Stress Interview

What a stress interview is

A stress interview is a job interview in which the interviewer deliberately makes the candidate uncomfortable in order to see how they behave under that discomfort: by interrupting, by holding a long silence after an answer, by repeating a question in a form that casts doubt on the candidate’s honesty, or by objecting to everything they say.

The condition that defines it is that the discomfort is intended and manufactured, not a product of the question itself. A difficult question about the substance of the work is not a stress interview, however hard it is on the candidate, because its difficulty lies in the content. In a stress interview the difficulty lies in how the interviewer treats the candidate.

The useful questions about it are what it actually measures, what it costs, and what can take its place when the real aim is to see how someone handles pressure.

What it measures is not what it is meant to measure

It is presented as a way to show resilience under pressure. What it actually creates is a situation that differs from pressure at work in four ways:

• The source of pressure is a person who holds the decision about the candidate. At work, pressure can come from a task and a deadline without anyone choosing to be difficult.
• The candidate has no context. They do not know the customs of the place, who is asking or why. An employee at work knows all of that and acts on it, which the candidate cannot.
• No resources and no colleagues. The ways pressure is handled at work (asking for help, checking, postponing) are not available in the room.
• Leaving is costly. The candidate cannot end the situation without paying for it; an employee can object and take the matter higher.

The last of these has an effect that turns the purpose upside down. How much the candidate appears to tolerate moves with how much they need the job, not with their ability to handle pressure. Someone with no alternative puts up with more; someone with alternatives walks away sooner. The score captures the candidate’s position in the market and attributes it to a trait in them.

A stress interview cannot be squared with what makes interviews comparable

In a structured interview the questions, their order and the rating scale are fixed, and an unstructured interview produces an impossible comparison because it compares two questions rather than two people.

The stress interview is built on the opposite by definition. The pressure is applied according to what the candidate shows in the room: this one is interrupted where that one was not, a silence follows this one’s answer and not that one’s. The situation differs from candidate to candidate, and the scores are compared as though it were the same.

That is why a stress interview cannot be called structured. If what is done to each candidate is fixed completely, it has become a standardised simulation, which is a different arrangement described below. If it is left to the interviewer, the basis for putting scores side by side has gone.

On top of that, the interviewer in this method rates a situation they created. They are opponent and judge in the same session, and someone who has been hard on a candidate and then rates them cannot separate what they saw from what they caused.

A worked example: who walks away

The visible cost of the method falls on the organisation’s reputation; the cost that can be counted falls on who it loses. What matters is that the loss does not fall evenly across candidates.

Take 30 candidates who reached the final stage, 10 of whom hold another live offer and 20 of whom do not. Assume that an interview run this way leads 60% of those with alternatives and 10% of the others to withdraw afterwards:

• Withdrawals among those with alternatives: 10 times 0.6, which is 6.
• Withdrawals among the others: 20 times 0.1, which is 2.
• So 8 of the 30 withdraw, and 22 remain.

The number that matters is not the eight. It is that 6 of the eight, 75% of those lost, are the candidates with alternatives. Only 4 of the ten candidates with alternatives remain.

So the method does sort people, but not by the trait it is meant to measure: it keeps those who cannot afford to leave and removes those who can. Candidates with alternatives can be the ones organisations compete for, so a tool built to select the strongest removes first the people with the most options. The rates in this example are assumed, to show how the calculation is done and which way the effect runs; they are not measured values. In the sources we reviewed we found no figures for the method’s effect on candidate withdrawal or on an organisation’s reputation.

There is also a cost that never appears in the pipeline: what the people who walked away say about the organisation, which is part of the wider candidate experience and of how the organisation is seen as an employer. Its effect falls on people who never applied, so it cannot be measured, because they are not in the data.

Why it stays in use anyway

We found no source, among those we reviewed, that puts a figure on the predictive validity or the reliability of stress interviews, or ranks them among selection tools. The method leaves the interviewer with a strong impression, and that alone explains a good deal. Three reasons are offered in its defence, and each is answered by what has already been said:

• It shows the candidate as they really are. What it shows is how they behave in a situation that will not recur, in front of someone who holds the decision about them, with no context and no resources. What appeared is real; the problem is that it says nothing about the role.
• The work here is hard, so we need to know who can take it. If the difficulty lies in the tasks, it belongs in an exercise built from the work itself. If it lies in how people treat one another, the interview does not fix it; it advertises it.
• This is how it has always been done here. That is not a reason. Its value is that it shows the method is passed on by habit rather than by a decision anyone reviews.

A fourth reason goes unspoken and weighs most: it needs no preparation. A structured interview needs written questions and a described rating scale; a stress interview can be run with nothing prepared. What is presented as rigour in selection can be the absence of preparation.

Practices that share the name but belong elsewhere

One question sorts every case: does the discomfort lie in the content of what the candidate was asked to do, or in how the person asking treats them? The practices below are filed under the label without belonging to it, and each sits on the content side of that line.

• A hard question on the substance of the work. A complex problem is put to the candidate to work through. The difficulty is in the content, the answer is scored against a key, and nobody is trying to make the candidate uncomfortable.
• Serious probing questions. Asking the candidate for detail on an answer, or to spell out what they summarised, is part of a structured interview run well, provided it does not turn into a new question that counts towards the score.
• An honest preview of a demanding role. Telling a candidate plainly, before they accept, about the real pressure the role carries belongs to a realistic job preview, whose material is information about what the job is, not discomfort produced in the room.

What happens on the other side of the table

The method hands the interviewer work they have not been trained for: producing discomfort in an amount they cannot gauge, then reading the reaction in a way that tells passing irritation from something meaningful. Neither skill is acquired by practice alone, and no written key governs either.

As a result, the intensity of the pressure varies with the interviewer and their day. An interviewer used to the method pushes harder than one running it for the first time, so a candidate who met the first is measured with a different ruler from one who met the second. Nothing in the session records this difference, so the gap between the two scores is read as a gap between the candidates.

The candidate, meanwhile, reads what happened as a description of what awaits them after hiring, and may be reading it correctly: an organisation that uses the method in the room may treat people the same way inside it. Those who accept have accepted what they saw, and those who walk away have read a signal nobody can call false.

Whatever the interviewer writes down about the candidate in such a session is personal data about them, and the rules on personal data apply to it. We found no provision, in the sources we reviewed, that sets out rules for the manner in which interviews are conducted in the Saudi private sector or the limits of that manner, so the points above describe what the tool measures and what follows from it; they are not a legal ruling and not legal advice.

When pressure is part of the role

Some roles are made of dealing with people who are angry or objecting: customer service at the point of complaint, collections, field work in emergencies. Measuring behaviour in those situations is genuinely needed, and the tool for it is not a stress interview but a standardised simulation with four features:

• The situation is the same for every candidate, from a written script followed by whoever plays the other party, so performance varies and the situation does not.
• The person producing the pressure is not the person rating, which separates making the situation from judging it.
• A written key is prepared before it is used, describing what counts as weak, acceptable and strong behaviour. How such a key is built is set out under work sample test.
• The candidate is told the nature of the exercise beforehand, because the aim is to see how they behave in the situation, not to surprise them with the fact that they are being tested.

The difference is that a simulation reproduces work the candidate will do, while a stress interview reproduces treatment that will not recur after hiring. What the stress interview measures is how well someone tolerates a person who holds power over them, and that is not a requirement of any role.

What to do if an organisation uses it today

What is needed is not a sudden ban but a question put to every place the method is used: what trait were we trying to measure here, and are candidates being measured on it in the same way? If the trait cannot be named, the justification falls away by itself; if it can, it may turn out to have a more precise tool.

The first practical step is to separate the person who rates from the person who creates the situation; the second is to write the rating scale before the session. Neither step abolishes anything, and together they turn the arrangement into something that can be defended, which is the basis of the structured interview. An organisation that takes both steps may find that manufactured pressure drops out by itself, because it has no place left on a written form.