| | |

Find The Odd One Out

Every so often the exam stops asking you to find the right answer and asks you to find the wrong one. Which of these is NOT a lawful basis. All of the following are AI harms EXCEPT. The task has flipped, and the odd one out, the single option that does not belong, is now the thing you are hunting. Read it on autopilot and you will pick a perfectly correct statement, feel good and lose the mark.

These negative questions are not harder than the rest. They punish a habit. You spend the whole paper trained to recognise the true, well-phrased option, so when three of the four options are true and one is false, your instinct points you at exactly the wrong kind of answer.

Why the odd one out is easy to miss

The design is quietly clever. In an ordinary question one option is right and three are wrong, so recognising correctness works. In a negative question the ratio inverts: three options are correct statements and one is not, and the correct statements are written to look attractive. Your eye lands on a true, tidy answer and your hand moves before the word NOT has registered. It is the plain-looking option you skimmed past because it did not seem impressive.

There is a second reason these bite. A normal question rewards partial knowledge; recognise the right answer and you score even if the rest are a blur. A negative question rewards complete knowledge, because you can only name the misfit if you know the full set it is missing from. That is deliberate. The examiner is checking that you hold the whole list, not a comfortable corner of it. The failure, then, is procedural rather than a gap in what you know, which is good news, because a procedure can be fixed.

Reading for the odd one out

Two small habits turn these from traps into easy marks.

Circle the flip word first

Before you read a single option, find and mark the word that flips the task: NOT, EXCEPT, FALSE, LEAST. These are close cousins of the qualifier words that quietly decide other answers, and they deserve the same suspicion. Say the stem back to yourself in its real form: three of these are lawful bases, one is not, find the one that is not. Naming the task stops your autopilot from answering the question you expected instead of the one on the page.

Watch for the double flip too. A stem that asks which option is NOT an exception to a rule has turned the task twice, and the safe move is to untangle it in plain words before you look down: I am looking for the one that does follow the rule. Slow is fast here.

Score each option true or false

Now go option by option and label each one true or false against the rule, not good or bad. Quality is a distraction here; a beautifully worded option that is true is still a wrong answer. Take a stem asking which of four items is NOT one of the data protection principles, offering purpose limitation, accuracy, profitability and storage limitation. Scored coldly, three are principles and profitability is not, so profitability is the answer, however sensible it sounds as a business goal. You did not rank anything; you knew the set.

Three should come back true. The one that comes back false is your answer, and you can pick it with confidence rather than a hopeful shrug. If two come back false, you have misread the rule or the flip word. Go back to the stem, not deeper into the options.

When LEAST asks for a ranking, not an odd one out

One variant needs a different tool. LEAST and MOST questions are not asking for a true-or-false split; they are asking you to rank. “Which is the LEAST appropriate safeguard” can offer four options that are all defensible to some degree, and your job is to place them on a line and pick the end. Treat it as a true-or-false split and you will hunt for a false option that is not there and stall.

So sort the question type before you start. A NOT or EXCEPT question wants the false misfit. A LEAST or MOST question wants the extreme of a ranking. Same family, different move and knowing which one you are holding saves the half minute you would otherwise spend looking for the wrong thing.

Whether this week has you naming an AI characteristic or a data protection principle, the underlying skill is the same: read what the question actually asks before you reach for what you know. If you want a structured set to practise the negative variety on, the drills in the study resources at 22Academy are built for exactly this.

Similar Posts