Interrogation Techniques: Reid, PEACE and False Confessions

John E. Reid was a Chicago-based polygraph examiner who developed a systematic interrogation method in the 1950s through his work with Fred Inbau, then a law professor at Northwestern University. Their 1962 manual Criminal Interrogation and Confessions became the dominant police training text in North America, and the technique it describes remains widely taught in the United States, Canada, and parts of Latin America.

The BAI screening stage. The Reid Technique begins before the formal interrogation with a Behavioral Analysis Interview (BAI). The BAI is a structured interview in which the investigator asks a standardised set of questions (some behaviour-provoking, some designed to elicit direct answers) while observing the subject's verbal responses, non-verbal behaviour, and what Reid called "paralanguage" cues: voice quality, response latency, and speech characteristics. Reid trained investigators to read these cues as indicators of deception. Gaze aversion, crossing arms, touching the face, and delayed responses are among the behaviours taught as deception indicators.

The empirical problem with the BAI is severe. Multiple laboratory and field studies have found that trained Reid interviewers perform no better than chance at detecting deception, and in some studies perform significantly worse than untrained observers because training installs confident but inaccurate heuristics (Bond and DePaulo 2006 meta-analysis; Vrij, Granhag, and Porter 2010 review). The same pattern of overconfident but inaccurate behavioural reading underlies the critique of offender profiling, where trained profilers similarly fail to outperform untrained comparison groups on behavioural inference tasks. The Reid organisation has disputed these findings, arguing that laboratory studies do not capture the high-stakes realism of actual interrogations, but the field research has not been more encouraging.

The nine-step interrogation. When a BAI indicates (or is thought to indicate) guilt, the suspect proceeds to the nine-step interrogation. The steps are:

1. The positive confrontation: the investigator states, as a fact rather than a hypothesis, that the investigation has established the suspect's guilt.
2. Theme development: the investigator offers morally minimising explanations for the crime (you didn't plan it; you were provoked; the victim was partly to blame) to make confession seem psychologically palatable.
3. Handling denials: the investigator cuts off the suspect's denials, interrupting and redirecting when the suspect attempts to deny guilt.
4. Overcoming objections: factual objections (I was elsewhere at the time) are handled by the investigator without genuine investigation.
5. Keeping attention: the investigator maintains eye contact and physical closeness to keep the suspect focused on the interrogation.
6. Reading the suspect's signs of surrender: the investigator watches for behavioural cues indicating psychological capitulation.
7. Presenting the alternative question: the investigator presents a choice between two versions of the offense, one morally more serious than the other ("Was this your idea, or did someone pressure you into it?"). Either choice is an admission.
8. The oral confession: the suspect admits guilt.
9. The written confession: the admission is converted to a detailed written statement.

The false-confession mechanism within Reid. Steps 2 and 7 are particularly implicated in false confessions. Step 2 (moral minimisation) reduces the perceived psychological cost of confessing. Research by Kassin and McNall (1991) demonstrated that minimisation constitutes an implied promise of leniency, a legally impermissible inducement in US law, even when no explicit promise is made. Step 7 (the alternative question) creates a false dichotomy in which both options involve admitting to the act. A suspect who initially denies guilt may accept the less-serious option under pressure without fully understanding that either answer constitutes a confession.

The impetus for developing an alternative to confession-centred interrogation methods in England and Wales came from two convergent pressures in the early 1990s. First, the Royal Commission on Criminal Justice (the Runciman Commission), convened after a series of high-profile miscarriages of justice including the Birmingham Six (convicted in 1975, acquitted in 1991) and the Guildford Four (convicted in 1975, acquitted in 1989), examined police interrogation practice and found it confession-focused and legally vulnerable. Second, psychologists including Gisli Gudjonsson, Michael Shepherd, and David Canter had been accumulating research demonstrating that coercive questioning produced unreliable information and, in vulnerable populations, false confessions.

The PEACE model was developed by a consortium including the Home Office, ACPO (Association of Chief Police Officers), and the Central Planning and Training Unit, published in 1992, and rolled out as mandatory training across England and Wales. PEACE is an acronym:

Planning and Preparation: the interviewer prepares by reviewing available evidence, formulating hypotheses (not a presumption of guilt), and designing the interview to test those hypotheses. The interview has an explicit structure.

Engage and Explain: the interviewer opens by establishing rapport, explaining the purpose of the interview, and setting ground rules including the interviewee's right to decline to answer. The tone is non-confrontational.

Account: the interviewer elicits the subject's own account using open-ended, non-leading questions. The PEACE model privileges free narrative and open prompts ("Tell me everything that happened from the beginning") over the closed, confirmatory questioning of the Reid BAI.

Closure: the interviewer summarises the account, gives the subject an opportunity to correct or add to it, and ends the session constructively.

Evaluate: after the interview, the interviewer reviews the account against available evidence to identify consistency, inconsistency, and gaps to be followed up.

What PEACE does not do. PEACE does not have a stage for "breaking down" the subject. It does not use minimisation. It does not present false evidence as established fact (a technique the Reid Technique permits, and which is explicitly illegal under UK law: R v. Mason [1988] 1 WLR 139). It treats confession as information rather than as the goal.

International adoption. Norway adopted a variant of the PEACE model (KREATIV) in the early 2000s, partly in response to research by Asbjørn Rachlew showing that Norwegian police interrogators using Reid-type techniques produced unreliable confessions. Australia (Victoria Police 2004 interview manual) and New Zealand (PEACE-based guidelines) followed. Canada has moved significantly toward information-gathering approaches, particularly following the Kaufman Commission (1998) on the wrongful conviction of Guy Paul Morin. In the United States, the High-Value Detainee Interrogation Group (HIG) has funded research into evidence-based, non-coercive interrogation since 2009, with the Strategic Use of Evidence (SUE) technique developed by Pär-Anders Granhag as one outcome.

Saul Kassin, a psychology professor at John Jay College of Criminal Justice in New York, is the leading researcher on false confessions. Together with Lawrence Wrightsman, Kassin proposed the three-category taxonomy in 1985 that remains the standard framework in the field.

Voluntary false confessions are given without external pressure, often by mentally ill individuals who confabulate a role in a well-publicised crime, or by individuals who confess to protect someone else. The first type (confabulation in the absence of pressure) is a recognised clinical phenomenon. In high-profile cases attracting intense media coverage, there are often multiple individuals who contact police claiming responsibility. Most are easily excluded by the details.

Compliant false confessions occur when the suspect knows they are innocent but confesses to escape the immediate pressure of interrogation, to gain a promised benefit, or to avoid a threatened harm. The confession is strategic, not a genuine belief in guilt. The suspect who has been interrogated for sixteen hours without sleep, has been told falsely that their co-accused has already confessed and named them, and has been told that cooperation will lead to leniency, may calculate that confessing now and retracting later is preferable to continued interrogation. This type is particularly common among juveniles, individuals with intellectual disabilities, and individuals with low frustration tolerance.

Brendan Dassey (Wisconsin, 2005) is a canonical case. Dassey, aged sixteen and with below-average measured intelligence, was interrogated multiple times by two detectives without a parent or attorney present. The detectives made repeated implicit promises ("We'll help you if you tell the truth") and fed Dassey details of the crime that appeared in his subsequent confession, which was later demonstrated to track detective-supplied information rather than independent knowledge. The US Seventh Circuit Court of Appeals ruled in 2017 that Dassey's confession was involuntary under the totality of the circumstances test, though the full court subsequently reversed that ruling on grounds related to deference to state-court findings.

Internalised false confessions occur when the suspect comes genuinely to believe that they committed the crime, typically as a result of memory contamination, extreme fatigue, or psychological vulnerability. Peter Reilly's case is an example: following sixteen hours of interrogation and false polygraph feedback, he appears to have genuinely doubted his own memory and constructed a confessional narrative. Steven Titus (Washington State, 1981) was wrongly convicted of rape he had not committed, a case driven primarily by eyewitness misidentification rather than a classic internalised false confession.

Why false confessions are believed. Kassin, Drizin, Grisso, Gudjonsson, Leo, and Redlich (2010) reviewed the literature on police-induced confessions and found that false confessions are persuasive to juries, to judges, and to reviewing courts precisely because confessions are behaviourally implausible under an innocence assumption. This persuasive weight compounds the problem documented in the eyewitness testimony literature: cases relying on both a false identification and a false confession are among the hardest wrongful convictions to reverse. The intuition that no innocent person would confess to a serious crime is deeply held and very difficult to displace with evidence. Mock-juror studies consistently show that confession evidence produces the highest conviction rates of any evidence type, and that the conviction effect persists even after judges instruct the jury that the confession was involuntary.

The legal controls on interrogation and confessions reflect the tension between effective criminal investigation and the protection of the suspect's rights. The frameworks in the United States, India, and the United Kingdom each address this tension differently, with different implications for the admissibility and reliability of confession evidence.

United States: Miranda v. Arizona (1966). The US Supreme Court's ruling in Miranda v. Arizona (384 US 436, 1966) required that suspects in custody be informed of their rights before interrogation: the right to remain silent; the warning that anything they say may be used against them; the right to have an attorney present; and the right to an appointed attorney if unable to afford one. Failure to give the Miranda warnings before a custodial interrogation renders any confession obtained inadmissible. The Court's reasoning was rooted in the inherently coercive nature of custodial interrogation, which it described as "psychologically rather than physically oriented."

The practical limitation of Miranda is that suspects routinely waive their rights. Studies by Kassin and Norwick (2004) found that innocent suspects were more likely to waive their Miranda rights than guilty ones, because innocent suspects trust that the truth will exonerate them. The waiver removes the procedural protection precisely in the category of suspects most vulnerable to false confession.

India: State of Bombay v. Kathi Kalu Oghad (1961). The constitutional provision relevant to self-incrimination in India is Article 20(3): "No person accused of any offence shall be compelled to be a witness against himself." The Supreme Court in Kathi Kalu Oghad (1961 SCR (2) 457) held that this protection applies to testimonial compulsion, meaning compelled oral or written statements. It does not extend to documents that pre-exist the investigation, or to physical samples (blood, fingerprints, hair). Critically for interrogation practice, the Court held that a voluntary confession is not covered by Article 20(3). BNSS 2023 § 23 (successor to the Indian Evidence Act and now BSA 2023) renders confessions made to police officers inadmissible, regardless of voluntariness, unless made in the presence of a magistrate. This provision, far more protective than the Miranda rule, has been interpreted by the Supreme Court as reflecting a recognition of the historically coercive character of police interrogation in India.

India: Selvi v. State of Karnataka (2010). The Supreme Court's ruling in Selvi v. State of Karnataka (2010 SC 1974) addressed the use of narcoanalysis, polygraph testing, and BEAP (Brain Electrical Activation Profile, brain-fingerprinting) in Indian criminal investigations. The Court held that compulsorily administered narcoanalysis constitutes a violation of Article 20(3), because the subject's mental contents are extracted involuntarily, constituting testimonial compulsion. The Court explicitly extended Article 20(3) to cover involuntary mental compulsion, not just physical compulsion. Voluntary narcoanalysis remains technically permissible with informed consent, but the Court required stringent procedural safeguards. The Selvi ruling is discussed in greater depth in the Forensic Neuroscience topic; it stands as the principal Indian constraint on coercive interrogation methods.

United Kingdom: PACE 1984 and the right to silence. In England and Wales, the Police and Criminal Evidence Act 1984 Part IV governs detention, and Code C regulates the interrogation of persons in detention. A suspect must be cautioned ("You do not have to say anything. But, it may harm your defence if you do not mention when questioned something which you later rely on in court") before questioning. Interviews must be audio-recorded under Code E, and in serious cases video-recorded. Suspects have the right to legal advice before and during questioning. The Criminal Justice and Public Order Act 1994 modified the right to silence by permitting adverse inference where the suspect fails to mention facts later relied on at trial, but the caution requirement remains.

The English false-confession safeguards. R v. Mason [1988] 1 WLR 139 (Court of Appeal) held that police deception of a solicitor (the defence lawyer was told, falsely, that the client's fingerprints had been found at the scene) rendered the resulting confession inadmissible. This case is the source of the principle that deliberate deception of the defence in obtaining a confession can lead to exclusion under PACE 1984 s 78 (discretion to exclude unfair evidence).

Gisli Gudjonsson, an Icelandic-British forensic psychologist at King's College London, developed the Gudjonsson Suggestibility Scales (GSS-1 and GSS-2) in the 1980s to provide a psychometric measure of interrogative suggestibility. The scales are now widely used in cases where the reliability of a confession is in question.

The scale structure. The GSS involves presenting the subject with a short narrative (a news story), asking for immediate free recall, asking twenty leading questions (some accurate, some misleading), and then applying "pressure" (telling the subject that they made several errors and the interview must be done again). The scale yields two measures: Yield (the tendency to accept misleading suggestions on the initial questioning) and Shift (the tendency to change answers following pressure). Both measures predict the likelihood that a suspect will provide information that is incorporated from the questioner's suggestions rather than from their own genuine memory.

Vulnerability populations. Research by Gudjonsson and others has established that several populations show elevated suggestibility: individuals with intellectual disabilities; individuals with high anxiety; individuals with low self-esteem; individuals in situations of social isolation (particularly juveniles separated from parents); and individuals with acute fatigue or substance intoxication or withdrawal. The intersection of these vulnerability factors with the Reid Technique's deliberate use of fatigue, isolation, and psychological pressure is the core of the false-confession problem.

Application in court. In England and Wales, Gudjonsson suggestibility evidence has been admitted in a series of cases where the defendant's confession was challenged. R v. Ward [1993] 1 WLR 619 was the most influential early case: Judith Ward had confessed to IRA bombings, but the Court of Appeal accepted that her intellectual functioning and heightened suggestibility rendered her confession unreliable. The Court allowed expert psychological evidence on her susceptibility to confabulation under interrogation pressure. In India, courts have not yet routinely accepted suggestibility expert testimony, though BSA 2023 § 39 provides the framework.

Intellectual disability and interrogation in international practice. The ADA (Americans with Disabilities Act) and the Criminal Justice Mental Health Standards of the American Bar Association both address the interrogation of individuals with intellectual disabilities. UK PACE Code C includes specific provisions for "appropriate adults" (a responsible adult who must be present during interrogation of individuals with mental disorders or intellectual disabilities). Australia's Crimes Act and the New Zealand Evidence Act include similar provisions. POCSO 2012 § 26(4) in India requires that a child witness be examined in the presence of the child's parents, guardians, or any other person the child trusts. There is no equivalent provision for intellectually disabled adult suspects under interrogation, a gap identified by Indian disability-rights advocates.

The US Senate Select Committee on Intelligence's 2014 report on the CIA's post-2001 enhanced interrogation programme documented that coercive techniques had produced false intelligence and that more productive information had consistently come from rapport-based interviews conducted before the coercive programme began. The report was significant not just as a political document but as a large-scale natural experiment: coercive methods produced compliance, but compliance produced confabulation rather than accurate information.

The High-Value Detainee Interrogation Group (HIG). Established by President Obama in 2009 and based at the FBI, the HIG funds scientific research into effective interrogation methods. The group has published research summaries and commissioned studies through a Science and Technology branch. Key outputs include documentation of the Strategic Use of Evidence (SUE) technique (withholding known evidence and presenting it strategically to identify inconsistencies in a deceptive account), the Motivational Interviewing approach (building rapport and eliciting cooperation), and cognitive load manipulations (asking suspects to report events in reverse order, which increases cognitive demand for deception relative to truth-telling).

The SUE technique. Developed by Pär-Anders Granhag and Leif Strömwall at the University of Gothenburg, the Strategic Use of Evidence approach maximises the diagnostic value of known evidence by using it strategically. The interviewer who immediately presents all known evidence gives the deceptive suspect an opportunity to incorporate that evidence into their false account. The SUE interviewer instead allows the suspect to give a free account, identifies where the suspect has committed to a position that conflicts with the known evidence, and then presents the evidence to maximise the inconsistency. Field validation studies in Sweden and Norway have found SUE interviews produce more case-relevant inconsistencies than standard techniques, which was confirmed as deceptive in case outcomes.

The trajectory of US domestic reform. The APLS (American Psychology-Law Society), the APA, and the ABA have all issued position statements calling for the replacement of confession-centred interrogation with evidence-based information-gathering approaches. As of 2025, a growing number of US law enforcement agencies have adopted modified approaches incorporating PEACE-influenced training, but the Reid Technique manual remains the most widely sold police interrogation training resource in North America.