Turning the doubts you already have into challenges that score.
If your scepticism marks keep coming back low, it is almost never because you missed the problem. It is because you wrote the problem as a question. On this skill, a question earns nothing — the marks go to whoever took the same doubt and turned it into a stated challenge. This page walks you through that one move, worked twice on a made-up company, so that by the end you can take the questions already sitting in your answers and rewrite them as challenges that score. If you have been spotting the right things and still losing the marks, this is almost certainly why.
When something in an exhibit looks off, the doubt usually arrives on its own. You notice the number leaning on an assumption nobody stated, or the glowing figure from the person who wants the plan approved. That noticing feels like scepticism — you have seen what others walked past, and you feel sharp.
But noticing is not the skill. A question points at the doubt and then hands the work back to the reader: someone should look into this. It never says what is wrong, or what it means, or what to do about it. So it earns nothing — not because the doubt was wrong, but because it stopped at the doubt. The skill is the stance you take on what you noticed.
Every strong sceptical point runs the same three steps, in the same order.
<aside> 🧭
Spot — name the thing that does not add up. This is the cheap step; the doubt turns up uninvited, which is why nearly every answer has it.
Position — turn the doubt into a challenge, not a question. Not is this right? but this is wrong, and here is why — naming the assumption, the gap, the contradiction or the wrong comparison that undermines it. This is where the marks start.
So what — say how far the evidence can be trusted, and the one thing to do before relying on it. This is the step almost nobody reaches, and it is what turns a good spot into a complete answer.
</aside>
Everything below runs on one made-up company. Kelby Foods is invented, built only for this example — a chilled ready-meals maker weighing up a six-million-pound automated packing line. The paper recommending the line was written by the operations director, who proposed it in the first place. Two of its claims are worth challenging.
The paper says the line pays for itself in three years — six million pounds of kit, two million pounds saved a year, so three years (six divided by two). It is the kind of number that ends a discussion, because it looks like arithmetic. But the three years only holds if Kelby's biggest customer — worth around forty per cent of revenue — renews its contract next year. And the same paper notes, elsewhere, that the contract is out to competitive tender right now.
<aside> 🔴
Weak — a question: "Is the three-year payback realistic? Will the supermarket actually renew the contract?"
</aside>
That is aimed at exactly the right place, and it earns nothing. It has not said anything — it has handed the reader a job.
<aside> ✅
Strong — a position: "The three-year payback assumes the supermarket contract renews next year — yet that customer is around forty per cent of revenue and the paper itself says the contract is out to competitive tender. The payback case treats a contested renewal as settled, so its central assumption is unproven. The board should place little reliance on the three-year figure as it stands, and either wait for the tender outcome or model the payback on losing the contract before committing the six million pounds."
</aside>