What Is Science-Based Interviewing? (And Why It Matters)

If you search "how to interview a suspect" or "interrogation techniques," you'll find no shortage of answers. Some come from retired detectives. Some come from body language experts. Some come from sales trainers who've repackaged persuasion tactics as investigative tools. There is even some really terrible advice that comes from Google and AI on how to interrogate. Some come from those who claim to teach research-backed methods. More come from people who simply worked a big or important case, had a successful interview, and, voila, an instructor is born. However, almost none of them will tell you what the research actually says.

Investigations should be about more reliable information and more of it, not a narrow focus on a confession alone.

Science-Based Interviewing is different—not because it sounds better, but because it is built on a different foundation entirely. It is an information-gathering approach focused on the quality and quantity of statement evidence versus being narrowed to a confession. That foundation is empirical research: peer-reviewed studies, field validation, and decades of psychological science on memory, communication, deception, and human behavior under pressure.

This post explains what science-based interviewing is, where it came from, why the old methods are inadequate, and why the interview and interrogation training distinction matters for law enforcement investigators, HR professionals, corporate security officers, and leaders who have to have difficult conversations.

The Problem Science-Based Interviewing Was Built to Solve

For most of the twentieth century, accusatory, confession-driven methods dominated investigative interviewing in the United States. The Reid Technique is the most widely recognized name in this tradition—developed in the 1950s, commercially taught for decades, and deeply embedded in how law enforcement has historically approached interviews, interrogation, and detecting deception. However, it would be misleading to treat Reid as the whole story. The market for interrogation training has produced a range of derivative and hybrid curricula—programs that often self-describe themselves as "Reid-style" (although not the technique) or that follow a structurally similar path under different branding. What these programs share, regardless of name, is a confession-driven architecture: identify guilt through behavioral assessment, confront the subject, apply psychological pressure, and move them toward admission. None of these approaches—the original technique or its many derivatives—have been subjected to independent, peer-reviewed empirical validation. The research that does exist on confession-driven, accusatory interrogation methods raises serious concerns about false confessions, coercion, and the reliability of the behavioral cues on which guilt presumption is based.

But the deeper problem with accusatory, confession-driven methods is not just what they get wrong. It is what they ignore entirely.

The confession obsession—the belief that a successful interrogation ends in an admission—is so deeply embedded in traditional interview and interrogation philosophy that it crowds out every other metric that actually matters to an investigation (Hartwig & Cory, 2025). I've stood in front of many rooms full of experienced investigators—from the Southern California Fraud Investigators Association to the International Homicide Investigators Association and training classes across the country—and I've asked all of them the same question: have you ever won a case without a confession? Every room answers the same way. Hands go up. A lot of them. Every time. Battle-tested investigators who have been in the field long enough know that confession is but one data point, not the whole picture. Why would you select an interviewing paradigm that limits you?

The metrics that also determine the value of an interview are broader and sometimes more consequential:

• Information yield—the quantity and quality of case-relevant detail gathered during the interview. Simply put you have more to work with.

• Statement-to-statement consistency—whether the account holds together internally across the interview

• Statement-evidence inconsistencies—contradictions and also called false exculpatory statements (FES), between what a person says and what the evidence independently shows, which can be just as powerful in court even without a confession (Brimbal & Jones, 2018).

• Cooperation—whether the subject remains engaged, continues providing information (even after the first interview), and experiences the process as fair also increases information yield

• Rapport—the degree to which the interviewee trusts the process enough to speak openly, which directly determines the depth and accuracy of everything else on this list and improves information yield

There is also a timing problem that confession-driven thinking ignores. Forensic results are slow. DNA, toxicology, drug panels, digital forensics, and lab analysis can take weeks or months to come back — and when they do, they may prove you right or they may prove you wrong. The interview, in some ways, can be perishable. The window when a subject, victim, or witness is willing to talk and able to remember is open now, and it closes. Memories fade. Once the adrenaline of an event passes, people care less, move on, and grow reluctant to get involved. You often cannot go back for that interview later. So the strategy is simple: get the most complete, most accurate information you can while you have the chance — because the evidence that confirms or contradicts it may not arrive until long after that chance is gone.

Accusatory methods are structurally blind to all of these factors. By narrowing the definition of success to a single outcome, explicitly shutting down new information, and then building a method around it, they sacrifice information, cooperation, and ultimately and ironically confession integrity in pursuit of an admission that may never come—and when it does come under coercive conditions, it may not hold up. What is your plan if they don't confess?

The field validation evidence makes this comparison concrete. When science-based methods are studied in real police cases — not laboratory simulations — the results are consistent: information yield increases, cooperation increases, and confessions do not drop. In fact, they often increase as well (Russano et al. 2024; Russano et al, 2026). Science-based methods do not trade confessions for information. They produce more of both. They beat accusatory styles across every meaningful interview metric—and they beat them decisively.

And here is the uncomfortable truth for those who resist this shift: it requires believing that law enforcement cannot upgrade how it communicates with people—that the 1950s playbook is still the pinnacle of law enforcement communication excellence. That is not true, and the private sector proves it. Sales, marketing, and leadership have spent decades refining how they communicate, build customer relationships, and build trust. They have invested in research, evolved their methods, and abandoned what the evidence showed did not work. They do not still sell the way they sold in 1955. Law enforcement has the same opportunity. The research is there. The methods are validated. The only thing standing in the way is the institutional inertia of a tradition that was never grounded in science to begin with.

Confession-driven, guilt-presumptive approaches have three additional specific failures that compound the problem. First, it starts with a presumption of guilt. The investigator has already decided—based on unvalidated behavioral cues, intuition, or early case information, sometimes all three—that the person being interviewed is responsible. Everything that follows is designed to confirm that conclusion rather than test it. This is confirmation bias baked into the methodology and/or step-by-step process itself.

Second, it relies on unvalidated behavioral lie detection. Accusatory models train investigators to read deception from body language, eye movement, posture, nervousness, and speech patterns. The research on this topic is unambiguous. Bond and DePaulo's landmark 2006 meta-analysis of over 200 studies and 24,000 judgments found that people correctly identify truth and deception at a rate of approximately 54 percent — barely better than a coin flip. Trained law enforcement professionals did not perform meaningfully better than untrained civilians. The behavioral cues that investigators rely on most confidently have no reliable relationship to deception in the scientific literature.

And yet some vendors — including the big ones — market these "credibility" assessments as appropriate for victims and witnesses. Never mind that trauma reshapes memory and demeanor; that culture governs eye contact, expressiveness, and body language; and that cognitive and neurological conditions routinely produce the very "deception cues" these tools claim to decode. This is where investigative failures such as false confessions begin: with forced investigative errors built on bad tools.

Third, confession-driven methods produce false confessions. This issue is not just theoretical. The Innocence Project has documented hundreds of wrongful convictions in which DNA evidence exonerated individuals who had confessed. Many of those confessions were produced by the same high-pressure, guilt-presumptive tactics that accusatory training teaches as best practice. This despite early in my career being told by some of these instructors that "no one would ever confess to a crime they did not commit." Ooops.

Science-Based Interviewing emerged as the evidence-based answer to these failures.

Where Science-Based Interviewing Comes From

Science-Based Interviewing is not a single technique created by a single person. It is a framework built from converging lines of psychological research, developed in America, Europe, and Scandinavia over the past three decades, and increasingly adopted by law enforcement agencies, militaries, intelligence organizations, and private-sector investigators around the world.

Several research programs and methodologies form the core of what we now call science-based interviewing.

The Cognitive Interview, developed by Ronald Fisher and Edward Geiselman in the 1980s, drew on memory research to produce a technique specifically designed to maximize the quantity and quality of information retrieved from cooperative witnesses and victims. Rather than asking yes-or-no questions or leading questions that shape memory, the Cognitive Interview uses free narrative, context reinstatement, and mental reinstatement of the scene to access more complete and accurate recall.

The Strategic Use of Evidence (SUE) technique, developed by Pär Anders Granhag and Maria Hartwig at the University of Gothenburg, gives investigators a research-supported method for managing evidence during interviews. Instead of confronting a subject with evidence immediately—which allows them to shape their account around what the investigator has revealed—SUE uses strategic questioning to allow subjects to commit to an account before evidence is introduced. Truthful people tend to be consistent with the evidence. Deceptive people tend to generate statement-evidence inconsistencies that become meaningful and testable. SUE is not merely late disclosure; it is a planning technique, a questioning technique, and an evidence strategy.

ORBIT (Observing Rapport-Based Interpersonal Techniques), developed by Laurence and Emily Alison and their colleagues at the University of Liverpool, examined what actually works in high-stakes interrogations with resistant and uncooperative subjects — including terrorists. Their finding was clear: rapport-based, non-confrontational approaches produced significantly more information yield than coercive, accusatory tactics. Resistance increases under pressure. Information increases under rapport.

These are not competing methods. They are complementary research programs that together help define what science-based investigative interviewing looks like in practice.

What Science-Based Interviewing Actually Does

At its core, science-based interviewing does four things differently from traditional accusatory interrogation.

1. It starts with information, not confession. The goal of a science-based interview is not to obtain an admission. Its goal is to gather accurate, detailed, and verifiable information. That shift sounds subtle but it changes everything—the questions asked, the evidence strategy, the way resistance is handled, the way silence is used, and the way the interview is documented. A well-conducted science-based interview may produce a confession. It may produce a false exculpatory statement that can be proven false through evidence. It may produce a corroborated witness account or a detailed timeline. All of these are valuable. None of them require pressure, confrontation, or coercion to obtain.

2. It protects memory rather than contaminating it. Human memory is not a recording. It is reconstructive, malleable, and vulnerable to suggestion. Leading questions, premature evidence disclosure, and accusatory framing can all alter what a person remembers and reports. Science-based methods—particularly Cognitive Interviewing — are specifically designed to maximize memory compatibility and minimize contamination.

3. It uses evidence strategically. Rather than revealing what the investigator knows at the outset or trying to overwhelm a suspect with evidence into confessing, science-based interviewing uses the Strategic Use of Evidence to preserve the value of evidence as a testing tool. An uncontaminated account checked against your independently held evidence is far more valuable — legally and investigatively — than a statement shaped by premature disclosure. You get an uncontaminted, corroborated statement.

4. It builds rapport as an operational tool. Rapport is not about being nice. It is about creating the conditions under which people are willing to talk openly, share details, and correct the record. Research consistently shows that rapport improves information yield, reduces resistance, and increases the accuracy of accounts. It is not a soft skill — it is a core investigative competency. It is also not presented properly by confession-driven interviewing; they use rapport to manipulate.

What Science-Based Interviewing Is Not

Because the terms "science-based," "evidence-based," and "research-based" are increasingly used as a marketing label, it is worth being specific about what they do not include.

Science-based interviewing is not anxiety-based lie detection. Any training program that teaches investigators to read deception from body language, microexpressions, eye movement, or behavioral clusters is not science-based, regardless of what it calls itself. The research does not support these methods. They introduce false case information, misclassification errors, and confirmation bias into the investigative process.

Science-based interviewing is not a step-by-step script. Real interviews and interrogations do not follow prescribed sequences. An interviewer locked into a rigid phase structure will miss what the conversation is actually offering—unexpected disclosures, contradictions, leads to other witnesses, and alternative explanations—because they are too focused on getting back on script, because it is all about one goal, a confession. Everything else, including the truth, be darned. Science-based interviewing builds adaptability through planning, not compliance through choreography.

Science-based interviewing is not confession-driven. The moment confession becomes the primary metric of a successful interview, we have stopped thinking. The process has tilted toward pressure. False confessions are real, documented, and often traceable to exactly the kind of high-pressure, guilt-presumptive tactics that confession-driven training teaches. And they rarely appear out of nowhere. What precedes a false confession is a chain: false case information that frames the wrong person, biases that harden into certainty, and — in case after case — the poor interviewing methods that turn that certainty into psychological pressure. A successful interview is a science-based interview that produces no confession but yields a thoroughly documented, internally consistent, evidence-corroborated account. Those of you who have forged your skills through on-the-stand testimony know that a confession isn't the only way to win in court. Use methods that account for this.

Science-based interviewing is not the same as being passive or permissive. Rapport does not mean avoiding hard questions. Information-gathering does not mean accepting every account at face value. Science-based interviewers ask difficult questions—they just ask them in a sequence and framing that maximizes the value of the answers rather than contaminating them. It treats statement evidence like it should be treated. Like evidence.

Why It Matters Beyond Law Enforcement

Science-based interviewing was developed primarily in law enforcement and intelligence contexts; after all, this is accessible data, but its applications extend well beyond them.

In HR, compliance, and workplace investigations, the same principles apply. Investigators who arrive with a presumption of guilt ask confirmatory questions, misclassify cooperative witnesses as deceptive, and build reports on contaminated statements. Those investigations fail under scrutiny—at depositions, termination hearings, EEOC proceedings, and in court. Science-based methods produce defensible investigations because the process is transparent, the logic is documented, and the credibility assessments are grounded in the statement and the evidence rather than behavioral folklore.

In leadership and high-stakes communication, the principles of science-based interviewing—active listening as a cognitive discipline, open-ended questions that invite narrative, rapport as an operational tool, and red team thinking—apply directly to difficult conversations, performance discussions, conflict resolution, and workplace violence prevention.

In corporate security and loss prevention, the shift from accusatory tactics to information-gathering methods reduces legal exposure, protects statements from contamination, and produces investigation records that hold up when challenged.

The underlying principles of better communication travel across the public and private sectors because they are grounded in how human communication, memory, and psychology actually work—not in how a particular industry tradition assumed they worked.

The Interviewing Research Keeps Building

One of the distinguishing features of science-based interviewing is that it is not static. The research continues. New studies test existing methods in new contexts. Field validation work examines what happens when these techniques are applied in real investigative settings, not just laboratory simulations.

A practitioner-developed Science-Based Interviewing and Interrogation course was assessed with investigators from three law enforcement agencies. After training, investigators increased their use of science-based tactics, and those tactics were associated with greater cooperation, more information disclosure, and more admissions (Russano et al., 2026). Information-gathering does not mean ineffective. It means disciplined, evidence-led, and accountable.

The contrast with legacy methods is stark. Accusatory interrogation was developed by practitioners, refined through tradition, and sold on the basis of claimed field experience. Science-based interviewing was developed by researchers, refined through empirical testing, and validated in both laboratory and field settings. One of these foundations holds up under scrutiny. The other does not.

What This Means for Interview Training

Not all interview and interrogation training is equal, even if the name implies that it is. The labels matter less than the curriculum.

Be especially wary of hybrids. The most convincing bad interview training doesn't reject science outright—it wraps pseudoscience in with it. A course may teach genuine cognitive interviewing, real rapport-building, and sound planning, then fold in behavioral lie detection or "credibility" assessments as though they carry the same evidentiary weight. They don't. The credible material is what makes junk science harder to spot and wear it's pulling it's credibility from. So the question isn't just what a curriculum teaches—it's also what it teaches alongside other material and whether it can tell the difference.

When evaluating training—for your agency, your organization, or yourself—the questions to ask are:

• Is the curriculum grounded in peer-reviewed research? Yes ✅

• Does it treat information-gathering as the primary goal? Yes ✅

• Does it build rapport to increase trust, cooperation, and information yield—because a person who trusts the process talks more openly and accurately? Yes ✅

• Does it teach investigators to dump evidence early — or fabricate evidence that doesn't exist — to overcome a subject's objections and push them toward confessing? Yes 🚩

• Does it have behavioral lie detection, taught by truth-or-lie wizards, neurolinguistic programming (NLP), microexpressions, and pseudoscientific credibility assessments within the training? Yes 🚩

• Does it promise to help you judge whether a victim or witness is credible based on how they present — instead of testing the reliability of what they actually say against the evidence? Yes 🚩

• Does it teach how human memory actually works — that memory is reconstructive and fragile, not a recording to be extracted? Yes ✅

• Does it rely on a rigid, numbered step-by-step sequence applied the same way to everyone? Yes 🚩

• Does it pair researchers with practitioners on the instruction team, so the science and its real-world application are taught together? Yes ✅

• Does it teach minimization, maximization, themes, or the use of false evidence (lying to the subject about proof that doesn't exist)? Yes 🚩

• Does it use "rapport" as a way to identify a subject's weaknesses and vulnerabilities, so interrogation themes will stick and produce a confession? Yes 🚩

• Does it teach active listening as a cognitive discipline rather than a checklist? Yes ✅

• Does it build adaptability through interview planning rather than rigid step-by-step sequences? Yes ✅

• Does it use the Cognitive Interview, the Strategic Use of Evidence, and rapport-based techniques grounded in ORBIT research? Yes ✅

• Does it present credible, evidence-based methods like the Cognitive Interview or rapport-building right alongside behavioral lie detection, microexpressions, or "credibility" assessments—as if they all belong in the same toolkit? Yes 🚩

• Does it recommend recording interviews in full, as a transparency and quality-control measure? Yes ✅

• Does it prepare investigators to explain and challenge their assumptions through red teaming? Yes ✅

• Does it use "ethical persuasion" in pursuit of a confession? Yes 🚩

• When a subject shows distress — even crying — does it instruct you to "move in" and apply more pressure? Yes 🚩

• Does it use rapid-fire or "shock value" questioning to keep a subject off guard rather than to gather accurate information? Yes 🚩

• Does it stay silent on false confessions—dismiss them as a non-issue or blame the detective for "not running the method right" when things go horribly wrong? Yes 🚩

• Was the curriculum built through academic partnerships — researchers and practitioners collaborating — rather than passed down by reputation and anecdote? Yes ✅

If the answer to most of those questions is yes, the training is moving in the right direction. If the training still includes behavioral lie detection, phase-by-phase confession scripts, behavior-provoking questions, or techniques designed to overcome denials and apply psychological pressure to obtain a confession, it is the old playbook with a new name. And there is one thing that is simply incompatible with science-based interviewing: walking into the interview or interrogation phase with a confession as your goal. The moment that becomes the objective, you narrow the flow of information and step outside what SBI methods are built to do. But this should settle the debate: if you want more reliable admissions and confessions, the choice is clear. Science-based interviewing is more effective than legacy methods in every interview metric (except false confessions; legacy interviewing wins there). It isn't even close.

The Bottom Line on Science-Based Interviewing

Science-based interviewing is not a trend or a rebranding. It is the result of decades of research asking a straightforward question: what actually works when investigators need accurate, reliable, defensible information from another human being?

The answer is not pressure. It is not behavioral reading. It is not a confession checklist. It is a disciplined, research-supported approach to human communication that produces better information, stronger cases, more defensible investigations, and fewer catastrophic investigative errors. That is why it matters — in the precinct, in the boardroom, in the HR office, and anywhere that the quality of a conversation determines the quality of a decision.

Want to bring science-based interviewing into your agency, organization, or team? See if we're a fit.

Science-Based Interviewing References

Brimbal, L., & Jones, A. M. (2018). Perceptions of suspect statements: A comparison of exposed lies and confessions. Psychology, Crime & Law, 24(2), 156–176. https://doi.org/10.1080/1068316X.2017.1390111

Hartwig, M., & Cory, C. (2025, June 18). A paradigm shift in science-based interviewing: Interrogating to elicit true and false exculpatory statements. Police Chief Online. https://www.policechiefmagazine.org/paradigm-shift-science-based-interviewing/

Russano, M. B., Meissner, C. A., Jones, M. S., Rothweiler, J. N., Taylor, P. J., Cory, C., & Brandon, S. E. (2026). Evaluating the effectiveness of a practitioner-designed science-based interviewing and interrogation course: A collaborative training and research effort. Legal and Criminological Psychology. https://doi.org/10.1111/lcrp.70021

Wilson, G. I. (2010). Perspective on neurolinguistic programming (NLP). Police Chief Magazine. https://www.policechiefmagazine.org/perspective-on-neurolinguistic-programming-nlp/