Case File 076 · Investigation and Psychology

Why People Avoid Answering Questions Directly

The innocent and non-innocent reasons people sidestep questions, and why avoidance alone cannot reveal guilt.

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

People avoid answering questions directly for many reasons, and many of them have nothing to do with guilt. Communication research shows that people tend to give vague but truthful answers when every direct answer would cost them something, such as hurting someone, embarrassing themselves or starting an argument. People also sidestep questions because they don't know or can't remember the answer, misunderstand what is being asked, feel overwhelmed, want to protect their privacy or someone else, or fear they won't be believed. Avoidance can also be a way to hide wrongdoing without telling an outright lie. The behaviour alone cannot tell you which explanation applies. Research on deception finds that behavioural cues are weak, and investigators who treat evasion as a sign of guilt risk misjudging innocent people. Avoidance is a reason to ask more carefully, not a conclusion.

Key takeaways

Research on equivocation finds that people give indirect but true answers when all direct answers have negative consequences.
Innocent reasons for avoidance include not knowing, not remembering, stress, embarrassment, privacy, protecting others and fear of being disbelieved.
Avoidance can also serve to conceal wrongdoing, so it is worth noticing, but it is not evidence of guilt on its own.
Listeners often fail to notice when a question has been dodged, which is one reason accurate records of questions and answers matter.
Research suggests innocent suspects often assume their innocence is obvious, so openness and reticence are both poor guides to guilt.

What counts as avoiding a question?

Avoidance covers a range of responses. A person might give a partial answer, answer a slightly different question, change the subject, answer a question with a question, give a very general reply or say nothing at all.

Researchers who study political interviews call much of this equivocation. Bull has studied noncommittal replies in televised political interviews. In a 2008 article, he concluded that the central propositions of a theory of equivocation developed by Janet Bavelas and colleagues were well supported.

Everyday conversation is full of milder versions. Someone asked whether they liked a friend's cooking may talk about how much effort went into the meal. That is avoidance, but hardly a sign of guilt.

Why do honest people avoid direct answers?

Every direct answer has a cost

One influential explanation comes from Bavelas, Black, Chovil and Mullett (1990). They proposed that the nature of the communicative situation, rather than the individual, largely determines whether people answer clearly or equivocally. In some situations, every direct message, whether true or false, has negative consequences. The authors called these avoidance-avoidance conflicts.

In four experiments using hypothetical scenarios, people in these conflicts made their messages equivocal but true, while people in control conditions did not. Additional analyses showed that the equivocal messages were not lies of omission. In other words, indirectness was a way of staying truthful while avoiding harm.

Bull (2008) argued that equivocation can also be explained in terms of face and face management, meaning how people manage their own and others' public image. A direct answer that would damage someone's reputation, relationships or dignity creates a strong pull towards vagueness.

They don't know, or can't remember

A vague answer can simply reflect a vague memory. Memory fades, and people sometimes cannot bring to mind something they know. Psychologist Daniel Schacter grouped transience, absent-mindedness and blocking as the forgetting failures among his "seven sins of memory". A person who hedges may be trying not to guess. Why Memory Is Not a Perfect Recording explains this in more detail.

Stress and trauma

High stress impairs recall of details, according to a meta-analysis by Deffenbacher and colleagues (2004). The UK Ministry of Justice guidance Achieving Best Evidence in Criminal Proceedings also notes that some witnesses who have experienced a traumatic event may find an interview "too much", especially when emotional matters are discussed. It suggests breaks and moving to more neutral topics before returning to the matter being investigated. A person who deflects a painful question may be managing distress, not hiding facts.

Fear of not being believed

Being questioned is stressful even for people with nothing to hide. A 2024 review by Farley in the Journal of Nonverbal Behavior points out that truth tellers can be anxious for fear of being wrongly labelled a liar. That fear can make people cautious, hesitant or guarded.

The question itself is the problem

Some questions are hard to answer directly because they are unclear, contain several questions at once or assume something the person disagrees with. For early discussions with a witness, the Achieving Best Evidence guidance says questions should, as far as possible, be open-ended or specific-closed rather than forced-choice, leading or multiple. A person who pushes back on a loaded question is not necessarily evading it. Open-Ended Versus Leading Questions covers why question design matters.

Privacy, loyalty and legal caution

People also avoid questions to protect things that are unrelated to the matter at hand: a private relationship, a health condition, a friend's secret or an embarrassing but lawful activity. In formal settings, a person may also be cautious because of legal advice or uncertainty about their rights. Rules on this vary by jurisdiction, and this article does not offer legal advice.

What are the non-innocent explanations?

Avoidance can also serve deception. At a general level, a person might avoid a question to:

conceal their own involvement in wrongdoing;
avoid making a specific false statement that could later be checked;
protect someone else who has done something wrong;
avoid saying anything that implicates them while they decide how to respond.

Research on deception gives some support for a link between concealment and reduced openness. A meta-analysis of 158 cues by DePaulo and colleagues (2003) found that, in some ways, liars were less forthcoming than truth tellers. But the same study found that many behaviours showed no discernible links, or only weak links, to deceit.

Can avoidance tell you whether someone is guilty?

No. The same behaviour fits too many explanations.

Possible reason for avoidance	Innocent?	What could help clarify it
Every direct answer would hurt someone	Usually	Understanding the relationships and context
Genuine forgetting or uncertainty	Yes	Whether the detail would reasonably have been noticed or kept
Stress or trauma	Yes	The person's circumstances and wellbeing
Fear of being disbelieved	Yes	How the questioning has been framed
Unclear, loaded or multiple question	Yes	Asking a simpler, open question
Privacy about an unrelated matter	Yes	Whether the topic is actually relevant
Concealing involvement	No	Independent evidence, not demeanour

Several research findings add to the case for caution. Bond and DePaulo's (2006) meta-analysis found that people judge lies and truths correctly only 54% of the time on average. Saul Kassin's 2005 review of confession research described how investigators can presume innocent suspects guilty in pre-interrogation interviews, and how innocent suspects' denials can still lead to highly confrontational interrogations. Once an investigator believes someone is hiding something, avoidance can look like confirmation. See How Confirmation Bias Can Affect an Investigation.

Openness is not a reliable sign of innocence either. In a mock-crime experiment by Kassin and Norwick (2004), participants who were truly innocent were significantly more likely than guilty participants to waive their Miranda rights, apparently believing their innocence would protect them. Innocent people may talk freely because they assume the truth is obvious, and that confidence can carry its own risks. Why Innocent People Sometimes Confess looks at those risks.

Do people notice when a question is dodged?

Often, they don't. In four studies, Rogers and Norton (2011) found that listeners can fail to detect a dodge when a speaker answers a similar but different question. The authors suggested that listeners usually focus on social evaluation, such as whether they like the speaker, rather than on whether the question was actually answered. Detection improved when the question remained visible during the answer.

This cuts both ways. Real evasions can pass unnoticed, and a clear record of exactly what was asked and exactly what was said is more useful than an impression of whether someone "seemed evasive".

How do investigators respond to avoidance?

At a high level, good practice treats avoidance as information to explore rather than as an admission. That means noting precisely which question was not answered, checking that the question was clear, returning to the topic with open questions, considering the innocent explanations above and testing the matter against independent evidence. The difference between exploring and pressuring is discussed in Interview Versus Interrogation: What Is the Difference?

How does Murder Book handle avoidance?

Murder Book: The Art of Interrogation is an upcoming narrative interrogation game from Gungrounds, an independent studio in Zagreb, Croatia. Avoidance is one of the interrogation behaviours named in its press kit, and development screenshots show an in-game casebook that also lists Stonewalling and Non-Response. The same casebook lists Stress Response, and the press kit names Emotional Leakage.

Those entries sit well alongside the research in this article. A suspect who won't answer might be hiding something, but they might also be frightened, confused or protecting someone for reasons unrelated to the case. In Murder Book, behaviour is a reason to look closer, not proof, and 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.