Skip to content
Search resources Open menu

False Confessions

Miranda Warnings, Waiver, and Comprehension

Published

Snap Shot

  • The question: How well do suspects understand the Miranda warnings, why do most waive their rights, and what do Miranda and Article 31 protect against?
  • What the research shows: Most suspects waive their rights. Warnings vary widely in wording and reading level, and many adults, and more adolescents and people with intellectual or cognitive limitations, show incomplete understanding or hold misconceptions. In one mock-theft experiment with students, innocent participants waived more often than guilty ones (Kassin & Norwick, 2004), a finding not tested directly in field samples, and prior justice-system contact may reduce the effect; in laboratory studies, stress, intoxication, and framing the rights as a formality were associated with lower comprehension or higher waiver, in low-stakes settings.
  • What it does not show: The research does not show that a particular suspect failed to understand, that a waiver was invalid, or that waivers lead to false confessions in general. Much of it uses students or laboratory tasks, and a test given weeks later does not measure understanding at the time of the warning. Some suspects, including some with prior justice-system experience, understand and assert their rights.
  • The law: Custodial interrogation requires warnings and a knowing, intelligent, and voluntary waiver (Miranda; Dickerson); waiver may be implied, invocation must be unambiguous (Berghuis), and a Miranda violation does not support a civil damages claim (Vega). In courts-martial, Article 31(b) applies without custody and Mil. R. Evid. 305(e) requires affirmative acknowledgment and waiver. Arizona applies Miranda, and A.R.S. § 13-3988 makes advice of rights a voluntariness factor.
  • Issues for each side: For prosecutors, the research bears on recorded, plain-language warnings and an explicit, recorded waiver; for defense counsel, on the exact warning given, how it was delivered, and the person’s capacities and condition at that time.

The warnings required by Miranda v. Arizona (1966), a case that arose in Phoenix, are among the most familiar features of American criminal procedure. Before custodial interrogation, a suspect must be told of the right to remain silent, that anything said can be used in court, of the right to an attorney, and that an attorney will be appointed if the suspect cannot afford one; statements are admissible only if the suspect knowingly, intelligently, and voluntarily waived those rights. In courts-martial, Article 31(b) of the Uniform Code of Military Justice requires its own warnings, which predate Miranda.

Questions about warnings reach court in motions to suppress, in disputes over whether a suspect was in custody or invoked a right, and in claims that a waiver was not knowing and intelligent because of the suspect’s age, cognitive abilities, language, mental state, or intoxication. They also reach the jury, which may hear how rights were given as part of the circumstances bearing on a statement’s weight.

This paper summarizes the research on warning content, comprehension, and waiver decisions, and the federal, military, and Arizona law governing warnings and waivers. Forensic assessment of Miranda abilities is described in “Forensic Evaluation of a Disputed Confession or Miranda Waiver,” juveniles in “Juveniles and False Confessions,” the related questions of competency and accommodation in Competency, Miranda Waiver, and Accommodations in the Special Populations and Sex Offending resource, and issues specific to autistic suspects in Police Interviews, Miranda, Compliance, and Suggestibility in Autism in the Autism Online Offense resource.

How Often Suspects Waive

Research since shortly after Miranda has found that most suspects waive their rights and answer questions. Observational studies in the United States have found that roughly four out of five suspects, or more, waive, and in a survey, 631 North American investigators estimated that 81% of people waive (Kassin et al., 2025). Those findings are sometimes read as showing that Miranda is ineffective, and sometimes as showing that it has not prevented investigators from obtaining statements; both readings go beyond what a waiver rate alone can show. A high waiver rate is consistent with suspects who understand their rights and choose to talk, with suspects who do not understand them, and with a mixture of both.

The 2025 scientific review concluded that Miranda does not adequately safeguard the accused against unreliable confessions, citing the research on comprehension and waiver summarized below and the Supreme Court’s later decisions narrowing Miranda’s requirements (Kassin et al., 2025). Courts have taken a different view of Miranda’s role. The Supreme Court has declined to overrule Miranda, describing it as a constitutional rule embedded in routine police practice and observing that cases in which a defendant can make a colorable argument that a statement was compelled despite adherence to Miranda are rare (Dickerson v. United States, 2000). The two positions address different questions: the review asks whether warnings protect vulnerable or innocent suspects in practice, and the Court asks what the Constitution requires and whether warnings are a workable rule.

What the Warnings Say

The Supreme Court has not prescribed exact wording, and the warnings in use vary considerably (Florida v. Powell, 2010). In an analysis of 560 warnings from across the United States, reading levels ranged from about grade 2.8 to postgraduate level, the “continuing rights” component required reading comprehension about six grades higher on average than the right to silence, 45.9% said only that an attorney was available “during questioning,” and 31.8% said explicitly that appointed counsel was free (Rogers et al., 2007). A follow-up across more than 600 jurisdictions identified more than 900 uniquely worded warnings ranging from 21 to 231 words (Rogers et al., 2008, as summarized in Kassin et al., 2025). Verbal-only administration has been associated with lower comprehension than written plus verbal administration, and some Spanish-language versions omit components such as continuing rights (Kassin et al., 2025).

These findings concern the warnings, not the people who heard them. A complex warning may still be understood by a particular suspect, and a simple one may not be. They support examining the exact words used, whether they were read or provided in writing, and in what language.

Understanding and Misconceptions

Comprehension research distinguishes understanding the words of the warning, appreciating how the rights apply to one’s own situation, and reasoning about whether to waive. Studies using standardized instruments have found that many adults do not fully understand and appreciate their rights, even in benign testing conditions (Kassin et al., 2025). Misconceptions have been reported in survey research: in one line of studies, about one-third of respondents believed that once the right to silence is given up it cannot be reasserted, 46% believed that statements made “off the record” could not be used, and 62% believed that saying “I might want a lawyer” invokes the right to counsel (Rogers et al., 2010, as cited in Kassin et al., 2025; the respondent samples are not described in the source consulted). These are survey misconceptions, not rates for all suspects. Under current law, a person can reassert the right to silence, off-the-record statements can be used, and an ambiguous reference to a lawyer does not require questioning to stop (Davis v. United States, 1994).

Comprehension is lower, on average, in some groups. People with intellectual disability and people with mental illness show lower average comprehension and appreciation (Kassin et al., 2025). Among 820 pretrial defendants assessed with the Standardized Assessment of Miranda Abilities, those with limited cognitive capacities in the IQ range of about 70 to 85, who do not meet criteria for intellectual disability, also showed impaired recall of the warning and Miranda-related vocabulary, and their waiver decisions often reflected misconceptions such as viewing investigators as being on their side (Tazi & Rogers, 2023). Adolescents, particularly those 15 and younger, show lower comprehension than adults, while older adolescents in one large study performed similarly to adults (Grisso et al., 2003, as discussed in Welner et al., 2025). Lower average scores in a group do not mean that a member of that group failed to understand; many people with these characteristics understand their rights.

Transient conditions matter too. In an experiment with 136 adults, participants who consumed alcohol (mean breath alcohol concentration .07%) scored lower on understanding the warnings and Miranda vocabulary than sober participants, but intoxication did not significantly affect recognition of the warnings or appreciation of the rights (Mindthoff et al., 2022). In experiments, being accused before hearing the warning reduced comprehension compared with not being accused (Kassin et al., 2025). These studies used moderate doses, laboratory accusations, and community or student samples.

Why People Waive

Several lines of research address why suspects waive. In a mock-theft experiment with 72 undergraduate suspects (32 men and 40 women), those who were innocent were significantly more likely to sign a waiver than those who were guilty (81% vs. 36%). A male detective acted in a neutral, sympathetic, or hostile manner, and his demeanor did not affect waiver rates; 67% of innocent suspects waived even with the hostile detective, although the difference between innocent and guilty suspects was not significant in that condition alone. Of the innocent suspects who waived, 72% said they did so because they were innocent and had nothing to hide or fear (Kassin & Norwick, 2004). Kassin (2005) proposed that innocent suspects hold a “phenomenology of innocence,” a belief that their innocence will be apparent, and some later studies have reported similar effects (as summarized in Kassin et al., 2025). This is a research-supported hypothesis rather than an established fact about real suspects: the experiments involved low stakes, and the participants knew they were in a study.

The way rights are presented also matters. In an experiment in which 89 participants were accused of misconduct on a study task, participants told that a waiver had trivial rather than important implications were more likely to sign it and understood it less well, and comprehension was also lower when the misconduct was described as serious rather than minor (Scherr & Madon, 2013). The waiver in that study concerned a student advocate at a meeting with a professor, not counsel in a police interrogation. Observational research has described investigators presenting themselves as allies, framing rights as a formality, presenting the opportunity to talk as time-limited, and constructing implicit waivers (Leo, 1996, as described in Kassin et al., 2025). These practices are lawful in many circumstances; the research concerns their effect on waiver, not their legality.

Field data add a qualification from the other direction. A 2025 integrative review that emphasized the limits of laboratory paradigms noted that, in a field study of people who had been interrogated, waiving Miranda was associated with confessing, and that prior contact with the justice system, including among adolescents, may reduce vulnerability to false confession (Welner et al., 2025). Field associations of this kind do not distinguish true from false confessions.

Limits of the Evidence

Most comprehension research uses standardized instruments administered in calm conditions, often weeks or months after the warning, and cannot reconstruct what a person understood at the moment of waiver; a later evaluation can show current abilities and relevant history, not the past state of mind directly. The waiver experiments use students or community members in low-stakes mock crimes. Field waiver rates do not reveal why suspects waived or whether they understood. The warning-content studies describe forms, not their delivery. Several findings in this paper are reported as summarized in the 2025 review, which reached a policy conclusion about Miranda that courts have not adopted. Few studies analyzed sex or gender differences in comprehension or waiver; none of the studies relied on here reported such differences. Research identifies groups and conditions associated with lower comprehension; it cannot determine whether a particular waiver was knowing and intelligent.

Questions to Ask About This Research

When the warning-content studies (Rogers et al., 2007) are cited, ask what warning was actually given in this case, how it was delivered, and in what language. When group comprehension findings (Tazi & Rogers, 2023; Kassin et al., 2025) are cited, ask whether this person was evaluated, with what instrument, when, and whether the results address understanding at the time of the warning. When the innocence-and-waiver experiment (Kassin & Norwick, 2004) is cited, ask who the participants were and what was at stake. When the 2025 review’s conclusion that Miranda is inadequate is cited, ask whether it is offered as a research finding or as a policy view, and how it relates to the legal standard. When field data on waiver and confession (Welner et al., 2025) are cited, ask whether they distinguish true from false confessions. These questions are explained in “Reading Research Claims: A Guide for Attorneys” on this site.

Miranda, Article 31, and the Law

Federal and Military Standards

Miranda requires warnings before custodial interrogation and a knowing, intelligent, and voluntary waiver, and Dickerson v. United States (2000) held that Congress could not displace that rule by statute (18 U.S.C. § 3501). A waiver is knowing and intelligent if made with full awareness of the nature of the right and the consequences of abandoning it, and voluntary if it is the product of free choice rather than intimidation, coercion, or deception; police need not tell a suspect that an attorney is trying to reach him (Moran v. Burbine, 1986). Coercive police activity is required to find a waiver involuntary, and the government must prove waiver by a preponderance of the evidence (Colorado v. Connelly, 1986). For juveniles, courts apply the totality of the circumstances, including age, experience, education, background, and intelligence (Fare v. Michael C., 1979), and a child’s age informs the custody analysis (J.D.B. v. North Carolina, 2011).

Later decisions define how rights are waived and invoked. A waiver may be implied when a suspect who understood the warnings makes an uncoerced statement, and a suspect must invoke the right to silence unambiguously (Berghuis v. Thompkins, 2010); a request for counsel must also be unambiguous (Davis v. United States, 1994). Once counsel is requested, questioning must stop until counsel is present or the suspect initiates further communication (Edwards v. Arizona, 1981). In Missouri v. Seibert (2004), a four-Justice plurality held that midstream warnings after deliberate question-first interrogation may be ineffective and the postwarning statements inadmissible; Justice Kennedy, concurring in the judgment, would exclude them only where the two-step strategy was deliberate and no curative measures were taken, and his narrower test is generally treated as controlling. Statements obtained in violation of Miranda but otherwise voluntary may be used to impeach a testifying defendant (Harris v. New York, 1971), and a Miranda violation does not by itself support a civil claim for damages (Vega v. Tekoh, 2022). Under 18 U.S.C. § 3501(b), whether the defendant was advised of rights is among the factors bearing on voluntariness, and the jury gives the statement such weight as it deserves.

Case Example. In Berghuis v. Thompkins (2010), a shooting suspect in Michigan was given a written form of the warnings and, at a detective’s request, read one warning aloud so the detective could confirm that he could read English. He declined to sign the form and was largely silent for about two hours and forty-five minutes of a three-hour interrogation. He then answered “yes” when asked whether he prayed to God to forgive him for the shooting. In a five-to-four decision, the Supreme Court held that his silence did not invoke the right to remain silent, because invocation must be unambiguous, and that his uncoerced answer after understanding the warnings was an implied waiver. The case shows how courts treat evidence of understanding at the time of the warning, which is the question the comprehension research cannot answer for a particular person.

In courts-martial, Article 31(b) and Mil. R. Evid. 305(c)(1) require a person subject to the Code, or a knowing agent, to inform a suspect of the nature of the accusation, the right to remain silent, and that statements may be used at court-martial before any interrogation or request for a statement, whether or not the suspect is in custody. Counsel rights apply to custodial interrogation and, after preferral of charges, to interrogation about the preferred charges (Mil. R. Evid. 305(c)(2)–(3)). The Court of Military Appeals applied Miranda‘s counsel warnings to military interrogation in United States v. Tempia (1967). Military waiver rules are more specific than federal law: the accused must affirmatively acknowledge understanding the rights, affirmatively decline counsel, and affirmatively consent to making a statement, although a written waiver is not required, and if the accused does not affirmatively decline counsel, the prosecution must prove waiver of counsel by a preponderance (Mil. R. Evid. 305(e)(1)–(2)). Questioning must stop when the privilege is invoked (Mil. R. Evid. 305(c)(4)). A statement obtained in violation of Article 31 or Rule 305 is treated as involuntary under Mil. R. Evid. 304, though it may be used to impeach by contradiction (Mil. R. Evid. 304(e)). Whether a questioner acted in an official law-enforcement or disciplinary capacity is decided objectively (United States v. Jones, 2014). Military practice is addressed further in “Confessions in Courts-Martial: Article 31(b), Mil. R. Evid. 304 and 305, and Expert Assistance.”

Arizona

Miranda, Edwards, and Mincey v. Arizona (1978) all arose from Arizona prosecutions, and Arizona courts apply the federal Miranda framework alongside the state privilege against self-incrimination (Ariz. Const. art. 2, § 10). Under A.R.S. § 13-3988, the trial judge determines voluntariness outside the jury’s presence considering all the circumstances, including whether the defendant knew the nature of the offense, knew he was not required to make a statement and that any statement could be used against him, and was advised of the right to counsel; no single factor is conclusive. If the statement is admitted, the jury hears the evidence on voluntariness, and the standard instruction directs jurors not to consider the statement unless they find beyond a reasonable doubt that it was voluntary (RAJI (Criminal) Standard 19, 2022). The Arizona Supreme Court has applied the federal custody analysis, holding, for example, that a burglary suspect briefly placed in a patrol car and then questioned on a curb was not in custody for Miranda purposes, because a restraint on freedom of movement does not by itself establish custody without the inherently coercive pressures of station-house questioning (State v. Maciel, 2016, ¶¶ 12, 16, 29). Arizona confession law is treated more fully in “Arizona Confession Law.”

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 practices that make waivers more reliable and easier for a court to evaluate: plain-language warnings at a modest reading level, provided in writing and orally in the suspect’s language, a request that the suspect explain the rights in his own words, an explicit waiver rather than an implied one, and a complete recording of the advisement and waiver (Rogers et al., 2007; Kassin et al., 2025). In courts-martial, Rule 305(e) already requires affirmative acknowledgment, and a recorded colloquy meeting that standard is strong evidence. Many suspects understand and knowingly waive, the law does not require a perfect understanding, and comprehension research based on later testing does not establish what a suspect understood at the time.

The research also counsels caution. Courts weigh age, intellectual and cognitive limitations, mental illness, intoxication, and language under the totality of the circumstances, and the research indicates that these conditions are associated with lower comprehension (Mindthoff et al., 2022; Tazi & Rogers, 2023). Framing rights as a formality or moving into questioning without an explicit waiver increased waiver in experiments (Scherr & Madon, 2013), and a suspect’s statement that he understood may not show that he did. In such cases, a careful explanation and a recorded check of understanding give the court a clearer record of what the suspect understood.

Special Issues for Defense

The research supports a close look at how rights were given: the exact text, its reading level, whether it was read aloud only, the language used, whether the suspect was asked to explain the rights, and whether a waiver was explicit (Rogers et al., 2007). It also supports evaluating whether a client’s age, cognitive ability, mental state, or intoxication at the time may have limited understanding, including clients with limited cognitive capacity who do not meet criteria for intellectual disability (Tazi & Rogers, 2023; Mindthoff et al., 2022). Common misconceptions about the permanence of waiver, off-the-record statements, and how to invoke counsel are relevant when a client’s words or conduct during the interview suggest them. In courts-martial, the affirmative-waiver requirement of Rule 305(e) and the absence of a custody requirement under Article 31(b) are significant differences from federal practice.

The research also counsels caution. Group findings do not show that a particular client failed to understand, and a later evaluation measures present abilities. Most suspects waive, many with full understanding, and prior experience with the justice system may reduce vulnerability (Welner et al., 2025). Courts generally accept the waiver of a properly warned adult who indicated understanding, require coercive police activity for involuntariness (Connelly), accept implied waivers (Berghuis), and have described colorable claims that a statement was compelled despite adherence to Miranda as rare (Dickerson). A Miranda challenge supported by evidence of the specific warning and the person’s documented limitations at the time is better supported than one based on research alone.

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

Kassin, S. M. (2005). On the psychology of confessions: Does innocence put innocents at risk? American Psychologist, 60(3), 215–228. https://doi.org/10.1037/0003-066X.60.3.215

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., & Norwick, R. J. (2004). Why people waive their Miranda rights: The power of innocence. Law and Human Behavior, 28(2), 211–221. https://doi.org/10.1023/B:LAHU.0000022323.74584.f5

Mindthoff, A., Evans, J. R., Wolfs, A. C. F., Polanco, K., Goldstein, N. E. S., & Schreiber Compo, N. (2022). The detrimental impact of alcohol intoxication on facets of Miranda comprehension. Law and Human Behavior, 46(4), 264–276. https://doi.org/10.1037/lhb0000490

Rogers, R., Harrison, K. S., Shuman, D. W., Sewell, K. W., & Hazelwood, L. L. (2007). An analysis of Miranda warnings and waivers: Comprehension and coverage. Law and Human Behavior, 31(2), 177–192. https://doi.org/10.1007/s10979-006-9054-8

Scherr, K. C., & Madon, S. (2013). “Go ahead and sign”: An experimental examination of Miranda waivers and comprehension. Law and Human Behavior, 37(3), 208–218. https://doi.org/10.1037/lhb0000026

Tazi, K. Y., & Rogers, R. (2023). Falling through the cracks: Failing to identify compromised Miranda abilities for defendants with limited cognitive capacities. Behavioral Sciences & the Law, 41(5), 326–342. https://doi.org/10.1002/bsl.2610

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

Berghuis v. Thompkins, 560 U.S. 370 (2010).

Colorado v. Connelly, 479 U.S. 157 (1986).

Davis v. United States, 512 U.S. 452 (1994).

Dickerson v. United States, 530 U.S. 428 (2000).

Edwards v. Arizona, 451 U.S. 477 (1981).

Fare v. Michael C., 442 U.S. 707 (1979).

Florida v. Powell, 559 U.S. 50 (2010).

Harris v. New York, 401 U.S. 222 (1971).

J.D.B. v. North Carolina, 564 U.S. 261 (2011).

Mincey v. Arizona, 437 U.S. 385 (1978).

Miranda v. Arizona, 384 U.S. 436 (1966).

Missouri v. Seibert, 542 U.S. 600 (2004).

Moran v. Burbine, 475 U.S. 412 (1986).

Vega v. Tekoh, 597 U.S. 134 (2022).

State v. Maciel, 240 Ariz. 46, 375 P.3d 938 (2016).

United States v. Jones, 73 M.J. 357 (C.A.A.F. 2014).

United States v. Tempia, 16 U.S.C.M.A. 629, 37 C.M.R. 249 (C.M.A. 1967).

U.S. Const. amend. V.

Ariz. Const. art. 2, § 10.

Uniform Code of Military Justice art. 31, 10 U.S.C. § 831.

18 U.S.C. § 3501.

Ariz. Rev. Stat. Ann. § 13-3988.

Manual for Courts-Martial, United States, Mil. R. Evid. 304, 305 (2024 ed.).

Revised Arizona Jury Instructions (Criminal), Standard Criminal 19 (State Bar of Arizona 2022).

Related topic papers: Forensic Evaluation of a Disputed Confession or Miranda Waiver · Juveniles and False Confessions · Intellectual Disability and False Confessions · Voluntariness and Suppression of Confessions · Confessions in Courts-Martial: Article 31(b), Mil. R. Evid. 304 and 305, and Expert Assistance · Arizona Confession Law

Related topics from Arizona Forensic Services: Autism Online Offense · Courts-Martial