Decision Scientist Interview Questions UK: 8 Real Questions and What Interviewers Want

Before your Decision Scientist interview, it helps to see the questions from the interviewer's side. We have picked eight from our set of 12 and added the reason each one gets asked.

Eight Decision Scientist interview questions, and what the interviewer is thinking

1. How would you frame a business decision quantitatively, for example using a decision tree and expected value?

Why they ask: Decision science is about turning choices into structured, numerical comparisons.

2. When would you use optimisation, such as linear or integer programming, to support a decision, and how would you set one up?

Why they ask: Optimisation distinguishes decision scientists from analysts who only describe the past.

3. How is the role of a Decision Scientist different from that of a Data Scientist or Data Analyst, and why does it appeal to you?

Why they ask: The title is used differently across organisations, so they want to see you understand its focus.

4. What is uplift modelling, and how does it differ from a standard propensity model when deciding whom to target?

Why they ask: Targeting the right people is a classic decision science problem, and this tests causal thinking.

5. When would you use a multi-armed bandit instead of a conventional A/B test, and what are the trade-offs?

Why they ask: It tests judgement about experimentation approaches in business settings.

6. How would you use Monte Carlo simulation to support a business decision, and how do you present the results?

Why they ask: Simulation is a core tool for dealing with uncertainty in decisions.

7. A director has already decided on a course of action and asks you to produce analysis that supports it. What do you do?

Why they ask: They want to see integrity under pressure and tact in stakeholder handling.

8. Tell me about a time you turned a vague business problem into a quantified decision.

Why they ask: This is the heart of decision science, and they want evidence you can structure ambiguity.

A sample answer to question 1

How would you frame a business decision quantitatively, for example using a decision tree and expected value?

I'd start by clarifying the decision itself: who makes it, what the options are, and what they're trying to achieve. Then I'd map the choices, the uncertain events that follow each one, and the outcomes in money or another agreed unit. In a decision tree, each uncertain branch gets a probability, ideally from data, and I calculate the expected value of each option by weighting outcomes by their likelihood. For instance, launching a promotion might cost £200,000, with a sixty per cent chance of £500,000 extra margin and a forty per cent chance of £100,000. Expected value is then £340,000 less the cost. But expected value alone isn't enough. I'd also show the range of outcomes and the downside, as a board may care about avoiding a loss more than maximising the average, and run sensitivity analysis on the most uncertain probabilities to see which assumptions could flip the recommendation. If the data supports it, I'd go further with simulation.

Common mistake to avoid: Giving a single number without uncertainty; Using made-up probabilities without saying so

Want the rest?

Full model answers, key points and common mistakes for all 12 questions are on the Decision Scientist interview questions page. The first five are free, no sign-up needed.