Case File 278 · Investigation Psychology

Can Interrogation Pressure Change What Someone Believes Happened?

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

Yes, interrogation pressure can change what a suspect believes happened. In what researchers call a coerced-internalised false confession, an innocent person not only confesses but comes to believe they actually committed the crime. This was demonstrated in a landmark 1996 experiment by Kassin and Kiechel, in which innocent participants were accused of causing a computer crash. Under certain conditions, 69% signed a confession, 28% came to believe they had caused the crash, and 9% constructed false memories of doing so. These findings show that memory is not a fixed recording; it can be reshaped by sustained pressure, suggestion and the presentation of false evidence.

Key takeaways

In Kassin and Kiechel's 1996 experiment, 28% of innocent participants came to believe they had caused a computer malfunction they did not cause, and 9% created false memories of the act.
Coerced-internalised false confessions occur when an innocent person actually comes to believe they committed the offence in question.
The presence of false evidence, such as a supposed eyewitness, significantly increased both compliance and internalisation rates in laboratory studies.
Memory is reconstructive, not reproductive, meaning it can be altered by post-event information, suggestion and emotional pressure.
Real cases exist in which suspects became convinced of their own guilt during interrogation, only for evidence to later prove their innocence.

What is a coerced-internalised false confession?

Kassin and Wrightsman proposed three categories of false confession in 1985: voluntary, coerced-compliant and coerced-internalised. The coerced-compliant type occurs when a suspect confesses to end the pressure of questioning while knowing they are innocent. The coerced-internalised type goes further: the suspect actually comes to believe they did what they are accused of.

This is the most psychologically troubling category because the person's own memory changes. They may search their mind for fragments that seem to fit the accusation, reinterpret ambiguous moments as evidence of guilt, and gradually construct a narrative in which they are the perpetrator. Once this process takes hold, the person may express genuine distress and guilt about what they believe they did, making their confession appear especially convincing to investigators and jurors.

How did Kassin and Kiechel demonstrate this in the laboratory?

The 1996 study used an elegant experimental design. Participants were asked to type letters read aloud by a confederate. They were warned that pressing the Alt key would cause the computer to crash and all data to be lost. No participant actually pressed the key, but after the computer appeared to crash, the experimenter accused each participant of having done so.

The study manipulated two variables: the speed of the task, which affected how uncertain participants were about their own actions, and whether a confederate claimed to have seen the participant press the key. The results were striking:

Overall, 69% of participants signed a written confession.
28% internalised the false belief, later telling another person they had caused the crash.
9% confabulated specific details about how they had pressed the key, even though they had not.

When both the pace was fast and a false witness was present, compliance reached 100%, internalisation reached 65%, and confabulation reached 35%. The study showed that even in a low-stakes laboratory setting with no coercion beyond accusation and suggestion, innocent people could come to doubt and then revise their own memories.

Why does this happen?

The answer lies in how memory works. Memory is not a video recording that plays back exactly what happened. It is a reconstructive process, and each time a person recalls an event, their recollection can be influenced by new information, emotional state and social context.

In an interrogation, several factors push in the same direction. The investigator presents a confident account of what happened. The suspect may be tired, anxious or confused. If false evidence is introduced, such as a claim that forensic results link the suspect to the crime, the suspect may begin to question their own certainty. This is especially likely when the suspect has limited or unclear memory of the relevant period, which can occur due to intoxication, sleep deprivation or the simple passage of time.

Shaw and Porter published a 2015 study in Psychological Science that demonstrated how easily detailed false memories of committing a crime could be implanted in participants through suggestive interviewing techniques over just three sessions. By the end of the study, 70% of participants in the false-memory condition had come to recall detailed memories of events that never happened, including events involving police contact.

How does this relate to real criminal cases?

Cases of apparent internalised false confessions have been documented in the real criminal justice system. In these cases, suspects confessed with apparent sincerity, sometimes providing details that seemed to corroborate the crime, only for later evidence such as DNA testing or the identification of the actual perpetrator to establish their innocence.

The mechanism by which crime-scene details enter a false confession is important. Research on confession contamination shows that investigators can inadvertently or deliberately feed details to a suspect during questioning. When those details appear in the confession, they may seem to confirm its reliability, even though the suspect learned them during the interrogation itself.

This creates a dangerous cycle. A suspect who has been told details of the crime, and who is beginning to doubt their own memory, may incorporate those details into a narrative that feels like a genuine recollection. The confession then appears rich in detail, making it more persuasive to investigators, prosecutors and juries.

Can someone be protected from this process?

Several safeguards can reduce the risk. Recording entire interrogations allows reviewers to identify the point at which crime-scene details first appeared, whether they came from the suspect or the investigator. This is one of the most effective ways to assess whether a confession reflects genuine knowledge or contamination.

Time limits on interrogation length can also help. Research consistently shows that longer interrogations increase the risk of both compliant and internalised false confessions. An information-gathering approach, which focuses on open-ended questions rather than accusation, reduces the suggestive pressure that can reshape memory.

None of these safeguards is perfect. But understanding that interrogation can genuinely alter what a person believes happened is a necessary first step toward reducing the risk.

What does Murder Book ask players to consider?

Murder Book: The Art of Interrogation is built around the challenge of reading unreliable stories. Development screenshots show an in-game casebook that includes entries such as Fabrication and Emotional Fading, behaviours that may signal a problem in a narrative but do not prove why the narrative is wrong. The research on internalised false confessions adds another layer: sometimes the person telling the story genuinely believes it, even when it is not true.

In Murder Book, players examine behavioural changes, narrative gaps and contradictions. The game does not present behaviour as proof, and the research above explains why that restraint matters. A suspect who has come to believe a false account may show none of the signs that people associate with lying, precisely because, from their perspective, they are not lying. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.