Signaller multi-tasking test: what to expect and tips

What the three tasks are

From consistent applicant descriptions:

• Arithmetic. Applicants describe it as simple or basic mental arithmetic.
• Pattern or shape monitoring. A stream of shapes, watched for a repeat.
• Steering. Usually described as keeping a car from crashing, or as something like a simple arcade game — an object you move left and right to get through gaps.

Taken individually, all three are trivial. A ten-year-old can do any one of them. That is deliberate, and it is the key to the whole thing.

What it is actually measuring

Because each task is easy, the score cannot be about ability at any of them. What varies between candidates is how performance holds up when attention has to be divided — and specifically, how it fails.

There are two failure patterns. Some people share the loss: all three tasks get a bit worse, roughly evenly. Others abandon one: they lock on to the steering, get near-perfect at it, and stop answering sums entirely. Network Rail does not publish how it scores the exercise, but tests of this kind usually add up performance on each task — and a task you stop doing scores nothing at all, however well the others go. So the second pattern is the riskier one, and it is also the one that says least about you as a signaller.

A signaller with three things happening — a failed track circuit, a driver on the phone, and a train approaching a junction — cannot afford to solve the first two beautifully and forget the third. The job is not to be excellent at one thing under pressure. It is to keep everything at an acceptable standard and to know which one is closest to going wrong. That is the most plausible reason for setting an exercise like this, and it is why our version scores the three tasks separately instead of giving you one number.

The habit that actually improves the score

Most people, told to practise multi-tasking, try to get faster. That is the wrong target, and it usually makes things worse: speeding up narrows attention, which is exactly the mechanism that makes you drop a task.

The thing to practise is deliberate switching. Rather than doing whichever task is currently shouting loudest, cycle: check the lane, glance at the shape, answer the sum, back to the lane. It feels slower and inefficient. It scores better, because the cost of a missed switch is far higher than the cost of a slightly slower one.

Two specific things follow from that:

1. Do not chase perfection on the steering. It is the most viscerally engaging of the three and it will eat all your attention if you let it. Clipping one obstacle is a single error. Losing thirty seconds of arithmetic to avoid clipping it is a great many.
2. Answer the sum, even if you are not certain. An unanswered sum is almost certainly counted against you, and it also means you are still holding it in mind while the next one arrives. Commit and move on.

Where to put your attention when it goes wrong

It will go wrong. Everyone has a moment where the shapes have moved on twice, there is a sum on screen you have not read, and you are about to hit an obstacle. What you do in that second is most of your score.

The instinct is to try to catch up on everything. Do not. Deal with whichever task has a deadline — the obstacle, because it is arriving whether you look at it or not — and then take the tasks that are merely waiting. The sum will still be there. The shape stream will have cost you one item, which is one mark. Trying to recover all three at once usually loses all three.

This is a genuinely transferable skill and it is the same judgement the job simulation is testing in slower motion: deal with the thing that is about to happen, then deal with the things that are merely outstanding.

You can try the multi-tasking exercise free to see which task you drop first. It sits in the same battery as the concentration exercise, which tests the opposite thing — holding one standard steady for a long time — and the assessment explained covers the order everything comes in.

Practical setup for the day

• Use a proper keyboard and a mouse or trackpad you are used to. The exercise is sat at home, so this is entirely within your control. Fumbling for an arrow key is a mark.
• Sit somewhere you will not be interrupted. An interruption during a divided-attention test is not a minor annoyance, it is a fourth task.
• Do not sit it tired. Divided attention degrades with fatigue faster than almost any other cognitive function. Doing it at 11pm after a shift is throwing away marks you could have had for free.
• Do a warm-up run first. The first thirty seconds of anything unfamiliar are your worst. If the real thing is the first time you have ever done it, you spend those thirty seconds on the clock.

How our version differs from the real one

We do not know exactly what Network Rail's looks like, because they do not publish it, and anyone who tells you they have the real thing is either guessing or has copied something they should not have. Ours reproduces the structure applicants describe — the same three tasks, running concurrently — rather than claiming to be a copy, and adds difficulty that climbs through the run so it keeps stretching you.

What that gets you is the thing that transfers: the experience of having three demands and not enough attention, and the chance to find out which one you drop before it counts. That is not a small part of the preparation. For most people it is the whole of it.

The one thing not to do is abandon a task entirely under load, which is the commonest way marks are lost here and number five in nine ways people lose the signaller assessment.