Case File 310 · Investigation Psychology

Can Knowing a Suspect's Confession Bias Other Evidence?

Yes, knowing that a suspect has confessed can bias how other evidence in the case is evaluated. Researchers call this forensic confirmation bias: the tendency for a confession or other expectation of

By Murder Book Editorial TeamPublished by Gungrounds, the studio making Murder BookPublished 3 October 2026Updated 3 October 2026

Yes, knowing that a suspect has confessed can bias how other evidence in the case is evaluated. Researchers call this forensic confirmation bias: the tendency for a confession or other expectation of guilt to distort the interpretation of forensic evidence, eyewitness testimony and even expert analysis. Saul Kassin, Itiel Dror and Jeff Kukucka coined the term in a 2013 paper in the Journal of Applied Research in Memory and Cognition, drawing on studies showing that confessions do not just add to the evidence against a suspect but can actively corrupt it. Their review found that when forensic examiners, polygraph operators and other analysts are told a suspect has confessed, they are more likely to interpret ambiguous evidence as confirming guilt. The effect has been demonstrated in fingerprint analysis, handwriting comparison, polygraph interpretation and eyewitness identification.

Key takeaways

A confession can bias how forensic examiners, witnesses and other analysts interpret evidence, a phenomenon researchers call forensic confirmation bias.
Dror and Charlton (2006) showed that fingerprint experts changed their own prior conclusions when given misleading contextual information about a case.
Kassin, Bogart and Kerner's 2012 review of 241 DNA exoneration cases found that multiple errors appeared far more often in false confession cases, and the confession usually came first.
The bias means that evidence which appears to corroborate a confession may not be truly independent.
Blinding procedures, where forensic analysts are not told about confessions or other case information, can reduce the risk.

How does a confession affect forensic evidence?

One of the most striking demonstrations came from Dror and Charlton, who published their findings in the Journal of Forensic Identification in 2006. They recruited five experienced fingerprint examiners with a combined 85 years of experience. Each examiner was shown pairs of fingerprints they had previously analysed in real criminal cases, but this time the researchers added false contextual information: they were told the prints were from the case that had led to the wrongful arrest of Brandon Mayfield in the Madrid bombing investigation.

Four of the five examiners changed their conclusions. Three now said the prints did not match, one said a definite decision could not be made, and only one maintained the original match determination. These were the same prints the examiners had previously analysed correctly. The only difference was the contextual information they were given.

Dror and colleagues expanded this work in later studies, showing that emotional context and case-related information could influence judgements even when the physical evidence remained identical. The effect was strongest when the evidence was ambiguous: clear matches and clear non-matches were less susceptible to contextual influence.

What happens when a confession comes first in a case?

Kassin, Bogart and Kerner addressed this question in their 2012 study, published in Psychological Science. They reviewed the trial records of the first 241 people exonerated through DNA evidence by the Innocence Project. Of those 241 cases, 59 involved false confessions.

The findings were striking:

Multiple errors appeared in about 69% of false confession cases, compared with fewer than half of cases without false confessions.
In two thirds of the false confession cases, the confession came first, followed by other errors including invalid forensic science, mistaken eyewitness identifications and unreliable informant testimony.
Cases with false confessions were more likely to involve forensic errors (roughly two thirds) and multiple types of additional errors (about one third had at least two other error types).

Kassin's interpretation was that these patterns suggest the errors were not independent. The confession may have set the direction for the rest of the investigation, and subsequent evidence was gathered and interpreted through the lens of assumed guilt. As he put it, "there may well be the appearance of corroboration, but it is false evidence that was corrupted by the confession, not independent at all."

How does the bias affect specific types of evidence?

Research has documented confession-driven bias across several evidence types:

Fingerprint analysis. The Dror and Charlton (2006) study showed that experienced fingerprint examiners changed their determinations when given misleading contextual information. The Innocence Project has documented cases where forensic errors and false confessions appeared together.

Handwriting comparison. Kukucka and Kassin published a study showing that participants who knew a defendant had confessed were more likely to conclude that handwriting samples from the defendant and an unknown writer were authored by the same person, compared with a control group that received no confession information.

Polygraph interpretation. Kassin, Dror and Kukucka's 2013 review noted that polygraph examiners who were given contextual information about a suspect's expected guilt were more likely to interpret ambiguous results as indicating deception.

Eyewitness testimony. The same 2013 review documented that confession information can also affect eyewitness confidence and identification decisions, as witnesses who learn about a confession may become more certain of identifications they were previously unsure about.

Why is this a problem if the confession is true?

Even in cases where the confession is genuine, the bias creates a structural weakness in the evidence. If forensic results, witness identifications and expert opinions are all influenced by knowledge of the confession, they cannot serve as independent checks on its accuracy. Why is corroboration important in an investigation? explains why independent verification matters.

The problem is especially acute with false confessions. When a false confession contaminates other evidence, the entire case can appear solid even though it rests on a single unreliable foundation. Each piece of evidence appears to confirm the others, but all of them trace back to the same source of error.

What can be done to reduce confession bias in forensic evidence?

Researchers have proposed several safeguards:

Linear sequential unmasking. Dror and colleagues have advocated for a procedure where forensic examiners first analyse evidence without any contextual information about the case, then receive additional context only after forming their initial conclusions. This prevents expectations from shaping the initial analysis.
Blinding procedures. Keeping forensic analysts unaware of confessions, suspect identities and other potentially biasing information is the most direct protection. Several forensic science organisations have begun recommending blinding as standard practice.
Documentation of initial conclusions. When examiners record their findings before receiving contextual information, any later changes can be identified and evaluated.
Separation of roles. Having different people handle the investigation and the forensic analysis reduces the chance that case theories will influence evidence interpretation.

How does Murder Book connect to confession bias?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. Its development screenshots show an evidence board where statements are linked to physical details, and a technique menu that includes Confrontation, Bluffing and Baiting alongside Build Trust. The game asks players to weigh whether a suspect's statements are consistent with the evidence, but the evidence itself must be examined carefully.

The research on forensic confirmation bias suggests that in real investigations, the relationship between confession and evidence runs both ways: evidence shapes the interrogation, but the interrogation can also shape how evidence is read. Murder Book's structure, where players examine both statements and evidence, mirrors this challenge. A player who takes a suspect's admission at face value may miss contradictions that a more careful reading would reveal. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.