Why Do Investigators Ask the Same Question in Different Ways?

Investigators ask the same question in different ways to test whether the subject's account stays consistent or shifts depending on how the question is framed. The reasoning is that a truthful person

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

Investigators ask the same question in different ways to test whether the subject's account stays consistent or shifts depending on how the question is framed. The reasoning is that a truthful person drawing on real memory should produce a broadly consistent account regardless of how the question is worded, while a person maintaining a fabricated story may struggle when the same topic is approached from an unexpected angle. However, research complicates this assumption. Studies by Granhag and Strömwall found that liars can be just as consistent as truth tellers when they anticipate the questions, because they rehearse and repeat a fixed story. The technique works best when the different phrasings include genuinely unexpected questions that the subject has not prepared for.

Key takeaways

Investigators rephrase questions to probe whether the subject's account holds up from multiple angles.
The "repeat versus reconstruct" hypothesis suggests that liars deliberately repeat a rehearsed story, while truth tellers naturally reconstruct from memory, sometimes introducing minor variations.
Consistency alone does not reliably distinguish liars from truth tellers, especially for anticipated questions.
Unanticipated questions, which the subject has not prepared for, are more effective at revealing differences between truthful and fabricated accounts.
Research found that unanticipated questions correctly classified up to 80% of liars and truth tellers, compared with near-chance accuracy for anticipated questions.

Why do investigators rephrase questions?

The basic logic is straightforward. If someone is telling the truth, they are describing events they actually experienced. Asking about those events from a different angle, such as asking about the timeline first and then asking about what they saw, should produce compatible answers because both draw on the same genuine memory. If someone has invented their account, each rephrased question forces them to generate or retrieve a response that is consistent with everything they have already said, which requires ongoing mental effort.

This technique is sometimes called "cross-questioning." Investigators use it not as a single decisive test but as a way to generate material that can be checked for internal consistency and compared against external evidence. For more on the broader process of testing accounts, see how investigators find inconsistencies.

What does research say about consistency and deception?

The relationship between consistency and deception is less clear-cut than many practitioners assume. Granhag and Strömwall proposed the "repeat versus reconstruct" hypothesis, which draws a distinction between how liars and truth tellers approach repeated questioning.

Liars tend to repeat. A person who has fabricated a story knows that changing details could expose them. They therefore try to recall exactly what they said before and say it again. This deliberate repetition can produce a highly consistent account across multiple interviews.

Truth tellers tend to reconstruct. A person drawing on genuine memory does not memorise their previous answers. Each time they are asked, they reconstruct the event from memory, which is an imperfect process. As a result, their accounts may contain minor variations: different word choices, slightly different ordering of events, or small details that appear in one telling but not another.

The practical implication is counterintuitive. Perfect consistency can actually be a feature of rehearsed accounts, while natural inconsistency can be a feature of genuine memory. This does not mean that inconsistency proves truthfulness, only that consistency alone is not a reliable indicator either way. For more on why truthful accounts can be inconsistent, memory research explains the mechanism.

Do unanticipated questions make the technique more effective?

This is where the research becomes more promising. Vrij, Leal, Granhag and colleagues published a study in 2009 titled "Outsmarting the liars," which tested whether unanticipated questions could improve lie detection accuracy.

In their experiment, pairs of liars and pairs of truth tellers were interviewed individually about a shared experience (having lunch together at a restaurant). The interviewer asked both anticipated questions, which the subjects could reasonably have prepared for, and unanticipated questions, such as asking each person separately to draw the layout of the restaurant.

The results were clear. For anticipated questions, the correspondence between paired subjects' answers did not differ significantly between liars and truth tellers. For unanticipated questions, particularly the drawing task, truth-telling pairs produced much more consistent answers than lying pairs. Classification accuracy reached up to 80% based on correspondence in unanticipated question responses.

The explanation is that truth tellers can answer unanticipated questions by accessing their actual memory of the event, while liars, who have typically only rehearsed answers to predictable questions, must improvise. When two liars improvise independently, their answers are less likely to match.

What types of unexpected questions work?

Research has identified several categories of questions that tend to catch prepared liars off guard.

Spatial questions. Asking about the physical layout of a place, such as "Draw me the room" or "Where was the door relative to the window?" These are hard to prepare for because they require detailed spatial memory that a person who was not actually there is unlikely to have.

Temporal questions. Asking about the sequence and timing of events in detail, such as "What happened between arriving and ordering?" A fabricated account often includes key events but lacks the mundane connective tissue between them.

Sensory questions. Asking about what the subject saw, heard or smelled during the event. These details are naturally encoded in genuine memories but are rarely part of a rehearsed cover story.

Reverse-order recall. Asking the subject to tell their story backwards. This is particularly demanding for liars because their story is typically organised as a forward narrative, and reversing it requires restructuring a fabricated sequence in real time. For more on this approach, see why investigators ask someone to tell a story backwards.

What are the limits of rephrasing questions?

Several important caveats apply.

Repetition can backfire. If an investigator simply asks the same question again without changing the angle, the subject may interpret it as disbelief and become defensive or anxious, regardless of whether they are lying. This can make an honest person appear evasive and a liar appear consistent, the opposite of what the technique intends.

Practice reduces the effect. Liars who have had time to rehearse extensively, or who have been through previous interviews, may be able to handle rephrased questions competently. The advantage of unanticipated questions comes precisely from their being unexpected, which means an interviewer who uses the same approach every time may lose the element of surprise.

Memory limitations apply to everyone. A truthful person who is stressed, tired or interviewed long after the event may produce inconsistent answers not because they are lying but because their memory has degraded. Research on how sleep deprivation affects memory shows that fatigue alone can create apparent inconsistencies.

How does Murder Book explore this?

In Murder Book: The Art of Interrogation, an upcoming narrative interrogation game from Gungrounds, players question suspects across focused 5-10-minute cases. Development screenshots show an in-game casebook that tracks behaviours such as Fabrication, Avoidance and Narrative Gaps. Players can approach the same topic from different angles, watching for whether the suspect's story holds together or shifts under pressure.

The game reflects the research finding that the way a question is asked matters as much as what is asked. A suspect who answers confidently to a direct question may falter when the same ground is covered from an unexpected direction. Murder Book is in development for PC (Steam) and mobile, with release currently planned for the end of Q2 2027.