Simultaneous vs. Sequential Lineups: What the Evidence Shows
Author: Ukrainian Psychological Hub · Published: October 11, 2026 · Editorial Policy
Ukrainian Psychological Hub · October 11, 2026 · Evidence review | Forensic & Legal Psychology
Simultaneous and sequential police lineups ask an eyewitness to make the same consequential decision in two different ways. In a simultaneous lineup, the witness views every lineup member together. In a sequential lineup, the witness examines members individually, ordinarily making a decision about each person before seeing the next. The procedures can influence whether a witness chooses someone, whether the actual perpetrator is recognized, and whether an innocent person is selected. Their comparative accuracy has been debated for four decades.
The central finding is that no single statistic makes one format universally superior. Older meta-analyses often found that sequential presentation reduced mistaken selections, but it also tended to reduce correct selections. Later research separated witnesses' willingness to choose from their ability to distinguish the perpetrator from an innocent suspect. Some receiver operating characteristic (ROC) studies, including a 2025 synthesis of laboratory and field findings, favor simultaneous lineups on discriminability. The National Academies' comprehensive review therefore emphasized the measurement problem and advised caution in declaring either format categorically preferable.
For a court, investigator, defense attorney, policymaker, or researcher, the relevant question is more precise: Which procedure, under which conditions, produces the best balance of correct suspect identifications, mistaken suspect identifications, and meaningful decisions not to identify anyone? This article compares the evidence, explains why studies sometimes disagree, and distinguishes psychological findings from procedural rules and legal judgments.
Simultaneous vs. sequential lineups: the essential comparison
A simultaneous photo lineup presents the suspect alongside several fillers—people who are not suspected of committing the offense but are included to test the witness's recognition. All images can be inspected together. A sequential photo lineup presents those individuals one at a time. Depending on the protocol, the witness may make a yes/no judgment for each image, continue after a selection, revisit images under specified rules, or complete only one viewing. These procedural details alter the task and must be described when interpreting research.
A live lineup involves physical persons, while a photo array involves photographs; either can theoretically be displayed simultaneously or sequentially. A showup, in which one person is presented for identification, is a different procedure and should not be treated as an ordinary sequential lineup. The foundational distinction is described by the National Institute of Justice.
In either format, police should have an independently grounded suspect before assembling the lineup. Fillers are not competing suspects. A witness's choice is evidence arising from a memory test, subject to contamination, error, and case-specific scrutiny. An identification alone does not establish that the selected person committed a crime, and a rejection does not conclusively establish innocence.
What changes when faces are viewed together?
Simultaneous presentation permits direct comparison of multiple faces. Researchers have proposed that this could encourage a relative judgment: selecting the person who looks most like the remembered perpetrator, even when the perpetrator is absent. That mechanism is plausible but does not describe every simultaneous choice. Seeing several faces together can also allow comparison of distinctive, diagnostically useful features and improve discrimination under some conditions.
Sequential presentation has been hypothesized to encourage an absolute judgment, with each face evaluated against the witness's memory rather than against neighboring faces. Yet sequential choices can still depend on earlier images, expectations about the number of images, how strictly the witness applies a decision threshold, and whether repeated viewings are allowed. Studies of signal detection and recognition support a more nuanced account than the simple relative-versus-absolute slogan. See Meissner and colleagues' experimental analysis and Palmer and Brewer's reanalysis.
A comparison of the main features
In the simultaneous format, lineup members are visible together; within-lineup comparisons are possible; the decision may occur after inspecting all faces; and the witness can potentially compare facial features repeatedly. In the sequential format, faces appear individually; the witness's decisions depend more explicitly on the sequence and any stopping or continuation rule; and an administrator may limit or allow a second pass. Both formats can be conducted blind, accompanied by neutral instructions, documented on video, and followed by an immediate confidence statement.
Neither format automatically guarantees a fair lineup. The suspect may stand out, fillers may be poorly selected, the administrator may convey cues, or a witness may have seen the suspect's photo before the procedure. These features are often more important to the integrity of a particular identification than the simultaneous–sequential label by itself. The 2020 American Psychology-Law Society policy paper treats careful construction, administration, and documentation as system-wide requirements.
The accuracy question has several different answers
The word accuracy is used for distinct outcomes. A study that counts only mistaken choices can prefer one procedure while a study examining true-positive identifications or discriminability prefers another. The distinction matters because every lineup can yield three broad kinds of decision: selecting the actual perpetrator, selecting an innocent person, or selecting no one. An innocent selection can be either a filler identification or, in a culprit-absent lineup, an innocent-suspect identification.
Culprit-present and culprit-absent lineups
In laboratory experiments, researchers can construct culprit-present lineups that include the person seen in a simulated offense and culprit-absent lineups that exclude that person. They can therefore calculate correct culprit identifications, mistaken suspect identifications, filler selections, and correct rejections using known ground truth.
In real police cases, investigators ordinarily do not know with certainty whether the suspect in the lineup committed the crime. They may possess independent evidence, but the lineup itself cannot reliably establish the suspect's ground-truth status for scientific calculation. This is the key reason that laboratory hit rates and field identification rates cannot be compared as though they measure identical outcomes. The distinction is central to the Houston field analysis and the 2025 field–laboratory comparison.
Correct identifications, false identifications, and filler identifications
A correct identification occurs when the witness selects the actual perpetrator from a culprit-present lineup. A false suspect identification occurs when the witness selects an innocent designated suspect. A filler identification occurs when the witness selects a known-innocent lineup member other than the suspect. Filler choices are important evidence of error in the procedure, but the proportion of filler identifications is not the same thing as the proportion of innocent suspects wrongfully identified.
In a field study, a filler is known to be innocent by construction, so filler identification is directly observable. The true innocence or guilt of the suspect is typically unknown. Investigators therefore cannot simply use the rate of selecting suspects as a true-positive rate or the rate of selecting fillers as the real-world false-suspect rate without assumptions. The 2015 randomized field experiment makes this distinction consequential.
Response bias and discriminability
Response bias describes a witness's threshold for deciding that a face matches memory closely enough to select. A conservative threshold yields fewer choices overall, potentially reducing both correct and false identifications. A liberal threshold yields more choices, potentially increasing both. Discriminability concerns the ability to distinguish the perpetrator from an innocent person.
A procedure can reduce false identifications mainly by persuading witnesses to choose less often, while leaving their underlying ability to tell guilty from innocent lineup targets unchanged. That is a real practical effect, but it is analytically different from improving memory discrimination. Palmer and Brewer's analysis of 22 experiments found that sequential presentation produced a more conservative response criterion without improved discriminability in the dataset they examined.
Why the early research favored sequential lineups
Early laboratory research was motivated by the danger of relative judgment when the perpetrator is absent. Sequential presentation appeared promising because it required witnesses to evaluate images individually, reducing the temptation to choose merely the closest match within the array.
A 2001 meta-analysis pooled 30 comparisons involving 4,145 participants. It reported that simultaneous presentation produced more correct perpetrator identifications in culprit-present lineups, whereas sequential presentation produced more correct rejections when the perpetrator was absent. The authors examined moderators related to the resemblance of laboratory procedures to real-world practice.
A later meta-analysis of 72 tests involving 13,143 participants found that sequential lineups were less likely to produce a suspect selection, yet suspect identifications obtained from them had a higher diagnosticity ratio under the analyses used. The review also documented the cost in correct identifications. These findings helped motivate policy changes and police training, but they did not settle how to compare procedures across different decision thresholds.
The strongest valid interpretation of these older results is that sequential presentation often altered the pattern of choosing in a manner that reduced some errors. Calling this universally superior memory accuracy would conflate response policy with discrimination. A lineup that yields fewer false selections can be advantageous under some priorities, particularly when avoiding innocent-suspect identifications receives high weight, even if fewer perpetrators are identified.
What receiver operating characteristic research changed
Receiver operating characteristic analysis evaluates discrimination across multiple response thresholds rather than comparing formats at just one chosen threshold. In many lineup experiments, confidence ratings create a set of thresholds: higher-confidence identifications represent stricter selection criteria, while lower-confidence identifications include more liberal choices.
The major methodological insight is that two lineups can produce different correct- and false-identification rates because witnesses adopt different thresholds, even if their memory discriminability is equal. Comparing a single diagnosticity ratio therefore does not fully isolate the ability to separate guilty from innocent suspects. This issue is explained in the National Academies' analysis of applied eyewitness research.
In a 2012 ROC study, Mickes, Flowe, and Wixted reported better discrimination for simultaneous than sequential lineups in the experimental conditions examined. That result challenged the assumption that sequential superiority established by conventional ratios meant stronger identification memory.
ROC methods have their own assumptions. Analysts must decide how to handle filler selections, lineup size, confidence thresholds, and culprit-absent false alarms. Some eyewitness studies estimate innocent-suspect selections from filler identifications and assumed probabilities rather than observing every false-suspect identification directly. The appropriateness of those estimates depends on lineup fairness and the underlying model. Different analytical choices can produce different substantive interpretations, so ROC results require methodological scrutiny rather than automatic acceptance.
Why the same evidence can lead to different policy judgments
Suppose two hypothetical procedures produce 80 and 65 correct identifications, respectively, and 20 and 10 false-suspect identifications, respectively, in equally constructed sets of test lineups. The first produces more correct identifications but also more false selections; the second produces fewer false selections but misses more perpetrators. These numbers are illustrative, not estimates from a particular experiment.
A policymaker may value the reduction in mistaken identifications while another may weigh the loss of successful identifications more heavily. Research can estimate the trade-off but cannot, by itself, establish a universal legal or ethical cost ratio between these errors. The decision also depends on the prevalence of innocent suspects among lineups, available corroborating evidence, and the safeguards attached to either procedure.
The National Academies' 2014 review explicitly concluded that then-available evidence did not justify a definitive endorsement of one presentation method and urged care before changing procedure solely on the claimed superiority of a format. Its recommendations emphasized stronger safeguards and better research.
What randomized police field studies show
Field research is essential because mock crimes differ from real investigations. Genuine witnesses face different emotions, delays, stakes, media exposure, interviewing histories, and memory conditions. Yet field studies also face a severe ground-truth problem: the designated suspect's guilt is generally uncertain.
The four-jurisdiction randomized trial
Wells, Steblay, and Dysart's 2015 study randomly assigned 494 eyewitnesses to simultaneous or sequential, double-blind photo lineups in four U.S. police jurisdictions. The sequential protocol reflected operational practice, including continuation through images and procedures for handling multiple choices.
Suspects were selected at similar rates across conditions—approximately one-quarter overall—with no statistically significant difference between presentation formats. Sequential lineups produced fewer filler selections: 11% versus 18% with simultaneous presentation. Among witnesses who selected someone, a sizable fraction chose a known-innocent filler. The observed filler difference supports a practical error-reduction result for that protocol, though it does not directly identify which innocent suspects would have been falsely accused.
That distinction is critical. Lower filler-selection rates can signal more conservative decisions; they do not prove that all sequential formats outperform all simultaneous formats in discriminating actual perpetrators from innocent suspects. The field experiment also did not randomly assign crimes or establish ground truth for each suspect.
The Houston Police Department experiment
The Houston analysis published in 2016 investigated eyewitness identifications made in robbery cases using simultaneous and sequential photo arrays and different blinded administration conditions. For a double-blind subset, researchers compared 187 simultaneous and 161 sequential lineups and analyzed choices and contemporaneous confidence through signal detection models.
The authors' modeling suggested that, if there was a discriminability advantage between formats, it likely favored simultaneous presentation. They also found evidence that high confidence, documented promptly under fair procedures, could be informative. These results differed from simplistic claims that sequential procedures are invariably superior.
The estimates rested partly on statistical modeling because the actual guilt of lineup suspects was unknown. Furthermore, participating investigations came from one police department and involved particular case types and operating procedures. Model-derived field estimates therefore require replication, transparent assumptions, and caution when generalized to other agencies or legal settings.
Why the two field studies are not contradictory in a simple way
The four-jurisdiction trial measured observable outcomes including how often witnesses chose known-innocent fillers. The Houston study asked additional questions about discrimination and accuracy using models in the absence of known suspect guilt. A reduction in filler selections and a simultaneous advantage in estimated discriminability can coexist; they concern different estimands.
Protocols varied in details of lineup administration, repeat viewing, confidence recording, and sample composition. Both investigations are important, and their apparent disagreement is best understood through design, outcome definition, model assumptions, and the limits of case-based inference.
Evidence published through 2025
A 2025 review and comparative analysis by Wixted and Mickes revisited the Houston field data alongside a newer, methodologically aligned mock-crime experiment. It reported broadly convergent patterns, including stronger simultaneous-lineup discriminability in the compared data and evidence that initial high-confidence suspect identifications can have substantial diagnostic value under controlled conditions.
The authors specifically confronted criticisms of the Houston model, including concerns about unfair lineups and potential bias. Their analyses offer an important modern case for simultaneous presentation on discrimination-based measures. They do not convert one study, one model, or one institutional protocol into a definitive rule governing every lineup.
A separate 2019 program of five experiments systematically varied lineup presentation alongside other design features and reported better discriminability from simultaneous than sequential presentation in a key comparison. Other studies have yielded null differences or findings that depend on rules and measurement. The cumulative record supports a serious evidence-based debate about presentation format, with stronger agreement on core administrative safeguards.
Claims of a settled universal answer should therefore be assessed against what each study measures: selection frequency, correct identifications, filler identifications, estimated innocent-suspect choices, ROC discriminability, or witness confidence conditioned on the way the test was conducted. Those are connected but not interchangeable quantities.
Psychological mechanisms: how could format change a witness's decision?
Eyewitness identification is a recognition-memory judgment influenced by perception during the original event, memory retention, decision thresholds, and features of the later testing procedure. Simultaneous and sequential lineups primarily alter retrieval and decision conditions, not the historical quality of the witness's original exposure to the perpetrator.
Relative comparison and memory matching
In a simultaneous lineup, a witness may compare lineup members to one another. If a witness feels obligated to choose someone, this can be hazardous when the actual perpetrator is absent. The same presentation arrangement can also aid discrimination by revealing that a face has distinctive features that are not shared by others. The effect of comparison depends on what information the witness retains and how the lineup is assembled.
Conservative responding in sequential presentation
Requiring individual decisions may raise the threshold for a positive identification. If witnesses become more reluctant to choose, false identifications can fall together with correct ones. Evidence from Meissner and colleagues and Palmer and Brewer is consistent with criterion shifts playing a substantial role.
Sequential position effects
A sequence is also a temporal context. A witness may apply a strict standard to early images, anticipate a better match later, or relax the threshold after passing over several faces. A 2018 experiment on serial position and ROC outcomes observed shifts in response criterion across sequential lineup positions that were not found in the simultaneous condition. The study cautions against assuming that sequential administration yields uniform decisions regardless of image order.
Stopping rules and second viewings
Research sometimes uses a strict stopping rule: the lineup ends after the first selection. Other sequential protocols allow all faces to be viewed even after an initial choice; some allow a second pass. These versions should not be pooled conceptually without examining their effects. What counts as a sequential lineup in a laboratory may differ substantially from what a police department implements.
When an eyewitness changes an answer after reviewing images, the initial judgment, final judgment, number of passes, and intervening instructions become important evidence. Protocol fidelity is part of scientific validity, not an administrative footnote. The 2015 field experiment explicitly sought to test sequential practice as it was actually conducted.
The role of lineup fairness
A fair lineup is designed so the witness's memory, rather than an obvious difference between the suspect and fillers, guides the decision. If the suspect is the only person closely matching the witness's earlier description, a choice may reflect lineup construction rather than reliable recognition. This problem applies to both formats.
Investigators should select fillers according to a consistent procedure grounded in the witness's description and relevant resemblance criteria; avoid making one face visually distinctive through image quality, clothing, background, or markings; and document who selected the fillers and how. Science-based principles and local rules should inform exact composition, including the number of fillers. The 2020 expert recommendations provide a central professional foundation.
The ability to assess fairness is especially important when researchers interpret rates of filler selection as proxies for erroneous suspect selection. If lineup positions or appearances make the suspect disproportionately salient, assumptions that each innocent member is equally likely to attract a mistaken choice are questionable.
Double-blind administration matters in both formats
Double-blind administration means that the witness is not told who the suspect is and the person administering the identification does not know which lineup member is the suspect. In some systems, a blinded alternative prevents an otherwise knowledgeable administrator from seeing which image the witness is examining. Both arrangements aim to reduce deliberate or inadvertent cues.
A knowledgeable administrator might unintentionally guide a witness by gaze, timing, tone of voice, facial reaction, or confirmatory remarks. Such influence can affect choosing or later recollections of certainty. Double-blind administration offers an important protection independent of whether images appear together or sequentially.
The National Academies' recommendations call for blind procedures, standardized witness instructions, immediate confidence documentation, and full recording. The American Psychology-Law Society's 2020 recommendations likewise emphasize preserving the integrity of the first memory test. These are consensus-oriented procedural recommendations, not claims that any one safeguard makes eyewitness evidence infallible.
Neutral instructions, confidence, and documentation
Before an identification procedure, the witness should receive standardized, understandable instructions that the perpetrator may be absent, that selecting no one is permissible, and that the investigation continues whether or not an identification is made. The witness should not be steered toward the idea that police already know the right answer. The National Academies report provides an authoritative rationale.
An immediate confidence statement, recorded in the witness's own words before feedback, preserves information about the first identification. Later confidence may be affected by investigators' reactions, discussion with others, or repeated exposure to a suspect. Confidence research is substantial, but the format comparison does not reduce to confidence alone; a separately dedicated article should address confidence–accuracy relationships comprehensively.
Whenever feasible and lawful, agencies should retain the full recording or an equally reliable contemporaneous record of the initial instructions, complete lineup, order of faces, choices, exact words, confidence report, administrator actions, and deviations from protocol. The record should include a nonidentification as carefully as it records a positive selection.
Laboratory studies versus real investigations
Laboratory experiments offer the rare advantage of known ground truth. Researchers can randomly assign participants, manipulate culprit presence and viewing conditions, hold filler characteristics stable, and observe false-suspect selections directly. Their limitations include simulated stakes, selective samples, often brief delays, and differences between staged encounters and lived criminal events.
Field studies use genuine eyewitnesses and police procedures, enhancing ecological relevance. But investigators cannot usually label a selected suspect as guilty or innocent from the identification itself. Cases are heterogeneous, and procedural deviations may be associated with witness or investigator characteristics. Missing information and post-identification events may distort interpretations.
Neither study type replaces the other. Strong policy evaluation triangulates controlled laboratory evidence, blinded field experiments, transparent model comparisons, recorded procedural implementation, and independent replication. The 2025 lab–field analysis represents one approach to this triangulation while also illustrating the assumptions on which field conclusions depend.
Statistical significance, practical size, and precision
A statistically significant difference does not indicate that one procedure is always better in every population. Conversely, a nonsignificant difference does not establish equivalence. Investigators and researchers should report effect sizes, uncertainty intervals where available, sample sizes, the exact outcome, and any relevant changes in threshold or base rate.
If a study reports fewer filler selections, the practical benefit should be evaluated alongside changes in culprit identifications and no-choice decisions. If a study reports improved ROC discriminability, readers should inspect how the ROC was constructed, whether models were used, and whether the result persists across lineup conditions. The scientific unit of comparison is the complete identification task.
Does one lineup type reduce wrongful convictions?
Mistaken eyewitness identification is a recognized contributor to wrongful convictions. However, the effects of changing lineup format on wrongful-conviction rates cannot be inferred directly from laboratory false-alarm rates or field filler-selection differences. Convictions depend on investigation, disclosure, corroboration, charging, plea decisions, expert evidence, trial procedure, and judicial findings.
A sequential format that reduces one kind of erroneous selection may reduce a route to wrongful accusation in some contexts. A simultaneous format with stronger discrimination in some studies may offer a different advantage. Translating either effect into a number of wrongful convictions prevented requires additional assumptions and justice-system outcome data.
A sound wrongful-conviction prevention strategy therefore improves lineup fairness, blindness, instructions, recording, evidence preservation, disclosure, and independent corroboration regardless of format. It also ensures that decision-makers understand the uncertainty attached to identification evidence. National Academies guidance emphasizes the full pathway from police procedure to courtroom evaluation.
Implications for U.S. police policy
Departments should treat the presentation format as one component of a documented eyewitness-identification policy. Policymakers can evaluate whether a locally adopted simultaneous or sequential protocol has been tested with suitable conditions, whether staff follow it consistently, and whether witnesses receive fair instructions, blind administration, and prompt confidence recording.
An agency considering a switch should specify its desired outcome before interpreting evidence. Is it prioritizing fewer erroneous selections, maximizing discrimination, reducing suspect-identification errors at a defined threshold, or improving completeness of records? Each objective demands different measurements. Training should explain why a procedure works and what it cannot establish.
Policies should also address image order, number and matching of fillers, continuation and stopping rules, repeat viewings, translations and accessibility, known exposure to suspect images, multiple witnesses, and retention of electronic audit records. The American Psychology-Law Society's 2020 paper is a useful professional starting point, with the National Academies supplying independent institutional guidance.
State law, evidence rules, local departmental directives, and judicial decisions vary. Some U.S. jurisdictions regulate lineup administration or the instructions given to witnesses, while others use internal guidelines. An agency or attorney should verify the currently controlling local standard instead of treating a psychological recommendation as a universally binding legal requirement.
What lawyers and courts should examine
The first question in a case is what actually happened during the memory test: Was the lineup simultaneous or sequential, and was the selected procedure implemented as specified? The second is whether the array was fair, the administrator blind, the instructions neutral, and the witness's original words and confidence preserved. The third is whether independent evidence corroborates or contradicts the selection.
A court should distinguish an identification made at the first properly administered lineup from an identification made after repeated exposure, feedback, or other contamination. Expert testimony may explain general research on recognition memory, response criteria, and documented procedural factors while acknowledging uncertainty about the individual witness. The expert's role is to clarify science, not to pronounce guilt or innocence from a generalized probability.
Admissibility and suppression are legal questions determined under the applicable jurisdiction's rules and facts. Psychological discriminability is a research construct; a legal finding of reliability or admissibility applies a legal standard. The two must remain distinct, even when scientific evidence informs legal decisions.
Cultural, demographic, and accessibility considerations
An observed advantage in one sample should not be assumed to apply identically across age groups, racial or ethnic groups, cultural contexts, language backgrounds, visual abilities, or neurocognitive profiles. Facial recognition may differ by familiarity, experience, and quality of original viewing. Language comprehension influences whether a witness understands that the perpetrator may be absent.
The key fairness question is whether the entire procedure gives each witness a meaningful opportunity to understand the task and express uncertainty without pressure. An interpreter, accessible presentation, or adjusted timing may be necessary. These safeguards should be documented, and modifications evaluated rather than treated as automatically harmless.
Evidence comparing simultaneous and sequential presentation within every relevant demographic subgroup remains limited. Where subgroup sample sizes are small or results unreplicated, policy claims about differential accuracy should be described as preliminary. Differential impact concerns require actual group-specific validation and analysis of procedural context, not stereotypes about witnesses.
Frequent misconceptions
“Sequential lineups are scientifically proven to be more accurate.”
That conclusion overstates the evidence. Early meta-analyses favored sequential formats under some diagnosticity measures and found fewer false choices, while later ROC studies and field models produced neutral or simultaneous-favoring results on discrimination. The correct conclusion depends on the outcome, study design, and operational protocol. See the 2011 meta-analysis, 2012 ROC study, and 2025 comparison.
“Simultaneous lineups force witnesses to choose the closest-looking person.”
A simultaneous array can encourage relative comparison, but witnesses may reject every person, and comparison can also help detect distinctive features. Neutral instructions and lineup design influence whether a witness treats choosing as obligatory. A mechanism proposed in theory is not a fixed law of every witness's behavior.
“Fewer filler identifications prove fewer innocent suspects are accused.”
Filler selections are observable errors involving known-innocent non-suspects. False identifications of designated innocent suspects have a different definition and are difficult to measure directly in field cases. Inferences between them require assumptions about fairness and selection probabilities.
“A confident identification is always correct.”
Confidence is probabilistic and depends on when and how it was recorded, the fairness of the first test, and the possibility of contamination. An initial confidence statement may be informative under controlled conditions; it is never a guarantee about one case. See Wixted and Mickes (2025).
“Double-blind administration and sequential presentation are the same reform.”
They address different concerns. Blind administration aims to prevent administrator influence. Sequential presentation changes how the witness sees the faces and decides. The former can accompany either simultaneous or sequential presentation.
Frequently asked questions
Which is better: a simultaneous lineup or a sequential lineup?
Research gives different answers for different criteria. Some studies find fewer erroneous selections in sequential lineups, while some newer discrimination-focused analyses favor simultaneous lineups. A defensible choice requires attention to the protocol, outcome definition, and safeguards.
What is the principal disadvantage of simultaneous lineups?
A potential disadvantage is that comparison among faces may encourage a witness to choose the person who looks most like the perpetrator even if the perpetrator is absent. Neutral instructions, fair fillers, and blind administration help reduce this hazard.
What is the principal disadvantage of sequential lineups?
Sequential procedures may reduce correct identifications as well as mistaken selections. They can also introduce position and threshold effects, and their outcomes depend on rules concerning continuation and repeat viewing.
Do sequential lineups produce fewer false identifications?
Some meta-analyses and a randomized field experiment report fewer mistaken or filler choices under sequential procedures. That finding should be stated with the exact kind of error measured, the associated changes in correct identifications, and the uncertainty attached to extrapolating from fillers to innocent suspects.
Why do newer studies sometimes favor simultaneous lineups?
Newer analyses often emphasize discriminability at comparable response criteria rather than a single diagnosticity ratio. Some studies find simultaneous presentation better supports the separation of culprit-present from culprit-absent decisions across confidence thresholds.
Is a photo lineup more reliable than a live lineup?
Photo versus live presentation is a distinct comparison from simultaneous versus sequential format. Reliability depends on construction, administration, witness conditions, and local implementation. An answer for one comparison does not automatically apply to the other.
What does double-blind mean in a lineup?
The administrator does not know which person is the suspect, and the witness is not informed of that identity. This helps prevent suggestion through inadvertent verbal or nonverbal cues and is recommended for both lineup formats.
Must witnesses identify someone?
No. Standardized instructions should communicate that the perpetrator may not be present and that the investigation proceeds regardless of whether the witness chooses a person. Choosing no one is a legitimate response to a memory test.
Can a lineup prove a suspect is guilty?
No. An eyewitness selection is one piece of evidence, potentially helpful but fallible. Criminal guilt is a legal finding under a jurisdiction's applicable burden of proof and evidentiary rules. It is established through the legal process, not by a psychometric or recognition-memory result alone.
Can a court reject evidence because a lineup was simultaneous?
That is a jurisdiction-specific legal question. Presentation format may be relevant to the scientific evaluation of reliability, but admissibility depends on controlling law, case facts, and the complete identification procedure.
Why are eyewitness confidence ratings collected immediately?
Prompt documentation helps preserve the witness's first expressed level of confidence before feedback, repeated exposure, or discussion can alter it. The value of confidence depends on the testing conditions and is not equivalent to proof.
Should police switch to sequential lineups immediately?
A universal switch is not supported by a single uncontested scientific result. Agencies should assess available evidence against their goals and existing safeguards, their applicable laws, and the implementation requirements of the particular sequential or simultaneous protocol.
Evidence-based conclusion
Simultaneous and sequential lineups differ in presentation, response dynamics, and the balance between recognizing perpetrators and avoiding innocent-person selections. The historically influential finding that sequential presentation can reduce mistaken choices remains relevant. Equally relevant is the later evidence that conservative responding is not the same as better discriminability, along with studies finding a simultaneous advantage on discrimination-based measures.
An evidence-based decision starts by identifying the outcome being optimized and by preserving an independently fair, blind, documented first memory test. The reliability of a specific eyewitness identification then depends on the entire history of observation, memory, administration, and corroboration. Research informs that evaluation; the applicable legal standard governs the legal finding.
For broader context, see Eyewitness Testimony: How Memory Works in Legal Cases.
Related Articles
References
Meissner, C. A., Tredoux, C. G., Parker, J. F., & MacLin, O. H. (2005). Eyewitness decisions in simultaneous and sequential lineups: A dual-process signal detection theory analysis. Memory & Cognition, 33, 783–792. https://doi.org/10.3758/BF03193074.
Mickes, L., Flowe, H. D., & Wixted, J. T. (2012). Receiver operating characteristic analysis of eyewitness memory: Comparing the diagnostic accuracy of simultaneous versus sequential lineups. Journal of Experimental Psychology: Applied, 18, 361–376. https://doi.org/10.1037/a0030609.
National Institute of Justice. (2009). Eyewitness identification: Simultaneous vs. sequential lineups. U.S. Department of Justice. https://nij.ojp.gov/topics/articles/eyewitness-identification-simultaneous-vs-sequential-lineups.
National Research Council. (2014). Identifying the culprit: Assessing eyewitness identification. National Academies Press. https://doi.org/10.17226/18891.
Palmer, M. A., & Brewer, N. (2012). Sequential lineup presentation promotes less-biased criterion setting but does not improve discriminability. Law and Human Behavior, 36, 247–255. https://doi.org/10.1037/h0093923.
Steblay, N. K., Dysart, J. E., & Wells, G. L. (2011). Seventy-two tests of the sequential lineup superiority effect: A meta-analysis and policy discussion. Psychology, Public Policy, and Law, 17, 99–139. https://doi.org/10.1037/a0021650.
Steblay, N. K., Dysart, J. E., Fulero, S., & Lindsay, R. C. L. (2001). Eyewitness accuracy rates in sequential and simultaneous lineup presentations: A meta-analytic comparison. Law and Human Behavior, 25, 459–473. https://doi.org/10.1023/A:1012888715007.
Wells, G. L., Kovera, M. B., Douglass, A. B., Brewer, N., Meissner, C. A., & Wixted, J. T. (2020). Policy and procedure recommendations for the collection and preservation of eyewitness identification evidence. Law and Human Behavior, 44, 3–36. https://doi.org/10.1037/lhb0000359.
Wells, G. L., Steblay, N. K., & Dysart, J. E. (2015). Double-blind photo lineups using actual eyewitnesses: An experimental test of a sequential versus simultaneous lineup procedure. Law and Human Behavior, 39, 1–14. https://doi.org/10.1037/lhb0000096.
Wixted, J. T., & Mickes, L. (2025). Suspect identification accuracy from lineups, in the lab and in the field. Cognitive Research: Principles and Implications, 10, Article 60. https://doi.org/10.1186/s41235-025-00670-1.
Wixted, J. T., Mickes, L., Dunn, J. C., Clark, S. E., & Wells, W. (2016). Estimating the reliability of eyewitness identifications from police lineups. Proceedings of the National Academy of Sciences, 113, 304–309. https://doi.org/10.1073/pnas.1516814112.