Presumptive questions, bait questions, and the other things that can be done to you in an interview
Most workplace investigation interviews are conducted properly, by people trying to find out what happened. But not all of them are, and the difference is visible in the questions themselves — if you know what you are listening for.
This is not about how to dodge anything. Every technique below has the same correct response: describe what you know, accurately, and decline to accept things you have no basis to accept. What changes is your ability to notice when a question is doing something other than asking.
Two families of interview
There are broadly two philosophies, and workplace investigators are trained in one or the other, or in some blend they may not have examined closely.
Information-gathering is the model behind PEACE, developed in the UK and adopted widely since. The interviewer is meant to be open-minded, to let you give a free account without interruption, and to pursue accurate information rather than a confession. Challenge exists in this model — you will be asked hard questions — but it comes after your account, and it is aimed at testing what you said rather than at breaking you.
Accusatory is the model behind the Reid technique and its descendants, developed in American law enforcement and exported into corporate investigations by people trained in it there. Its structure assumes guilt has already been established by some earlier assessment, and the interrogation that follows is designed to move a person from denial to admission. Denials are interrupted. Pressure is applied deliberately.
The research on the second family is not kind. The FBI's own Law Enforcement Bulletin describes a documented sequence behind wrongful convictions: investigators misclassify an innocent person as guilty, subject them to guilt-presumptive interrogation involving lying about evidence, and pressure them into a narrative — sometimes supplying the facts themselves. Information-gathering approaches show meaningfully lower rates of false confession.
There is a further finding worth carrying with you. People are worse at detecting deception when they can see the person than when they only hear them, and trained investigators perform at roughly chance. Anyone in that room who believes they can read you is wrong, and if they are working from that belief, their reading of your nerves is not evidence of anything.
You are unlikely to be told which model you are in. The questions will tell you.
The question types
Open invitations
"Tell me in your own words what happened." "What happened next?"
These are the good ones, and they are the ones people waste. An open question with no interruption afterwards is the most reliable thing an investigator can obtain, which is exactly why trained ones ask it and then go quiet. Take the space. Describe events in order. This is where a clear account does most of its work.
Probing and specific questions
"What time did you send that?" "Who else was in the room?"
Normal, necessary, and not a tactic. An investigator who never asks these has not done their job.
Leading questions
"You were annoyed by then, weren't you?"
The answer is contained in the question. Sometimes this is laziness rather than strategy — leading questions are easy to ask by accident when you already have a theory. Either way the record ends up containing the interviewer's words in your mouth.
The response is not to argue about the question. It is to answer in your own terms: "I wouldn't put it that way. What I remember is…"
Presumptive questions
"When you took the laptop home, who did you tell?"
Look at what happened there. The question appears to be about who you told. The thing it actually establishes is that you took the laptop home. Answer the surface question — "nobody" — and you have conceded the premise without ever being asked about it.
This is the most common tactic you will meet, and it is often not deliberate. An investigator who has already decided what happened will drift into presumptive phrasing naturally, because to them the premise is settled.
The response is to separate the premise from the question, out loud and without heat:
I didn't take the laptop home, so there's nobody I told. If you have something showing it left the building, I'd like to see it.
That is not evasion. You answered. You just declined to answer a question about an event that did not occur.
Bait questions
"Is there any reason your badge would show you entering the building at eleven that night? I'm not accusing you of anything — maybe you came back for something."
This one is worth understanding in detail, because it is the most effective thing in the accusatory toolkit and almost nobody on the receiving end recognises it.
Reid publishes the structure openly. A bait question is non-accusatory by design. It raises a plausible probability that some piece of evidence exists which implicates you. It supplies you with an innocent explanation in advance, so that taking the bait feels like agreeing with a helpful suggestion rather than changing your story. It is phrased as is there any reason why rather than as a claim. And its purpose is to get someone who has committed to a denial to reconsider it.
The evidence in a bait question does not have to exist. That is the mechanism. There may be no badge record. There may be no neighbour. The question is testing what you will do when you believe there might be.
Reid's own guidance is that a bait question should be used once per interview, because repeated use destroys its effect. That tells you something about what it is: not a request for information, but a single planned move.
There is published research finding that bait questions can act as a source of misinformation in their own right — that they don't merely pressure the person but can contaminate what that person subsequently believes they remember. This is the part that should concern you most. A bait question does not only risk pushing you into a false admission. It can quietly rewrite your recollection, and you will not feel it happen.
How to answer one. Answer the literal question, truthfully, and go no further:
No, there's no reason my badge would show that. If there's a record showing it, I'd like to see it, because it doesn't match my recollection.
What you must not do is adjust your account to accommodate evidence you have not seen. The instinct is powerful — they must have something, I'd better find a version that fits — and it is the single most damaging thing you can do in an investigation interview. An account that shifts each time it meets resistance is the pattern investigators are most alert to, and if the evidence turns out not to exist, you have moved for nothing and there is no moving back.
If you genuinely do not know, say so. "I don't remember being there at eleven. I'm not going to guess."
Minimization
"Look, anyone would have done the same thing. It's not a big deal. Just tell me and we can all move on."
Minimization downplays the moral seriousness of the conduct while saying nothing about the actual consequences. It is designed to make admission feel cheap. It is not cheap — the person offering you the easy exit is very rarely the person who decides what happens next, and the summary that reaches the decision-maker will not contain the words not a big deal.
Warmth in an interview is not itself a tactic; rapport is a legitimate and effective technique, and people give fuller accounts when they are not frightened. The tell is not friendliness. It is friendliness attached to a specific invitation to concede something.
Maximization
"We already have everything we need. This is going to a criminal referral unless you help yourself here."
The mirror image: overstating the strength of the evidence and the severity of what is coming. If the evidence really were conclusive, there would be no reason to tell you so in order to obtain something from you.
You do not have to challenge the claim. You can simply not move: "I've told you what happened. If there's something specific you want me to respond to, I'll respond to it."
The same question, three times, in different words
Legitimate and expected. Free accounts are accurate but thin, and a good interviewer circles back to get detail. It is also how inconsistency surfaces.
The response is the same each time, which is easy if you are describing and hard if you are constructing. If you notice you are trying to remember what you said rather than what happened, stop and slow down.
Forced-choice questions
"Did you do it because you were under pressure, or because you didn't think anyone would notice?"
Both options concede the act. There is a third answer and you are allowed to give it: "Neither. I didn't do it."
Challenge is not a tactic
This matters, and it cuts the other way.
A competent investigator is expected to put the difficult material to you — the document that contradicts your account, the colleague who remembers it differently, the gap in your timeline. That is not an accusatory technique. It is procedural fairness, and it is required. If nobody ever put the hard point to you and it appeared in the findings anyway, that is a failed process.
So the arrival of an uncomfortable question is usually a sign that things are being done properly, not improperly. Do not read every hard question as an attack. The distinction is whether the question is asking you about something, or working on you to produce something.
If you think you are in an accusatory interview
Some signals, taken together rather than individually: you are interrupted every time you deny something; you are told the evidence is overwhelming but never shown any of it; you are offered a version of events and invited to agree with it; the same admission is pursued from several directions regardless of what you answer; someone tells you the meeting can end as soon as you say the thing.
What you can do about it, in the room:
Ask for the scope, in plain terms. "Can you tell me specifically what I'm here to respond to?" You should have been told this at the start.
Ask to see documents you are being asked about. "You've said the record shows that. Can I see it?" A properly run investigation will normally put the material in front of you, because your response to it is the point.
Correct the summary at the end, specifically and out loud. That summary is what goes into the report.
Ask for a break. Then take it.
Ask to stop. You can say that you would like to continue at another time, or after taking advice. This may have consequences in an employment context and you should understand that before you say it — which is precisely why the next section exists.
Where this stops being about interview technique
A workplace investigation is not a police interview. There is no caution, no right to silence in the criminal sense, and in most jurisdictions declining to participate can itself carry employment consequences. Whether that is fair, and what it means for you specifically, is a legal question about your jurisdiction and your contract, and this article cannot answer it.
Speak to an employment lawyer before the meeting if any of these apply:
- There is a potential criminal dimension
- A regulator is or may become involved
- Dismissal is clearly on the table
- You have been offered a settlement or severance agreement
- The employer's own lawyers are running the investigation
- A professional registration could be affected
In those situations the interview is not only an internal process, and an hour of advice beforehand is worth a great deal more than an hour afterwards.
The short version
Learn to hear the shape of a question. A presumptive question smuggles a conclusion into the premise — answer the premise, not the surface. A bait question raises evidence that may not exist — answer it literally and never reshape your account around something you have not seen. Minimization makes conceding feel cheap; maximization makes resistance feel futile; both are aimed at moving you rather than informing anyone.
And underneath all of it, one thing does not change. Describe what you know. Say when you do not know. Do not accept a premise you have no basis to accept, and do not adjust your account to fit a document nobody has shown you.
That answer works in a good interview and a bad one, which is why it is the only one worth learning.
The First Account includes a module on question types with the specific forms you are likely to meet in matters like yours, and a printable note sheet for recording the meeting as it happens. Built by someone who ran these interviews for thirty years.