What Is a Double-Blind Police Lineup?

A double-blind lineup is one where neither the witness nor the person administering the lineup knows which member is the suspect. The purpose is to prevent the administrator from influencing the witness's choice, whether deliberately or through unconscious cues. Research has shown that administrators who know which person is the suspect are more likely to emit behavioural signals, subtle shifts in body language, tone or pacing, that steer the witness towards that person. They are also more likely to give confirming feedback after an identification, inflating the witness's confidence. Double-blind administration removes both risks by ensuring the administrator has no information to leak. It is recommended by the National Research Council, the American Psychology-Law Society and a growing number of jurisdictions in the United States and beyond.

Key takeaways

• In a double-blind lineup, neither the witness nor the administrator knows who the suspect is.
• Administrators who know the suspect's identity can unconsciously influence the witness through behavioural cues.
• Post-identification feedback from a non-blind administrator can inflate witness confidence without improving accuracy.
• Double-blind procedures reduce mistaken identifications without significantly reducing correct ones.
• The method also protects against later claims that the identification was improperly influenced.

Why can a non-blind administrator affect the outcome?

The influence is usually unintentional. An administrator who knows the suspect is in position four might pause slightly longer at that photograph, lean forward, or change their tone of voice when the witness considers it. These cues are often too subtle for either party to notice consciously, but research has demonstrated that they are sufficient to shift witnesses' choices.

The effect is analogous to the "Clever Hans" phenomenon in psychology, where an experimenter's unconscious body language communicated the correct answer to a performing horse. In lineup administration, the "correct answer" the administrator holds in mind can leak through in similar ways. Studies have found that non-blind administrators are more likely than blind ones to emit cues that guide the witness towards the suspect, increasing identifications of the suspect regardless of whether that person is the actual perpetrator.

What is the post-identification feedback effect?

One of the most well-documented problems with non-blind administration is what happens after the witness makes a choice. In a landmark study, Wells and Bradfield had witnesses identify someone from a lineup that did not include the actual target, so every choice was wrong. Some witnesses were told "Good, you identified the suspect." Those who received this confirming feedback reported feeling more confident in their choice, remembering a better view, and having paid more attention than those who received no feedback.

The feedback did not change the accuracy of the identification, which was zero in all cases. It changed how the witness remembered the experience. By the time that witness reaches a courtroom, they may honestly testify that they were certain from the start, with no awareness that their certainty was manufactured by a comment made after the fact. This feedback effect, explored further in can a confident eyewitness be wrong?, is one of the strongest arguments for double-blind procedures.

How does double-blind administration work in practice?

The most common implementation uses a lineup administrator who is not involved in the investigation and does not know which position the suspect occupies. In some jurisdictions, computer-administered lineups achieve the same effect: the software presents photographs in a randomised order with standardised instructions, eliminating the human administrator entirely.

The procedure typically includes:

1. An administrator who does not know the suspect's identity presents the lineup.
2. A standard instruction tells the witness that the perpetrator may or may not be present, reducing pressure to choose.
3. No feedback is given during or after the identification, regardless of the witness's choice.
4. The witness's confidence is recorded in their own words at the moment of any identification, before any other interaction.
5. The entire procedure is recorded, preferably on video, so that it can be reviewed later.

These steps work together. Double-blind administration prevents cueing and feedback; the unbiased instruction prevents forced choice; confidence recording captures the witness's certainty before it can be inflated. Together, they produce an identification that is as close as possible to a pure test of the witness's memory.

Does it reduce correct identifications?

Research suggests that double-blind procedures primarily reduce mistaken identifications rather than correct ones. When the perpetrator is actually in the lineup, witnesses with a genuine memory of the face are still able to recognise them without help from the administrator. What changes is that witnesses with weaker memories, who might have been steered towards the suspect by an informed administrator, are less likely to make a choice at all.

This is the desired outcome. A correct identification that depends on administrator cueing is not reliable evidence. If the witness would not have made the same choice without guidance, the identification reflects the administrator's knowledge rather than the witness's memory.

What are the limitations?

Double-blind administration is not a complete solution. Research has shown that even when witnesses know the administrator is blind, behavioural feedback from the administrator can still influence decisions and confidence. For example, if the administrator reacts to the witness's choice with surprise or satisfaction, despite not knowing who the suspect is, this general social feedback can still affect the witness.

Additionally, double-blind procedures do not address other sources of error in identification:

• Poor lineup construction. If the suspect stands out because of inappropriate fillers, a blind administrator will not prevent a biased identification.
• Delay. A witness whose memory has faded over time, as covered in how does time affect eyewitness memory?, may still make a mistake regardless of how the lineup is administered.
• Cross-racial identification. The well-documented difficulty of recognising faces of a different race is not addressed by the administrator's knowledge.
• Prior exposure. If the witness has seen the suspect's photograph in media or in a previous lineup, they may identify the face from that exposure rather than from the event, a source-monitoring error.

Double-blind administration removes one important source of error. It works best as part of a package of safeguards rather than as a standalone fix.

Where has it been adopted?

Following the National Research Council's 2014 recommendations, a growing number of US states and law enforcement agencies have adopted double-blind or "blinded" lineup procedures. Some jurisdictions mandate it by law; others include it in departmental policy. The trend reflects a broader movement towards evidence-based identification practices, alongside reforms such as recording confidence statements and providing unbiased instructions.

Outside the United States, similar reforms have been adopted in England and Wales, where identification procedures already incorporate many of these safeguards through the Police and Criminal Evidence Act codes of practice.

How does Murder Book explore the question of influence?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. Its design reflects the reality that how evidence is gathered shapes what it means. The game is inspired by real police footage and psychological interviews, and it asks players to examine not just what people say, but how the circumstances of the conversation might have shaped their account.

That mirrors the lesson of double-blind research: a piece of evidence is only as independent as the process that produced it. Cases are designed for focused 5–10-minute sessions. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.

Sources

1. National Research Council (2014), Identifying the Culprit: Assessing Eyewitness Identification. National Academies Press
2. Wells and Bradfield (1998), "Good, you identified the suspect": feedback to eyewitnesses distorts their reports. Journal of Applied Psychology
3. Wixted and Wells (2017), The relationship between eyewitness confidence and identification accuracy: a new synthesis. Psychological Science in the Public Interest
4. Kovera and Evelo (2017), The case for double-blind lineup administration. Psychology, Public Policy, and Law
5. Innocence Project, Reevaluating Lineups: Why witnesses make mistakes and how to reduce the chance of a misidentification
6. Gungrounds, Murder Book fact sheet