A syllogism gives you a few premises — statements to be taken as true — and asks which conclusion follows necessarily. The catch is that one option is always 'no logical answer can be determined', and it is sometimes the correct one. You must accept the premises exactly as given and reason strictly, rejecting any conclusion that merely sounds plausible or happens to match what you know about the real world.
Why airlines test it
This is deductive reasoning under discipline: it measures whether you can follow logic to its actual conclusion instead of jumping to what feels right. Airlines care because aviation runs on rules, limitations and procedures that must be applied precisely — a candidate who bends logic to fit an assumption is a candidate who might one day bend a limitation. Syllogisms also test whether you can recognise when the available information simply does not support a conclusion, which is its own kind of good judgement.
In the cockpit
Regulations, minima and standard operating procedures are formal logic in disguise: if conditions are met, the approach is legal; if they are not, it is not — no matter how much you would like to land. Diagnosing a fault from indications, deciding whether the requirements for a procedure are truly satisfied, knowing when you do not yet have enough information to commit — all of it demands strict deduction and the discipline to say 'the data does not support that.' Wishful reasoning is how good crews talk themselves into bad approaches.
For example: the destination reports 600 m visibility, your approach requires a minimum of 550 m, and your alternate is forecast well above minima — so the approach is legal and you may continue. Change that one report to 500 m and the very same strict logic now says you may not begin the approach, no matter how much fuel, daylight or determination you have. The premises decide the conclusion; your preferences do not get a vote.
How to practise
The Syllogisms game below rewards exactly this discipline. Practise treating every premise literally, checking whether a conclusion truly must follow, and getting comfortable choosing 'no logical answer' when it genuinely does not.