Snap Shot
- The question: In what different ways do false confessions come about, and how does the distinction between voluntary, compliant, and internalized false confessions help, or fail to help, in analyzing a disputed confession?
- What the research shows: Case studies, self-report surveys, and experiments are generally used to describe three broad types (the classic typology): false confessions volunteered without police pressure, compliant confessions by suspects who know they are innocent but confess to escape pressure or gain a benefit, and internalized confessions in which the suspect comes to believe, at least for a time, that he may have committed the crime. Later work has proposed refinements and added categories, such as confessions to protect another person, that have not been tested.
- What it does not show: A typology describes how known false confessions arose; it is not a test. No type can be assigned with confidence without first establishing on independent grounds that the confession is false, and the conditions associated with each type can also accompany true confessions.
- The law: A confession volunteered without police coercion does not violate due process however unreliable it may be (Colorado v. Connelly, 1986); its reliability is then a question for evidence law, corroboration rules, and the fact-finder (Crane v. Kentucky, 1986). Federal, military, and Arizona voluntariness rules all focus on the conduct of the questioner.
- Issues for each side: Prosecutors can point out that a typology cannot classify a confession and that true confessions share many of the same features; defense counsel can draw on the typology to organize the evidence about why a suspect might have confessed, while tying each point to the record rather than to a label.
When a defendant says that a confession was false, the first question a fact-finder asks is why anyone would confess to a crime he did not commit. The research on false confessions answers that question in more than one way, because false confessions come about through different processes. Some are volunteered; some are given by people who know they are innocent but decide that confessing is the quickest way out of an intolerable situation; and some are given by people who have come to doubt their own memory.
The distinctions matter in court because they point to different evidence: for a volunteered confession, the confessor’s mental state and motives; for a compliant confession, the interrogation and what the suspect expected to gain; for an internalized confession, the suspect’s memory, any intoxication or amnesia, and what he was told. They also matter legally, because constitutional voluntariness doctrine reaches police conduct, not a confession that no official pressure produced.
This paper describes the classic typology and its refinements, the research behind each type, related distinctions, and the limits of any typology for evaluating a particular confession. The memory processes behind internalized confessions are developed in “Internalized False Confessions and False Memories of Committing a Crime” and in “False Memory” and “Source Monitoring” in the Memory in the Courtroom resource.
The Classic Typology
Drawing on legal history and the social-psychology literature on influence, Kassin and Wrightsman (1985) distinguished three types of false confession: voluntary, given without external pressure; coerced-compliant, given by suspects who know they are innocent but confess to escape an aversive interrogation, secure a promised benefit, or avoid a threatened harm; and coerced-internalized, given by suspects who come to believe that they committed the crime (Kassin & Kiechel, 1996). The scheme has been criticized and refined for four decades and remains the most widely used framework (Gudjonsson, 2021; Kassin et al., 2025). Current writing generally drops “coerced,” in part because the processes need not involve coercion in the legal sense (Kassin et al., 2025; Welner et al., 2025).
The typology rests on three kinds of evidence. Case studies of proven false confessions supplied the original categories. Self-report surveys of prisoners, suspects, and students asked people who said they had confessed falsely why they did so (Gudjonsson, 2021). Laboratory experiments produced analogues of compliance and internalization under controlled conditions (Kassin & Kiechel, 1996; Russano et al., 2005). Each line of evidence has the limits described below.
Voluntary False Confessions
Some people claim responsibility for crimes they did not commit with no prompting from police. High-profile cases have drawn many such confessions, and the reasons described include a desire for notoriety, a need for self-punishment, difficulty distinguishing fact from fantasy, and, most commonly in self-report studies, a wish to protect the actual offender (Kassin et al., 2025). In interviews with 193 incarcerated boys aged 14 to 17, 17.1% said they had falsely confessed to police, and about half of them said they did so to protect someone else (Malloy et al., 2014; Kassin et al., 2025); Icelandic prisoners also commonly gave protecting a peer as a reason (Sigurdsson & Gudjonsson, 1996, as discussed in Welner et al., 2025).
An integrative review separates these into two categories: a protective false confession, given to shield another person, which may arise during an interrogation or be volunteered; and a voluntary false confession in the narrower sense, which the authors attribute to serious psychopathology, such as delusional guilt or a pathological desire for attention, and in which the confessor typically comes forward to police (Welner et al., 2025). The distinction matters in practice. A protective confession may be given by a person with no mental illness, may be consistent with the facts because the confessor knows them, and may be accompanied by loyalty or fear that discourages recantation.
Hypothetical Example. Consider a hypothetical shooting outside a party. The next morning a 19-year-old tells police that he, not his older brother, fired the shots, and accurately describes where the shooter stood and how many shots were fired. A neighbor’s video later shows the older brother holding the gun, and the younger brother’s phone places him across town. His details came from being told about the event. Accurate detail alone does not separate a true confession from a protective one; independent evidence does. Had the video shown the younger brother firing, the same analysis would have supported the confession.
Compliant False Confessions
A compliant false confession is an act of public acquiescence by a suspect who privately knows that he is innocent. The suspect concludes that the short-term benefits of confessing, such as ending a stressful interrogation, being allowed to sleep, make a phone call, or go home, outweigh long-term costs that seem distant or uncertain (Kassin et al., 2025). Case analyses identify incentives of exactly this kind, and the risk is described as greatest for people who are young, isolated, socially dependent, or fearful of detention (Kassin et al., 2025). In case accounts, compliant confessions are described as often retracted soon after the pressure ends (Welner et al., 2025); a retraction is one fact to examine, not a marker. Measurement of compliance is discussed in “Interrogative Suggestibility and Compliance.”
The laboratory analogue is the cheating paradigm, in which students who did or did not break an experimental rule are accused of cheating. In the original study, minimization and an explicit offer of leniency each increased both true and false confessions; with neither tactic, 46% of guilty and 6% of innocent participants signed a confession, and with both, 87% of guilty and 43% of innocent participants did so (Russano et al., 2005). The innocent participants knew they had not cheated. They signed anyway; the authors suggested that participants inferred leniency from the experimenter’s tone and, in the deal condition, believed they were better off accepting known consequences (Russano et al., 2005).
Internalized False Confessions
In an internalized false confession, an innocent suspect comes to believe that he may have committed the crime. Gudjonsson and MacKeith (1982) described the underlying condition as “memory distrust syndrome,” a state in which people come to distrust their own recollections and rely instead on external cues and suggestions (Gudjonsson, 2021). The belief is often tentative and stated in the language of inference (“I must have done it”) rather than recollection, and it may later be abandoned; for that reason Ofshe and Leo (1997) proposed calling these “persuaded” false confessions. In some cases the belief has lasted for years and has been accompanied by confabulated details (Gudjonsson, 2021; Kassin et al., 2025).
Accounts of internalized confessions identify recurring conditions: a suspect who is shocked or grieving after the death of someone close, who was intoxicated or has no memory of the period, who trusts the investigators, and who is told that strong evidence shows his guilt and offered an explanation for why he cannot remember (Kassin et al., 2025; Welner et al., 2025). One review argues that internalized confessions have not been shown to occur without both memory distrust and a manipulated alliance with the questioner (Welner et al., 2025).
Laboratory analogues are consistent with this account but limited. In the computer-crash experiment, 79 students (75 analyzed) were falsely accused of pressing a forbidden key; 69% signed a confession, 28% told another person they had pressed it, and 9% confabulated details. In the condition combining a fast pace with a confederate who falsely said she saw the key pressed, all 17 participants signed, 65% internalized, and 35% confabulated (Kassin & Kiechel, 1996). A later study classified 70% of participants interviewed repeatedly about a fabricated adolescent crime as having a false memory of committing it (Shaw & Porter, 2015); a reanalysis using stricter criteria found a substantially smaller share, with many better described as holding a false belief without recollection (Wade et al., 2018). Because the computer-crash task accuses participants of accidentally committing a highly plausible “crime,” which leaves open that some were unsure whether they were innocent (Russano et al., 2005), “internalization” there may reflect inference more than changed memory.
Refinements and Alternative Typologies
Several refinements have been proposed. Ofshe and Leo (1997) distinguished stress-compliant confessions, given to escape the stress of the interrogation itself, from coerced-compliant confessions, given in response to threats or promises, and divided persuaded confessions into coerced and non-coerced forms, alongside voluntary confessions. McCann (1998) added a coerced-reactive type for confessions produced by pressure from someone other than the police, such as a peer or family member. Gudjonsson has reported that, despite these refinements, the three-type structure has held up (Gudjonsson, 2021).
A 2025 integrative review proposed five types: coerced-compliant, internalized, leverage, protective, and voluntary (Welner et al., 2025). The leverage type describes confessions obtained when investigators truthfully tell one suspect that another has admitted involvement and named him, so that an innocent suspect admits a lesser role to avoid being portrayed as the principal actor. The authors emphasize that most of their types do not depend on police misconduct and that leverage confessions can follow professional questioning; they argue that this is one reason such confessions may be hard to recognize. Whether these categories add predictive value has not been tested.
Related Distinctions
A false confession should be distinguished from related statements. A false admission is a self-incriminating statement short of a full acknowledgment of guilt; military law draws the same line between a “confession” and an “admission” (Mil. R. Evid. 304(a)(1)). A false guilty plea is entered with counsel, usually well after arrest, and its causes may differ; self-reported false guilty pleas are at least as common as self-reported false confessions in some samples (Malloy et al., 2014; Redlich et al., 2010), as discussed in “False Guilty Pleas.” A statement that minimizes the speaker’s role or shifts blame is not a confession, though it may contain incriminating admissions (Welner et al., 2025).
A confession may also be partly true. A guilty suspect may confess to the offense but adopt details suggested by the interviewer, exaggerate or minimize his role, or confess to additional acts he did not commit. A retracted confession is not necessarily false, and a confession that was not retracted is not necessarily true. These distinctions often matter more in a real case than the choice among the three classic types.
Why the Typology Matters in a Case
The typology is useful for organizing questions. For a claimed compliant confession, the relevant evidence concerns the length and conditions of the interrogation, what the suspect was told about the evidence and consequences, what he expected to gain, and how soon he retracted. For a claimed internalized confession, it concerns his memory for the period, any intoxication or amnesia, what he was told about the evidence, and whether he spoke in terms of recollection or inference. For a volunteered confession, it concerns his mental state, his knowledge of the facts, and anyone he may have been protecting.
The typology is not a method for deciding that a confession is false. The types were derived from cases already known to be false, and the conditions associated with each type, such as pressure, fatigue, intoxication, or inference-based language, also occur in true confessions; assigning a type from the same evidence it is said to explain adds nothing. The questions that do the work are whether independent evidence corroborates the confession, whether it contains accurate details the suspect supplied before they were disclosed, and whether it led to evidence investigators did not already have (see “Corroboration and Assessing the Reliability of a Confession”).
Limits of the Evidence
The case material on which the typology rests consists of proven or probable false confessions, a selected and mostly serious set of cases. Self-report studies depend on unverified accounts in samples of prisoners, detained youth, and people with mental illness, and respondents’ explanations are themselves reconstructions (Gudjonsson, 2021). Laboratory analogues use students accused of minor acts, the dominant paradigms have been criticized as measures of low-stakes compliance (Mourtgos & Adams, 2026b), and a widely cited false-memory figure was substantially reduced on reanalysis (Wade et al., 2018). The newer five-type scheme has not been tested for reliability or predictive value (Welner et al., 2025).
Few studies report sex differences by type. In the computer-crash experiment, whose 79 participants were 40 men and 39 women, men and women confessed and internalized at comparable rates (Kassin & Kiechel, 1996); in a variant of that paradigm, women were more likely than men to confess falsely when the alleged error was highly plausible (Klaver et al., 2008). The youth self-report study sampled only boys (Malloy et al., 2014). Most proven false confessors in the case collections are men. These are group-level observations and say nothing about any individual.
No typology, test, or expert can determine whether a particular confession is true or false. The types describe processes that have produced false confessions; whether such a process occurred, and whether it produced a false statement, depends on the record and the corroborating evidence.
Questions to Ask About This Research
When a confession is described as “compliant” or “internalized,” ask how the label was assigned: from independent evidence about the interrogation and the suspect, or by working backward from the assumption that the confession is false. Ask whether the features said to mark the type, such as fatigue, inference-based phrasing, or quick retraction, also occur in true confessions. When the computer-crash or false-memory studies are cited, ask what participants were accused of, how internalization was measured, and whether the reanalysis (Wade et al., 2018) changes the figure. When a protective or leverage confession is suggested, ask what evidence identifies the person protected or the statement used as leverage. And when the absence of any type’s features is offered to show that a confession is true, ask whether the research supports that inference, since the typology describes risks, not markers of truth. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.
Types of False Confessions and the Law
Federal and Military Standards
Constitutional voluntariness doctrine is directed at official conduct. A volunteered confession can also be driven by mental illness rather than by police pressure. In Colorado v. Connelly (1986), the Supreme Court held that such a statement is not involuntary under the Due Process Clause without coercive police activity, leaving its reliability to state evidentiary law and the factfinder; the case is described more fully in “Voluntariness and Suppression of Confessions.”
A voluntary false confession of this kind is therefore unlikely to be suppressed on due-process grounds, whatever its reliability. By contrast, in Leyra v. Denno (1954), a suspect in his parents’ killing was questioned by a state-employed psychiatrist introduced as a physician who would relieve his sinus pain, and the Court held that the confessions that followed the same day were part of one coercive process and involuntary; one review discusses the case in connection with internalized confessions (Welner et al., 2025). Once a confession is admitted, the defendant may present evidence about the circumstances in which it was obtained, including evidence bearing on its reliability (Crane v. Kentucky, 1986), and federal courts require independent evidence tending to establish its trustworthiness (Opper v. United States, 1954).
In courts-martial, Article 31(d) and Military Rule of Evidence 304(a) exclude statements obtained through coercion, unlawful influence, or unlawful inducement, and Rule 304(a)(1) distinguishes confessions from admissions. Rule 304(c) requires independent evidence tending to establish trustworthiness before a confession or admission may be considered on guilt, and Rule 304(g) allows the defense to present evidence on voluntariness to the members, who give the statement the weight it deserves.
Arizona
Arizona’s standard instruction defines an involuntary statement as one resulting from the defendant’s will being overcome by a law enforcement officer’s violence, coercion, threats, or “direct or implied promise, however slight” (RAJI (Criminal) Standard 19, 2022; A.R.S. § 13-3988), so it too is directed at official conduct; a protective or volunteered false confession would be addressed instead through corroboration and the weight the jury gives the statement. Arizona’s corpus delicti rule requires evidence from which jurors could reasonably infer that the charged crime occurred before a confession may be used (State v. Carlson, 2015). In Carlson (2015), the Arizona Supreme Court upheld the exclusion of an expert’s testimony relaying the defendant’s own claim that his confession to a television reporter was false, and of risk-factor testimony by an expert who had not examined the defendant and whose testimony went to a general propensity to lie rather than to mental or physical circumstances affecting the statement; the court did not hold that expert testimony on confession research is inadmissible; admissibility of research-based testimony generally is governed by Rule 702 and State v. Salazar-Mercado (2014) and discussed in “Expert Testimony on Interrogations and False Confessions.”
Special Issues in Legal Practice
The research described above is the same whichever party relies on it. The science operates the same way regardless of who is asking the questions, and the research summarized here does not by itself favor either side. What differs between the parties is not the science but the vantage point — the particular places where these issues tend to surface, and the particular errors each party is best positioned to guard against. The two summaries below identify those recurring issues. They are offered for orientation, not as litigation strategy; in every case the controlling question is what the evidence in that case will, and will not, support, and the expert’s role is to inform the trier of fact rather than to advocate for either side.
Special Issues for Prosecutors
The research supports several points prosecutors can legitimately make. The typology was derived from cases already known to be false and cannot classify a confession whose truth is in dispute; the conditions associated with each type also accompany true confessions; and a label applied after the fact is not evidence (Welner et al., 2025). Several types do not depend on police misconduct, and a confession volunteered without police pressure raises no due-process issue (Connelly). Internalized confessions have been described mainly where memory distrust was present, such as amnesia or intoxication for the period, so a detailed account given from recollection, including verified details the suspect supplied first, is the kind of account that the corroboration analysis treats as more informative about reliability (Kassin et al., 2025); whether a particular statement is reliable is for the fact finder. A widely cited laboratory false-memory figure was substantially reduced on reanalysis (Wade et al., 2018).
The research also counsels caution. Compliant false confessions are given by people who know they are innocent and may be detailed if details were disclosed during questioning (Kassin et al., 2025); protective and leverage confessions can be consistent with the facts and can follow professional questioning (Welner et al., 2025). Where the suspect was told he had failed a test or that evidence proved his guilt and responded in inferential language, or where a codefendant’s statement was used to obtain the confession, the confession warrants the same corroboration analysis the law requires in any case, and in Arizona an implied promise of leniency, however slight, bears on voluntariness under Standard 19.
Special Issues for Defense
The research supports using the typology to organize the evidence: for a compliant confession, the conditions of the interrogation and what the suspect expected to gain; for an internalized confession, memory distrust, intoxication or amnesia, false or exaggerated evidence, and inferential language; for a protective or leverage confession, the relationship to the actual offender or the use of a codefendant’s statement (Kassin et al., 2025; Welner et al., 2025). Laboratory studies show that tactics implying leniency increase false as well as true confessions (Russano et al., 2005) and that false evidence can lead some innocent people to accept blame (Kassin & Kiechel, 1996). Even when a confession survives suppression, the circumstances of its making are admissible on weight (Crane; Mil. R. Evid. 304(g); § 13-3988).
The research also counsels caution. A type is a description, not a diagnosis, and labeling a confession “internalized” or “compliant” without independent evidence that it is false assumes the conclusion the label is offered to support. Laboratory internalization rates come from a low-stakes paradigm that has been criticized as a measure of compliance (Mourtgos & Adams, 2026b), and the 70% false-memory figure should be presented only with the reanalysis (Wade et al., 2018). A voluntary or protective confession is unlikely to support a due-process suppression claim (Connelly), and an expert who would only relay the defendant’s own account of a false confession may be excluded (Carlson). The research supports tying each feature of the claimed process to the record and to the corroboration analysis, not to the label.
This paper is for educational purposes only and is not legal advice. The author is a psychologist, not an attorney; summaries of law are general and must be verified against current authority or with a licensed attorney.
References
Gudjonsson, G. H. (2021). The science-based pathways to understanding false confessions and wrongful convictions. Frontiers in Psychology, 12, Article 633936. https://doi.org/10.3389/fpsyg.2021.633936
Gudjonsson, G. H., & MacKeith, J. A. C. (1982). False confessions: Psychological effects of interrogation. In A. Trankell (Ed.), Reconstructing the past: The role of psychologists in criminal trials (pp. 253–269). Kluwer.
Kassin, S. M., Cleary, H. M. D., Gudjonsson, G. H., Leo, R. A., Meissner, C. A., Redlich, A. D., & Scherr, K. C. (2025). Police-induced confessions, 2.0: Risk factors and recommendations. Law and Human Behavior, 49(1), 7–53. https://doi.org/10.1037/lhb0000593
Kassin, S. M., & Kiechel, K. L. (1996). The social psychology of false confessions: Compliance, internalization, and confabulation. Psychological Science, 7(3), 125–128. https://doi.org/10.1111/j.1467-9280.1996.tb00344.x
Kassin, S. M., & Wrightsman, L. S. (1985). Confession evidence. In S. M. Kassin & L. S. Wrightsman (Eds.), The psychology of evidence and trial procedure (pp. 67–94). Sage.
Klaver, J. R., Lee, Z., & Rose, V. G. (2008). Effects of personality, interrogation techniques and plausibility in an experimental false confession paradigm. Legal and Criminological Psychology, 13(1), 71–88. https://doi.org/10.1348/135532507X193051
Malloy, L. C., Shulman, E. P., & Cauffman, E. (2014). Interrogations, confessions, and guilty pleas among serious adolescent offenders. Law and Human Behavior, 38(2), 181–193. https://doi.org/10.1037/lhb0000065
McCann, J. T. (1998). Broadening the typology of false confessions. American Psychologist, 53(3), 319–320. https://doi.org/10.1037/h0092165
Mourtgos, S. M., & Adams, I. T. (2026b). What do laboratory false confession paradigms measure? A calibration meta-analysis. Journal of Quantitative Criminology. Advance online publication. https://doi.org/10.1007/s10940-026-09687-1
Ofshe, R. J., & Leo, R. A. (1997). The decision to confess falsely: Rational choice and irrational action. Denver University Law Review, 74(4), 979–1122.
Redlich, A. D., Summers, A., & Hoover, S. (2010). Self-reported false confessions and false guilty pleas among offenders with mental illness. Law and Human Behavior, 34(1), 79–90. https://doi.org/10.1007/s10979-009-9194-8
Russano, M. B., Meissner, C. A., Narchet, F. M., & Kassin, S. M. (2005). Investigating true and false confessions within a novel experimental paradigm. Psychological Science, 16(6), 481–486. https://doi.org/10.1111/j.0956-7976.2005.01560.x
Shaw, J., & Porter, S. (2015). Constructing rich false memories of committing crime. Psychological Science, 26(3), 291–301. https://doi.org/10.1177/0956797614562862
Sigurdsson, J. F., & Gudjonsson, G. H. (1996). The psychological characteristics of “false confessors”: A study among Icelandic prison inmates and juvenile offenders. Personality and Individual Differences, 20(3), 321–329. https://doi.org/10.1016/0191-8869(95)00184-0
Wade, K. A., Garry, M., & Pezdek, K. (2018). Deconstructing rich false memories of committing crime: Commentary on Shaw and Porter (2015). Psychological Science, 29(3), 471–476. https://doi.org/10.1177/0956797617703667
Welner, M., DeLisi, M., & Janusewski, T. (2025). False confessions: An integrative review of the phenomenon. Behavioral Sciences & the Law, 43(2), 185–202. https://doi.org/10.1002/bsl.2707
Legal Authorities
Colorado v. Connelly, 479 U.S. 157 (1986).
Crane v. Kentucky, 476 U.S. 683 (1986).
Leyra v. Denno, 347 U.S. 556 (1954).
Opper v. United States, 348 U.S. 84 (1954).
State v. Carlson, 237 Ariz. 381, 351 P.3d 1079 (2015).
State v. Salazar-Mercado, 234 Ariz. 590, 325 P.3d 996 (2014).
U.S. Const. amend. XIV.
Uniform Code of Military Justice art. 31, 10 U.S.C. § 831.
Ariz. Rev. Stat. Ann. § 13-3988.
Ariz. R. Evid. 702.
Manual for Courts-Martial, United States, Mil. R. Evid. 304 (2024 ed.).
Revised Arizona Jury Instructions (Criminal), Standard Criminal 19 (State Bar of Arizona 2022).
Related topic papers: Introduction to False Confessions and Police Interrogation · Internalized False Confessions and False Memories of Committing a Crime · Interrogative Suggestibility and Compliance · Corroboration and Assessing the Reliability of a Confession · False Guilty Pleas · Secondary Confessions: Informants, Co-Accused, and Jailhouse Statements
Related topics from Arizona Forensic Services: Memory in the Courtroom