the ruling that let the evidence in (opinion).
The workplace version is the EEOC’s guidance on harassment investigations. It lists demeanor as one of five credibility factors, alongside plausibility, motive to falsify, corroboration, and past record. Then it says none of the factors is determinative (EEOC, 1999). Demeanor can be considered. It can’t carry a finding.
Box 2 sometimes points in two contradictory directions
Jeffrey Deskovic was a suspect at 16 in part because he seemed overly distraught about the victim’s death. He visited the victim's wake three times. He was convicted and later exonerated by DNA (National Registry of Exonerations). Michael Morton was suspected in his wife’s murder in part because he seemed to show too little emotion. He, too, was convicted and later exonerated (Colloff, 2012).
One man looked guilty because he showed too much. The other looked guilty because he showed too little. That isn’t a coincidence, because some interview and interrogation curriculums teach it both ways. Too calm can mean deception. Too emotional can mean deception. Too much eye contact is a sign, and so is too little. When every behavior counts as a clue, no behavior can count against your theory. A rule that convicts in both directions isn’t a rule. It’s a story you tell after you’ve already decided.
That's the real danger to being a more effective investigator. When you go looking for liars with lie detection or detecting deception training, you don’t discover your suspicion. You bring guilt-presumption with you, and every reaction confirms it. Confirmation bias gets stronger, and the “tells” you think you saw become false case information that goes into your notes, your report, and the next interview. It’s an unforced investigative error: one caused by your own training and experience, not by a bad outside actor or source. No witness lied to you and no evidence misled you. You supplied the mistake yourself.
Here’s a question: have you ever seen a course on truth detection? Almost every program in this space is sold as lie detection: how to catch the person who is lying. But an investigation has to know when someone is telling the truth too, because that’s how the innocent get cleared. In Science-Based Interviewing, we don’t try to read the truth off a person’s face, posture, or tick. We teach how to get a full account and test it against the record, so a truthful statement gets confirmed and an untruthful one gets exposed. Both come out of the same process: better interviews.
Box 2 costs you Box 1
Box 2 isn’t harmless noise. It takes things from Box 1.
It takes attention. Every minute you spend watching for a tell is a minute you aren’t listening. The best follow-up questions and the path to more information comes from what the person just said, not which way their eyes were pointing.
It changes the interview. Behavior-provoking questions and “shock” questions don’t just observe a reaction. They create one. Then you read the reaction you caused.
It raises confidence without raising accuracy. In one study, people trained with a popular accusatory program’s own materials judged truth and deception less accurately than people with no training. They were also more confident (Kassin & Fong, 1999). Investigators with that kind of training and experience leaned toward calling people deceptive without getting better at telling who was (Meissner & Kassin, 2002). A review of 30 training studies found that teaching nonverbal cues did little, while training that focused on what people say did better (Hauch et al., 2016).
Once you’ve decided someone is lying, you stop collecting information and start collecting confirmation. That’s how a Box 2 read quietly damages Box 1. It matters most with a complainant describing something traumatic. Distress, flat affect, and a jumbled sequence can be what trauma looks like. They aren’t reliable signs of lying. This is also why investigative bias deserves its own attention in any workplace investigation.
What fills Box 1
If Box 1 is the input, the skill is getting more information and getting information that is more accurate. That’s a teachable interviewing skill. It’s also rarer than you’d expect. In a study of 80 real police interviews, the interviewers did most of the talking, and open-ended questions made up less than 1% of what they asked (Snook et al., 2012).
The methods that fill Box 1 are well documented and field-tested:
ORBIT, rapport and active listening, built to get an accurate account, not a confession. In field research with trained investigators, science-based tactics were linked to more cooperation and more information, while accusatory tactics were linked to less cooperation (Russano et al., 2026).
Planning your questioning objectives around the evidence you hold, meaning the topic areas you need covered, such as where someone was during a specific afternoon (see Strategic Use of Evidence below).
Cognitive interviewing, which asks memory-compatible questions. A major meta-analysis found a large increase in correct details with accuracy essentially unchanged (Memon et al., 2010).
Strategic Use of Evidence (SUE). It’s a planning and questioning technique, not just “late disclosure.” In a controlled study, interviewers trained in it told liars from truth-tellers far more often than untrained interviewers did, about 85% versus 56% (Hartwig et al., 2006).
Red teaming your own case, so your assumptions get challenged before a decision-maker does it for you.
Notice the pattern. Every method with support examines the account or the evidence. None of them examines the person’s hands, eyes, ticks, or other tea leaves.
Before you buy the training: five questions
If you’re a training director, an HR leader, a compliance chief, or a police executive, these questions sort Box 1 programs from Box 2 programs quickly.
What’s the false positive rate? How often does the method label a truthful person deceptive? If the answer is "we don't know," nobody has measured it.
What’s the full batting average? A highlight reel of hits isn’t an accuracy rate; it is a sales pitch. Ask for the misses too. If they don't have them, they don't have a batting average.
Can it be taught to someone who isn’t the instructor? If the answer is twenty years of experience, I wrote a book on it, etc., etc., you’re buying a résumé, not a method that can't be transferred to others. Ask whether it works in settings the instructor has never seen.
What does the curriculum teach about memory-compatible questioning, cognitive interviewing, Strategic Use of Evidence, and red teaming? If those chapters are missing, most of the course is Box 2.
Who’s the arbiter? When two instructors disagree, whose résumé wins? The only fair answer is the research. The arbiter is the science
Two more warnings. First, watch for rebranding. Many programs now say "non-accusatory" and "non-coercive," but if a big share of the course is baselining, behavior provoking questions, deviation from baseline, behavioral cueing, and reading deceptive answers, it’s the same Box 2 in new packaging. Not long ago someone sent me a clip of a well-known program being taught to a corporate audience. The instructor waved off who, what, when, where, and why as a footnote; yeah, we do all that, then moved straight to judging guilt from the way people answered questions. In box terms, he skipped Box 1, the most important input to your case files and built the whole class on Box 2.
Second, if a program claims it can assess the credibility of victims describing traumatic events, treat that as a red flag. Ask for the receipts. A science-based program can show you its research, its error rates, and its limits.
Where to start
Sort your last report. Take the findings in your most recent investigation and mark each one: Box 1 or Box 3 (a statement, or something you can corroborate), or Box 2 (an impression). Anything in Box 2 comes out. If the report falls apart without it, you’ve found the gap.
Ask the raised-hand question. In your next training session, ask who would testify to a microexpression, or any of them, as proof of lying. See how many hands go up.
Spend your next interview on Box 1. Get a full account in the person’s own words, and test what you hear against the record. Write down contradictions as facts, and treat the reason for each one as a question still to ask.
If you want to build better interviewing skills, backed by research, IXI teaches them. Start with workplace investigations training, see what’s coming in online training, or browse the workshops. For the research behind the method, begin with science-based interviewing and the post on why you can’t learn lie detection from your own mistakes.
Better interviewing carries the day; dump the junk science.
Better Interviews. Better Investigations.™
Frequently asked questions about body language training
Does body language training work?
Not for detecting deception, your intuition alone is better. Across more than 200 studies, people were right about 54% of the time, barely above a coin flip, and professionals did no better than students. Training that teaches nonverbal cues has shown little benefit, and some versions have made people more confident without making them more accurate.
Can body language tell you if someone is lying?
No reliable cue has been found. People are nervous in investigations for many reasons: fear of losing a job, damage to their reputation, or worry about the process. A truthful person and a deceptive person can look identical. What you can do is test what they say against the record.
Are microexpressions reliable?
Research on microexpressions found them to be rare and nearly as common in truthful people as in deceptive ones (Porter & ten Brinke, 2008). I have yet to meet an investigator who would testify to one as proof of lying.
What should investigators use instead of body language?
Methods that build the account and test it: rapport, cognitive interviewing, active listening, planning your questioning around the evidence you hold, strategic use of evidence, and red teaming.
Can demeanor be considered at all?
It can be a prompt. Visible distress is a reason to slow down, check in, and ask another question. It isn’t evidence of deception. The EEOC lists demeanor as one credibility factor and says none of its factors is determinative.
Body Language Training Sources
Bond, C. F., & DePaulo, B. M. (2006). Accuracy of deception judgments. Personality and Social Psychology Review, 10(3), 214–234. https://doi.org/10.1207/s15327957pspr1003_2
Hartwig, M., Granhag, P. A., Strömwall, L. A., & Kronkvist, O. (2006). Strategic use of evidence during police interviews: When training to detect deception works. Law and Human Behavior, 30(5), 603–619. https://doi.org/10.1007/s10979-006-9053-9
Hauch, V., Sporer, S. L., Michael, S. W., & Meissner, C. A. (2016). Does training improve the detection of deception? A meta-analysis. Communication Research, 43(3), 283–343.
Kassin, S. M., & Fong, C. T. (1999). “I’m innocent!”: Effects of training on judgments of truth and deception in the interrogation room. Law and Human Behavior, 23(5), 499–516.
Meissner, C. A., & Kassin, S. M. (2002). “He’s guilty!”: Investigator bias in judgments of truth and deception. Law and Human Behavior, 26(5), 469–480.
Memon, A., Meissner, C. A., & Fraser, J. (2010). The Cognitive Interview: A meta-analytic review and study space analysis of the past 25 years. Psychology, Public Policy, and Law, 16(4), 340–372. https://doi.org/10.1037/a0020518
Porter, S., & ten Brinke, L. (2008). Reading between the lies: Identifying concealed and falsified emotions in universal facial expressions. Psychological Science, 19(5), 508–514.
Russano, M. B., Meissner, C. A., Jones, N. J., Rothweiler, J. N., Taylor, S., Cory, C., & Brandon, S. E. (2026). Evaluating the effectiveness of a practitioner-designed science-based interviewing and interrogation course. Legal and Criminological Psychology, 31, 273–300. https://doi.org/10.1111/lcrp.70021
Schrödinger, E. (1935). Die gegenwärtige Situation in der Quantenmechanik. Naturwissenschaften, 23(48), 807–812.
Snook, B., Luther, K., Quinlan, H., & Milne, R. (2012). Let ’em talk! A field study of police questioning practices of suspects and accused persons. Criminal Justice and Behavior, 39(10), 1328–1339. https://doi.org/10.1177/0093854812449216
U.S. Government Accountability Office. (2017). Aviation security: TSA does not have valid evidence supporting most of the revised behavioral indicators used in its behavior detection activities (GAO-17-608R). https://www.gao.gov/products/gao-17-608r
U.S. Equal Employment Opportunity Commission. (1999). Enforcement guidance: Vicarious liability for unlawful harassment by supervisors (EEOC-CVG-1999-2). https://www.eeoc.gov/laws/guidance/enforcement-guidance-vicarious-liability-unlawful-harassment-supervisors
State v. Elnicki, 279 Kan. 47, 105 P.3d 1222 (2005). https://law.justia.com/cases/kansas/supreme-court/2005/89003.html
United States v. Beck, 143 F.3d 1439 (8th Cir. 1998). https://caselaw.findlaw.com/court/us-8th-circuit/1455282.html
National Registry of Exonerations. Jeffrey Deskovic. https://exonerationregistry.org/cases/10405
Colloff, P. (2012, November). The innocent man, part one. Texas Monthly. https://www.texasmonthly.com/politics/the-innocent-man-part-one/
Comments