Case File 074 · Investigation and Psychology

What Makes a Statement Unreliable?

The many reasons a statement can be wrong, from stress and suggestion to source confusion and interview pressure, and why unreliable does not mean dishonest.

By Murder Book Editorial TeamPublished 16 September 2026Updated 16 September 2026

A statement is unreliable when there is good reason to doubt that it accurately reflects what happened. Lying is only one possible cause. An honest person's statement can be unreliable because of how they perceived the event, such as under high stress, or because of what happened to their memory afterwards: forgetting, conversations with others, news coverage or misleading questions. People can also confuse where a memory came from, remembering something they were told as something they saw. Finally, the way a statement is obtained matters: leading questions, repeated questioning and pressure can all change what people say. Reliability and honesty are separate questions. A sincere witness can be wrong, and a confident one is not automatically accurate.

Key takeaways

Unreliable does not mean dishonest. Many causes of unreliability are honest errors.
High stress was linked to poorer eyewitness identification and recall in a meta-analysis of dozens of tests.
Source monitoring errors happen when people misattribute where a memory came from.
Leading questions, discussions with other witnesses and repeated questioning can change what people report.
Confidence is informative only under good conditions, so it cannot be treated as a simple sign of accuracy.

Is an unreliable statement the same as a lie?

No. A lie is a deliberate attempt to mislead. An unreliable statement is any statement that may not match reality, whatever the reason.

The distinction matters because people are not good at telling the two apart by watching someone speak. In a meta-analysis of 206 documents and 24,483 judges, Bond and DePaulo (2006) found that people correctly classified lies and truths only 54% of the time on average. A separate meta-analysis of 158 possible cues to deception by DePaulo and colleagues (2003) found that many behaviours showed no discernible link, or only weak links, to deceit.

An honest witness who has absorbed a false detail will not look like a liar, because they are not lying. The difference between the two is explored in What Is the Difference Between a Lie and a False Memory?

What can make a statement unreliable?

Problems can enter at every stage, from the event itself to the moment the statement is written down.

Stage	Factor	What can go wrong
During the event	Stress	High stress can impair identification and recall of details
During the event	Viewing conditions and attention	A person may not have seen or noticed what they later describe
After the event	Forgetting	Details fade over time
After the event	Misinformation	Outside information can be absorbed into memory
After the event	Discussion with others	Witnesses can adopt details from each other
When remembering	Source confusion	A detail is remembered but attributed to the wrong source
When questioned	Leading or repeated questions	The question can shape the answer, and repetition can produce inconsistent responses
When questioned	Pressure and vulnerability	Certain tactics and personal characteristics raise the risk of false statements
At any stage	Deliberate deception	The person knowingly says something false

The sections below look at several of these in more detail.

How does stress affect a statement?

It might seem that a frightening event would be remembered especially well. The research suggests otherwise. A meta-analysis by Deffenbacher, Bornstein, Penrod and McGorty (2004) combined 27 tests of stress on eyewitness identification and 36 tests on recall of crime-related details. It found considerable support for the idea that high stress impairs both. The effects were notably larger in studies using a staged crime than in studies that induced stress in other ways.

A stressed witness may also feel very sure about what they remember. Research on emotional "flashbulb" memories found that emotion predicted later belief in a memory's accuracy, but not its consistency over time. How Stress Affects Testimony covers this topic in more depth.

What is source monitoring, and why does it matter?

Source monitoring is the process of working out where a memory came from. Did you see it, hear it from someone, read it, dream it or imagine it?

In an influential 1993 review, Johnson, Hashtroudi and Lindsay described a framework for these judgements and linked it to eyewitness testimony, misattributed familiarity and the "incorporation of fiction into fact". They argued that source judgements are based on the qualities of a remembered experience and are "subject to error and disruption".

In practice, a witness might remember a detail accurately but place it in the wrong context. A face seen at a shop may be remembered as a face seen at the scene. A detail heard in a conversation or a news report may later feel like something the witness saw personally. The memory is real. Its source is wrong.

How can suggestion change a statement?

Memory can absorb information that arrives after an event. Elizabeth Loftus reviewed three decades of research on this misinformation effect in 2005. She described how exposure to misleading information can impair memory for the past, noted that warnings help only in limited circumstances, and reported that some groups are more susceptible than others.

Questions are one route for misinformation. In Loftus and Palmer's classic 1974 study, changing a single verb in a question about a car crash changed people's speed estimates. For more on that study, see Why Memory Is Not a Perfect Recording.

Other witnesses are another route. In a study by Gabbert, Memon and Allan (2003), 71% of witnesses who discussed an event with a co-witness later recalled items they had picked up from the discussion. This is why the U.S. National Institute of Justice's eyewitness guide advises separating witnesses, avoiding leading questions and encouraging witnesses to avoid media coverage.

Can the interview itself make a statement unreliable?

Yes. The UK Ministry of Justice guidance Achieving Best Evidence in Criminal Proceedings warns that repeating the same focused questions over time could lead to unreliable or inconsistent responses in some witnesses. For early discussions with a witness, it says that any questions should, as far as possible, be open-ended or specific-closed rather than forced-choice, leading or multiple. Open-Ended Versus Leading Questions explains the difference.

One of the most serious forms of interview-induced unreliability is a false confession. A white paper by Saul Kassin and colleagues (2010), published in Law and Human Behavior, identified suspect characteristics that raise the risk, including adolescence, intellectual disability, mental illness and certain personality traits. It also identified interrogation factors, including excessive interrogation time, presentations of false evidence and minimisation. The authors strongly recommended mandatory electronic recording of interrogations. See Why Innocent People Sometimes Confess for more.

Does confidence mean a statement is reliable?

Not on its own. Confidence can be informative, but only under the right conditions.

Wixted and Wells (2017) reviewed research showing that an adult's high-confidence identification can be remarkably accurate when the memory is tested under "pristine" conditions: an initial, uncontaminated test, a fair line-up, no influence from the person running it and an immediate record of confidence. When those conditions are missing, even a high-confidence identification can be seriously compromised.

This means the history of a statement matters. Confidence expressed at the first, careful interview carries different weight from confidence that grew after weeks of conversations and coverage.

How do investigators judge whether a statement is reliable?

There is no single test. At a high level, reliability is judged by looking at the statement's history and checking it against other evidence:

When and how was the account first given? Early accounts obtained with open questions are less likely to have been shaped by outside information.
What might have influenced it? Conversations with other witnesses, media exposure and earlier questioning are all possible sources of contamination.
What conditions did the person experience? Stress, distance, lighting and attention all affect what could have been perceived.
Is it supported independently? Records, recordings and physical evidence can confirm or contradict a statement without relying on anyone's memory.

Comparing one person's statement with another's is useful but has its own traps, which are covered in How Investigators Compare Multiple Accounts.

How does Murder Book treat unreliable statements?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. It asks players to examine unreliable stories, narrative gaps and contradictions. The press kit names Fabrication as one behaviour among several, alongside Avoidance, Narrative Gaps and Emotional Leakage. Development screenshots also show a casebook entry for Stress Response.

That range is the point this article makes. A statement can be wrong because someone is lying, but it can also be wrong because they are frightened, misremembering or repeating something they heard. In Murder Book, behaviour is a reason to look closer, not proof, and each case is designed for a focused 5–10-minute session.

Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.