The Psychology of Wrongful Convictions: Memory, Confessions, Bias, and Error
Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy
Wrongful convictions arise when human judgments and legal procedures converge on an erroneous conclusion about guilt. Psychological science explains several recurring routes to that error: an eyewitness confidently identifies the wrong person; an innocent suspect makes an unreliable admission; an investigator becomes committed to an early theory; a forensic examiner encounters suggestive context; or a juror interprets several apparently independent pieces of evidence that actually reinforce the same initial mistake. These mechanisms can interact throughout an investigation, prosecution, plea negotiation, trial, and later review.
The psychology of wrongful convictions studies how memory, persuasion, authority, expectation, uncertainty, evidence evaluation, and institutional decisions affect the reliability of criminal-case conclusions. Its central contribution is to identify where errors can enter a case, how they spread, why sincere people may fail to detect them, and which safeguards can reduce their likelihood. Research can assess known vulnerabilities and the quality of procedures; guilt, innocence, admissibility, and the legal validity of a conviction are determined under applicable law.
This article integrates the major research streams and explains their evidentiary limits. The foundational scientific sources include the National Research Council's eyewitness-identification consensus report, the 2025 American Psychology–Law Society scientific review of police-induced confessions, the National Academies report on forensic science, and the National Registry of Exonerations' official 2025 report.
What Counts as a Wrongful Conviction?
A wrongful conviction is a criminal conviction in which an innocent person was convicted of the offense, or in which a legal system later recognizes a conviction as wrongful under its governing rules. The phrase is used in overlapping ways. Researchers studying factual innocence may restrict their analysis to people shown not to have committed the crime. Legal discussions may also address convictions obtained through serious procedural or constitutional errors even where factual innocence remains unresolved. An overturned conviction, an acquittal after retrial, an exoneration, and a finding of actual innocence are therefore distinct outcomes.
An exoneration is a postconviction outcome in which authorities formally clear a person of a crime, according to the standards of the source that records the case. Different registries use defined inclusion rules. A conviction vacated because evidence was improperly obtained may require a new trial without resolving who committed the offense. Conversely, compelling indications of innocence may arise in cases that never obtain formal exoneration. Careful analysis names the outcome and the underlying evidence rather than treating every reversal as proof of innocence.
In psychological terms, a wrongful conviction is a failure of a decision system under uncertainty. Investigators rarely know the ground truth when they begin. They assess hypotheses from incomplete witness reports, interviews, documents, physical traces, and the conduct of other actors. A court subsequently applies procedural rules and a defined burden of proof. Each stage can change what evidence survives, which interpretations gain prominence, and how later decision makers understand the original event.
How Often Do Wrongful Convictions Happen? What Exoneration Data Can Tell Us
The true rate of wrongful conviction cannot be read directly from an exoneration registry. Verified exonerations reveal that errors occur and permit detailed investigation of pathways in documented cases. They do not provide a random sample of all convictions. Discovery depends on access to preserved evidence, legal representation, biological material, judicial review, publicity, agency practices, and the resources needed to reopen a case. Some categories of crimes and defendants have far greater prospects of later proof of innocence than others.
The National Registry of Exonerations' 2025 annual report documented 97 exonerations in that reporting year and classified recurring contributing factors such as official misconduct, perjury or false accusation, mistaken witness identification, false confessions, misleading forensic evidence, and inadequate legal defense. These factors can overlap within one case: a mistaken identification may lead to pressure for a confession, while an erroneous forensic interpretation may appear to corroborate both. Its numbers describe the cases covered by the registry, not the prevalence of each error among all American convictions.
Research estimates derived from specialized settings, such as particular categories of death-sentence cases or prisoners later tested using DNA, also require their own denominators and eligibility assumptions. It is an error to extrapolate a single percentage to all prosecutions, pleas, crimes, states, or countries. The sound conclusion is that known wrongful convictions expose serious and preventable vulnerabilities, while their total population frequency remains uncertain and context dependent.
A Psychological Model: How Small Errors Become a Conviction
A criminal case unfolds as a sequence of contingent decisions. An initial statement directs an investigation toward a suspect; investigators collect confirming leads; an interviewer approaches the suspect as likely guilty; forensic material is evaluated with knowledge of the accusation; prosecutors decide which evidence supports a charge; defense counsel evaluates the likely consequences of trial; and a jury receives the resulting account. The later stages often rely on conclusions generated earlier. If the initial interpretation was mistaken, the same underlying error can become embedded in several sources.
This accumulation can produce evidence dependence. An eyewitness's identification, a confession, and a forensic opinion may look like separate confirmations. Yet the identification may have shaped interrogation; the interview may have supplied facts later repeated in the confession; and an examiner may have encountered the suspect's alleged admission before interpreting ambiguous material. Three sources do not provide three independent confirmations when they were influenced by one another. Case review should reconstruct who learned what, when, and from whom.
The distinction between evidence generation and evidence evaluation is essential. A procedure can change a witness's response or a suspect's statement before any judge considers the result. A second procedure can determine how reliably the result is documented. A third can determine what a jury hears about uncertainty, alternatives, and contamination. Psychology addresses these processes empirically; law decides how they bear on admissibility, discovery, burdens, remedies, and proof.
Eyewitness Memory: Why Sincere Identification Can Be Wrong
Human memory is a reconstructive capacity. People detect, encode, store, and retrieve information under limits imposed by attention, lighting, distance, duration, visibility, expectations, stress, and subsequent exposure to related information. Seeing an unfamiliar person briefly during a threatening event is a different task from recognizing a familiar person under good conditions. A witness can describe an experience sincerely and still make an identification error. The 2014 National Research Council consensus report documents the underlying perceptual and procedural issues.
A lineup is a decision task in which a witness may identify a suspect, identify a filler, or reject the lineup. Its evidentiary meaning depends on whether the perpetrator was actually present, how the suspect and fillers were chosen, whether administrators knew the suspect, which instructions were delivered, and how the result was recorded. A witness who believes the suspect must be present may compare faces and choose the best match even when the actual perpetrator is absent. The lineup's design and administrator behavior can alter this decision process.
The initial identification is particularly important. Subsequent viewings, photographs, publicity, courtroom exposure, and comments from investigators may reinforce familiarity with the selected person. An identification made for the first time in court is difficult to evaluate as a clean test of original memory, because the defendant is obvious and the witness may have encountered their image repeatedly. Readers can explore the specialized evidence and limitations in Eyewitness Testimony: How Memory Works in Legal Cases.
Confidence, Accuracy, and the Conditions of the First Identification
Eyewitness confidence requires precise interpretation. Under carefully controlled conditions and at an uncontaminated first identification, confidence can provide meaningful information about accuracy for certain identification decisions. That relationship becomes harder to interpret when a witness has received feedback, selected the same person repeatedly, learned case details, or revised a memory of earlier certainty. The National Academies recommendations therefore emphasize recording confidence immediately, in the witness's own words, and documenting the identification procedure itself.
The relevant question is not simply whether a witness appears confident at trial. It is whether the witness expressed confidence before feedback and repeated exposure, whether the chosen person matched an independently documented description, and whether the procedure offered a fair opportunity to reject the array. The National Research Council report also warns against treating some variables as universal or assuming that laboratory differences among lineup formats automatically resolve every field setting.
Misinformation, Co-Witness Discussion, and Familiarity
After witnessing an event, a person may hear descriptions from investigators, other witnesses, family members, social media, or journalists. Later recollection can incorporate elements of that post-event information or become less certain about their source. A witness may remember an image seen in the investigation and misattribute its familiarity to the original scene. This mechanism is particularly important when multiple witnesses talk before giving independent accounts or when interviews repeatedly suggest one interpretation.
The broader experimental literature on post-event misinformation supports the proposition that recall can be altered under identifiable conditions, although effect sizes and generalizability vary across materials and protocols. A recent systematic review of misinformation research provides a structured evidence base. The specialist article on The Misinformation Effect in Eyewitness Memory addresses that mechanism; the wrongful-conviction question is how contaminated memory later becomes apparently independent evidence.
What Better Identification Procedures Accomplish
The National Academies recommends blind or double-blind administration, standardized instructions explicitly stating that the perpetrator might be absent, verbatim recording of initial confidence, video documentation, and careful court review of earlier identification procedures. These safeguards protect the evidentiary history and reduce opportunities for inadvertent influence. They cannot guarantee that every identification is correct, because errors can arise before an identification procedure begins. The supporting recommendations are set out in Identifying the Culprit.
Independent lineup administration reduces the risk of the administrator's knowledge affecting the witness. Documentation preserves what happened before a witness or investigator had reason to reinterpret it. The related published guidance, Double-Blind Lineups and Confidence Statements: Why Procedure Matters, addresses implementation questions. Disputes about simultaneous and sequential presentation require attention to the exact measure of accuracy, response bias, and the jurisdiction's policy; the 2014 consensus report did not declare one method universally superior.
False Confessions: How Innocent People Come to Admit Guilt
A false confession is a self-incriminating statement admitting involvement in an offense the speaker did not commit. It differs from an unproven allegation that a confession was coerced, an inaccurate detail in a true account, or a retraction that has not been independently assessed. Exoneration cases and experimental studies show that innocent individuals can admit wrongdoing under some conditions. The 2025 scientific review by Kassin and colleagues, approved by the American Psychology–Law Society, synthesizes the principal psychological risk factors and safeguards.
Suspects encounter a powerful social situation in interrogation: authority, restricted knowledge of the evidence, uncertainty about consequences, fatigue, fear, expectations of cooperation, and pressure to resolve an accusation. Some innocent suspects may believe that their innocence will eventually be obvious, underestimate the significance of an admission, or conclude that saying what the interviewer wants offers the fastest exit from an intolerable encounter. Investigators may sincerely believe the person is guilty and interpret denials as evasive, creating a feedback loop between belief and tactics.
False confessions may arise through several pathways described in research: voluntary admissions made without direct police pressure, compliant admissions in which a person knowingly says something false, and cases in which a person becomes confused or comes to doubt their own memory. These are explanatory categories, not clinical diagnoses or automatic classifications of any disputed statement. The published EN-58 article False Confessions: Why Innocent People Confess examines these pathways in depth.
Interrogation Methods and the Risk of Contamination
A statement's detail is often treated as a marker of authenticity, especially if it describes facts not publicly known. That inference is strongest when investigators can document that the speaker supplied accurate nonpublic information independently. If interviewers previously disclosed the facts, offered choices, corrected answers, or rehearsed a narrative, apparent inside knowledge can reflect information transfer rather than firsthand involvement. Reconstructing the entire interview sequence is therefore necessary to interpret the confession.
Accusatory methods that convey certainty of guilt, deceptive evidence claims, threats or promises, intense repetition, and strong minimization of wrongdoing can create vulnerabilities. Their effects depend on exact conditions, legal restrictions, participant characteristics, and how the techniques are combined. A Campbell systematic review of interrogation and investigative interviewing methods concludes that approaches emphasizing information gathering perform better than accusatory approaches in experimental samples on important outcomes, while acknowledging limitations in transporting laboratory conditions to real investigations.
Youth, intellectual disability, suggestibility, language barriers, some mental-health-related difficulties, exhaustion, and limited understanding of the process can be relevant to vulnerability. None of these characteristics by itself establishes that a confession is false. The assessment asks whether the actual interview imposed identifiable pressures and whether independently verifiable details support or contradict the statement. The 2025 scientific review recommends more complete recording and protective interviewing procedures.
Why a Confession Can Change the Whole Case
Once a confession enters a case, it can redirect the investigation, reshape recollection of earlier witness evidence, alter the perceived meaning of forensic findings, and strongly influence charging and plea decisions. A confession may also lead investigators to stop pursuing alternative suspects. If the statement is false, its influence can create new appearances of corroboration because later evidence collection is no longer independent of the confession. The causal history of the evidence matters as much as its final presentation.
Electronic recording from the beginning to the end of a substantive interrogation helps later reviewers distinguish freely supplied information from interviewer-originated facts, understand breaks and pressure, and assess the context of admissions and denials. Video perspective, missing segments, transcription, language interpretation, and local legal rules still matter. The published article Recording Police Interrogations: Psychology, Safeguards, and Evidence explains the procedural and evidentiary issues.
Confirmation Bias and Investigative Tunnel Vision
Confirmation bias is a pattern in which a working belief can affect which information people seek, notice, remember, and interpret. In an investigation, early suspicion may change how an interviewer evaluates denials, which witnesses are recontacted, how alternative accounts are tested, and whether inconsistent evidence receives the same attention as supporting evidence. Tunnel vision refers to a broader case-level narrowing of investigation around a favored hypothesis. These phenomena can occur among experienced professionals because they arise from ordinary cognitive and organizational processes rather than a uniquely defective personality.
Initial assumptions often become anchored by seemingly distinctive behavior, a witness identification, an informant statement, or the reputational characteristics of a suspect. Once the hypothesis influences information collection, later results are harder to treat as independent evidence. A forensic scientist given an investigator's expectation, or a witness told that a particular person has confessed, may interpret ambiguous material through that context. Research on contextual bias in forensic settings supports measures designed to limit exposure to task-irrelevant information, while strength and transferability differ among specialties.
The National Institute of Standards and Technology's human-factors report on latent print examination explains how organizational practice, documentation, and contextual information can influence forensic workflows. Reviewers should distinguish a general demonstrated risk from a finding that bias actually altered a particular examiner's decision. Our specialist article Cognitive Bias in Criminal Investigations: Tunnel Vision and Confirmation Bias covers the psychology and investigative countermeasures.
Why Sincere Professionals Can Reach Reinforcing Errors
A person can be confident, conscientious, trained, and mistaken. Investigators may be rewarded for rapid closure and may have limited resources for testing alternatives. Witnesses often wish to help, experts are asked for decisive conclusions, prosecutors rely on investigators' summaries, and defense teams operate under competing deadlines. These institutional realities can create repeated opportunities for error even when participants believe they are acting responsibly. Psychological accounts of wrongful convictions therefore examine workflows and incentives as well as individual cognition.
A disciplined alternative is structured consideration of competing hypotheses, independent review of major evidence, preservation of disconfirming material, and documented separation of observed results from interpretations. These procedures are not magic debiasing techniques. Their value lies in making assumptions reviewable, reducing unnecessary information leakage, and requiring evidence to be tested against more than one explanation.
Forensic Science, Contextual Bias, and the Appearance of Certainty
Forensic evidence includes many different disciplines with different empirical foundations. DNA profiling based on suitable material, a fingerprint comparison, a toolmark opinion, a microscopic hair comparison, and a medical interpretation of injury cannot be given the same generic claim of accuracy. Reliability depends on validated methods, the quality and quantity of the sample, the task, error rates, examiner training, contextual information, and the reporting language. The National Academies' 2009 report on strengthening forensic science called for rigorous scientific foundations and consistent quality practices.
A match or association can be persuasive while still requiring careful attention to its scope. The probability of observing evidence under a specified proposition is not automatically the probability that the accused person is guilty. Laboratory error rates, likelihood ratios, and declarations of inclusion have meanings tied to a test and a comparison population. A forensic opinion cannot by itself resolve opportunity, intent, alternative transfer routes, or the factual identity of the perpetrator in every case.
Contextual information can shape examination when the signal is ambiguous. A laboratory should decide which information is necessary for the technical task, manage the sequence in which potentially biasing details are revealed, preserve documentation of analysis, and use appropriate independent verification. The NIST report on latent-print examination and human factors offers a systems perspective; its recommendations are especially relevant to describing procedures, not diagnosing the cause of every known forensic error.
Informants, False Accusations, and Incentives
Not every wrongful conviction turns on memory or interrogation. An informant may supply testimony under incentives to obtain leniency, favorable treatment, financial reward, or protection. A witness may intentionally accuse the wrong person, misunderstand events, or reconstruct an account after repeated discussion. These possibilities have different psychological and legal implications. False accusation is intentional only when the evidence supports that conclusion; an erroneous statement can be sincere.
Decision makers should examine the origins of the information, benefits available to the source, previous inconsistent accounts, exposure to case details, and independent corroboration. A story that becomes more detailed after contact with investigators or media requires attention to how its details were obtained. The National Registry of Exonerations report treats perjury and false accusation as a recurring recorded contributing factor but cannot establish the private motives of all participants through its category labels alone.
Juries, Narratives, and the Integration of Unreliable Evidence
Jurors must evaluate evidence within the instructions and legal standards of their jurisdiction. They seldom witness the investigation itself. Instead, they hear selected testimony, exhibits, and arguments arranged into competing narratives. A coherent story can feel more persuasive than isolated uncertainties, especially when several witnesses use similar language or when a confession appears to settle disputed identity. The psychological problem is that apparent coherence may reflect shared information or earlier investigative decisions rather than independently confirmed truth.
A courtroom claim may also be conveyed with more certainty than the original evidence warranted. A witness whose first identification was hesitant may express high confidence after several years; a forensic report may condense technical uncertainty into a categorical statement; a detective may summarize a lengthy interrogation without its disputed exchanges. Jurors need access to the original procedural record to evaluate how evidence changed. Experts and instructions can explain known mechanisms, while factual credibility and guilt remain for legally authorized decision makers.
Information about eyewitness confidence and accuracy illustrates the difficulty. Confidence and accuracy are conditionally related at an uncontaminated initial identification, but a confident in-court identification may reflect repeated exposure or feedback. The National Research Council's recommendations for courts call for consideration of original identifications, initial confidence, procedural documentation, and expert framework testimony where appropriate. Laboratory findings can clarify mechanisms without supplying a numerical probability that a particular defendant is innocent.
Plea Bargaining: Why Innocent Defendants May Accept Guilt
Most wrongful-conviction discussions focus on contested trials, yet many criminal prosecutions are resolved through guilty pleas. An innocent accused person may confront a difficult choice between the uncertainty of a trial, possible severe punishment, prolonged detention, legal expense, and a plea offer that promises a smaller or more predictable consequence. Decision-making under such conditions may favor minimizing immediate or catastrophic risk, even when a person maintains factual innocence.
Controlled experiments have demonstrated that innocent participants can accept an admission-like deal under pressure and incentives. A well-known study by Dervan and Edkins on the innocent defendant's dilemma illustrates the psychological mechanism but does not establish the frequency of false guilty pleas in actual criminal courts. Pleas are governed by local standards for voluntariness, factual basis, defense representation, and judicial acceptance; the experimental situation is only a partial model of those legal conditions.
Guilty pleas and false confessions should be distinguished. A confession may be used as investigative evidence, whereas a plea is a formal legal act with defined consequences. An individual might make both, either, or neither. A plea can be legally valid under the applicable jurisdiction while later facts raise questions about innocence. Assessing the choices available requires case-specific information about counsel, disclosed evidence, procedural rules, and actual sentencing exposure.
Police and Prosecutorial Misconduct, Disclosure, and Institutional Error
Wrongful convictions may involve intentional wrongdoing, negligent practice, structural pressure, or ordinary mistakes amplified by poor documentation. These are different causal and normative categories. The term official misconduct in exoneration data is a registry classification that should not be equated automatically with a judicial liability finding in each case. Psychological mechanisms such as motivated reasoning and loyalty to a theory may interact with policies governing evidence preservation, witness handling, forensic review, and supervision.
Disclosure rules, standards of professional responsibility, and remedies for violations are legal matters. In federal US practice, the Department of Justice's Justice Manual guidance on disclosure and trial issues discusses evidence-handling and prosecutorial obligations under applicable doctrines. State and international systems differ in what must be disclosed, when disclosure is required, and what legal remedy follows. Psychological analysis can explain how hidden inconsistencies undermine reliable fact-finding while the legal rule must be identified jurisdiction by jurisdiction.
Institutional safeguards include clear evidence inventories, complete recording, accessible defense review, procedures for documenting alternative leads, independent laboratory oversight, and correction mechanisms for newly discovered evidence. Their value is greatest when responsibility is distributed across a system rather than concentrated in one professional's intuition. A reliable system should make it possible to reconstruct an evidentiary decision years later.
How Different Errors Interact: A Hypothetical Case
Consider an illustrative case in which a robbery occurs at night. A witness observes the perpetrator briefly, hears a later news description, and identifies the person whose photograph investigators have prominently circulated. An interviewer treats that identification as decisive and tells a suspect that the evidence is overwhelming. During repeated questioning, the interviewer provides facts about the scene, and the suspect eventually accepts a statement containing those facts. A laboratory examiner then receives the case narrative before interpreting an inconclusive trace.
At trial, the witness identifies the accused confidently, the confession includes apparent inside details, and the forensic opinion is described as consistent with the prosecution's case. A juror could reasonably perceive three corroborating strands. Yet each may have inherited information from the same early assumption. The scientifically relevant task is to reconstruct the chronology: first description, first identification, interview disclosures, exact wording of the alleged admission, laboratory information access, and the uncertainty in every observation.
This illustration is fictional and represents a possible pathway, not a claim about a specific defendant or the inevitable effect of an investigative practice. Some cases contain powerful independent physical evidence; others show mixtures of reliable and unreliable statements. A well-supported forensic reconstruction asks which errors are demonstrated, which are plausible, and which remain untested.
How Scientific Evidence Should Be Evaluated in a Particular Case
The first step is to establish the question being answered. Is the issue whether a witness's identification process was suggestive, whether a confession was voluntary under law, whether a forensic method is scientifically valid, whether an expert applied the method reliably, or whether new evidence proves innocence? These questions require different methods and different decision makers. A scientist's view that a procedure increases the risk of error is not an automatic legal conclusion, and a legal ruling on admissibility is not a laboratory validation.
Next, separate direct observations from interpretations. Original witness descriptions, recorded interviews, contemporaneous notes, photographs, instrument data, complete laboratory files, discovery logs, and preserved physical evidence can illuminate the sequence of decision making. Later affidavits, trial summaries, retrospective recollections, and opinions can also matter, but they may be affected by intervening experiences. Chronology is critical to identifying when information or confidence changed.
Finally, test alternative hypotheses and specify what would falsify a claim. If a confession contains a nonpublic fact, establish whether the fact was independently supplied or disclosed by investigators. If an eyewitness made a correct identification in a blind procedure, record the full circumstances and examine independent corroboration. If a forensic score is reported, identify the validated task, comparison standard, false-positive and false-negative rates, relevant population, and uncertainty. The phrase 'the evidence was biased' should be replaced, where possible, with a documented process and a specified inferential limitation.
Predictive Validity, Base Rates, and the Limits of Quantifying Error
Psychological research often compares error frequencies between experimental conditions or measures associations across case records. A higher false-identification rate in one protocol does not assign a probability of guilt to a particular person. False positives and false negatives describe decisions relative to independently established ground truth; definitions change with the task, threshold, and sample. Calibration describes whether stated probabilities match observed frequencies within a relevant population. These concepts should be used only when the data and assessment design support them.
Selection is especially serious for exoneration studies. Known exonerations may overrepresent cases in which physical evidence survived, legal assistance was accessible, or a particular error could be established. Also, several factors may co-occur, and investigative records may be incomplete. The National Registry's reporting is invaluable for identifying patterns but should not be used to derive a defendant-level wrongful-conviction prediction score.
Population differences matter. Laboratory experiments may involve short videos and undergraduate participants; real cases involve trauma, language differences, varying crime durations, years of delay, attorney advice, cultural context, and complicated evidence. Comparisons between jurisdictions require attention to procedural law, recording mandates, rules about interrogation and counsel, criminal court structure, and the availability of postconviction remedies. A safeguard supported by controlled evidence warrants implementation and evaluation, not the claim that its effect is identical in every system.
Culture, Disability, Youth, and Unequal Exposure to Risk
Communication style, language proficiency, disability, developmental stage, and familiarity with legal institutions can shape comprehension, responsiveness to authority, and participation in investigative procedures. These variables deserve assessment as aspects of the specific interaction, not as labels for inherent unreliability or culpability. A person's diagnosis cannot prove a confession false, establish criminal intent, or serve as a general marker of danger.
Young people may have less experience evaluating long-term legal consequences and may face particular difficulties resisting authoritative questioning. Interpretation and access accommodations can affect how evidence is elicited and recorded. For eyewitnesses, cross-group familiarity and differences in interviewing practices may be relevant, but demographic categories are imperfect proxies for personal experience. The goal is equitable access to reliable procedures and transparent evaluation for each person.
Disparities also arise at institutional levels: access to counsel, resources for independent experts, forensic testing, language services, and postconviction review varies. Good psychological research distinguishes individual cognitive mechanisms from the effects of unequal institutional conditions. The related article Cultural Bias and Fairness in Forensic Psychological Assessment addresses interpretation and fairness in evaluations.
Evidence-Based Safeguards: Preventing Wrongful Convictions
During Investigation
Investigators can document multiple hypotheses, pursue potentially disconfirming leads, avoid implying that a suspect's guilt is established, and preserve complete contemporaneous witness statements. Identification procedures should use appropriate fillers, blind administration when possible, clear instructions that the perpetrator may be absent, verbatim initial-confidence records, and video documentation. These procedures follow the National Academies' eyewitness recommendations and reduce opportunities for evidence contamination.
During Suspect Interviews
Evidence-informed interviewing emphasizes gathering an accurate account, open questions, careful documentation, appropriate breaks, and avoidance of coercive pressure or misleading representations that can increase unreliable admissions. Recording should preserve the entire substantive encounter, including preliminary discussion, denials, disclosures of case facts, and any breaks. Vulnerabilities relating to youth, disability, health, and language should be accommodated. The 2025 American Psychology–Law Society review and the Campbell systematic review support these priorities, subject to known evidence gaps.
In Forensic Laboratories and Prosecutorial Review
Laboratories should use validated methods, proficiency testing, documented uncertainty, technical review, and controls on task-irrelevant contextual information. Investigative and prosecutorial teams should preserve original data and identify which corroborating items were obtained independently. Reviewers should ask whether the evidence genuinely distinguishes rival explanations or is merely consistent with the preferred narrative. The 2009 National Academies report on forensic science and NIST human-factors work provide evidence and institutional guidance.
During Plea Negotiations, Trial, and Appeal
Defendants and counsel need meaningful access to the evidence and its limitations under governing discovery rules. Courts can consider appropriate expert explanations, understandable instructions, accurate presentation of scientific uncertainty, and scrutiny of identification and interrogation histories. Appellate and postconviction procedures can preserve avenues for testing new evidence and correcting demonstrable error. Precise safeguards depend on applicable law; a safeguard recommended by science is not automatically mandatory in all jurisdictions.
Myths and Misconceptions
Myth: An Innocent Person Would Never Confess or Plead Guilty
An admission can arise under fatigue, pressure, misunderstanding, situational incentives, or distorted expectations about future consequences. A legal plea may also reflect an attempt to avoid an extreme trial risk. Neither phenomenon proves a particular defendant innocent. They establish why the content and circumstances of statements must be evaluated rather than treated as self-validating.
Myth: A Confident Eyewitness Must Be Correct
Confidence has evidentiary meaning only in context. Initial confidence under sound uncontaminated procedures can be informative; later confidence may be influenced by feedback, repeated exposure, and the passage of time. The witness's sincerity and the accuracy of the identification remain separate questions.
Myth: Several Consistent Pieces of Evidence Must Be Independent
Consistency can arise because sources share information. A witness may learn that a suspect confessed; a confession may include facts supplied by an interviewer; an examiner may know which result detectives expect. Independent corroboration requires independent evidence generation and a documented chain of information.
Myth: A Wrongful Conviction Always Means Every Actor Was Dishonest
Some cases involve misconduct; others involve perceptual limitations, mistaken beliefs, poor procedures, and systemic incentives. Distinguishing deliberate falsification, negligent practice, and error is essential for identifying what failed and selecting effective remedies.
Myth: A Scientific Study Can Tell Whether a Particular Defendant Is Guilty
A study estimates patterns under specified conditions. A forensic opinion may describe limitations in a procedure or the reliability of an inference. Neither automatically supplies the legal verdict. Individual-case conclusions require the full evidence, procedural history, and application of the jurisdiction's law.
Frequently Asked Questions
What is the most common psychological cause of wrongful convictions?
There is no single cause that can responsibly be ranked across all convictions. Mistaken identification, confession-related error, bias, misleading evidence, and decisions shaped by incentives can occur alone or in combination. Their recorded prevalence depends on the type of case, how wrongful convictions are identified, and the reporting system. The major research lesson is that errors compound across stages of a case.
Can eyewitness testimony be accurate?
Yes. Witness testimony can be highly useful, and some initial identifications are accurate and informative. Reliability depends on the conditions of observation and the fairness of subsequent procedures. The appropriate approach is to evaluate a witness's first account, the identification history, documented confidence, and corroboration instead of assuming either universal accuracy or universal unreliability.
Why do innocent people confess to crimes?
Pressure, immediate incentives, fatigue, interrogation tactics, misunderstanding, and individual vulnerabilities can contribute. A suspect may say what the interviewer seeks to end questioning or may become uncertain about memory after suggestive methods. These pathways are described in scientific reviews; determining whether a particular confession is false still requires case evidence.
What is investigative tunnel vision?
Tunnel vision is a narrowing of investigative attention around a favored suspect or explanatory theory. It can affect the search for evidence and the interpretation of results. Its existence in a specific case must be demonstrated through records of investigative choices, alternative leads, information exposure, and the handling of inconsistent evidence.
Does DNA evidence solve every wrongful conviction?
DNA testing can be decisive when suitable material exists and the relevant biological inference has been validated. Many crimes leave no usable DNA, some evidence was never retained, and biological transfer may require interpretation. Other reliable methods, disclosures, and independent evidence can also uncover error. The presence of DNA evidence does not exempt a case from careful evaluation.
Do polygraphs or body-language analysis reliably expose a lying witness?
No simple gesture, level of nervousness, or polygraph output establishes whether a particular statement is true. Deception-related behaviors overlap widely across truthful and deceptive individuals, and test results depend on protocols and context. The broader evidence is examined in Detecting Deception: Why Lie Detection Is Hard.
Is a conviction overturned for a legal error necessarily proof of factual innocence?
No. Reversal, vacatur, dismissal, acquittal, official exoneration, and judicial findings of factual innocence have different definitions and implications. An overturned conviction can reflect an error serious enough to require another proceeding, while innocence may remain unresolved. Terminology should track the actual legal disposition and the evidence.
How can researchers study wrongful convictions without knowing every hidden error?
Researchers combine verified exoneration cases, archival case reconstruction, laboratory experiments with known ground truth, field studies, professional consensus reviews, and procedural evaluations. Each method answers different questions. A laboratory study can demonstrate vulnerability to a manipulation; exoneration records show that comparable mechanisms can arise in real cases. Neither method alone supplies a universal prevalence estimate.
Are psychological findings about wrongful convictions the same in every country?
General constraints on perception, memory, attention, and judgment can inform many jurisdictions, but investigative procedures, admissibility doctrines, rules on counsel and disclosure, plea practices, and postconviction relief differ. Legal conclusions and recommended safeguards must be matched to the governing jurisdiction and tested against local practice.
What should be examined first in a potentially wrongful conviction?
A careful review reconstructs the earliest unaltered evidence and the exact sequence in which information moved between witnesses, interrogators, laboratory personnel, prosecutors, and fact-finders. It then separates direct observations, documented procedures, later interpretations, and independently verifiable corroboration. Legal review must also address preservation, disclosure, appeal deadlines, and the procedural rules that govern the actual case.
Related Articles
References
Catlin, M., Wilson, D. B., Redlich, A. D., Bettens, T., Meissner, C. A., Bhatt, S., & Brandon, S. E. (2024). Interview and interrogation methods and their effects on true and false confessions: A systematic review update and extension. Campbell Systematic Reviews, 20(4), e1441.
Dervan, L. E., & Edkins, V. A. (2013). The innocent defendant's dilemma: An innovative empirical study of plea bargaining's innocence problem. Journal of Criminal Law and Criminology, 103(1), 1–48.
Gudjonsson, G. H. (2021). The science-based pathways to understanding false confessions and wrongful convictions. Frontiers in Psychology, 12, 633936.
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.
Kassin, S. M., Drizin, S. A., Grisso, T., Gudjonsson, G. H., Leo, R. A., & Redlich, A. D. (2010). Police-induced confessions: Risk factors and recommendations. Law and Human Behavior, 34, 3–38.
National Institute of Standards and Technology. (2012). Latent print examination and human factors: Improving the practice through a systems approach. NIST Interagency/Internal Report 7842.
National Registry of Exonerations. (2026). 2025 annual report. University of Michigan Law School, Michigan State University College of Law, and University of California Irvine Newkirk Center for Science & Society.
National Research Council. (2009). Strengthening forensic science in the United States: A path forward. National Academies Press.
National Research Council. (2014). Identifying the culprit: Assessing eyewitness identification. National Academies Press.
U.S. Department of Justice. (n.d.). Justice Manual: Issues related to trials and other court proceedings. U.S. Department of Justice.