SHL Numerical Test Sample Questions — Playable, With Solutions

The four families every SHL numerical test draws from

1. Percentage change — growth rates, discounts, year-on-year deltas. The classic trap: confusing a percentage point change with a percentage change. 2. Ratios and proportion — market shares, staffing splits, currency conversion. Trap: applying a ratio to the wrong base. 3. Rates — output per hour, cost per unit, defect rates. Trap: averaging rates that have different denominators. 4. Multi-step table reading — the answer needs two cells and an operation, not one lookup. Trap: the one-lookup answer is always among the options.

Every distractor in a well-built SHL item encodes a specific error. If you finish a practice question and can name the mistake behind each wrong option, you have extracted everything the item can teach.

What the real test measures (and what it doesn't)

Nothing beyond GCSE-level arithmetic — the test measures speed and accuracy of data interpretation under time pressure, not mathematical depth. That is why familiarity moves scores so much: the candidate who has already met the four families spends their 90 seconds calculating, while the first-timer spends 30 of them understanding the table. Scores are reported against a norm group, typically as percentiles — you are ranked against other candidates, not against a pass mark.

How to prepare effectively

Untimed accuracy first: work through each family until the method is automatic. Then timed runs at adaptive difficulty, because the real test climbs while you answer. Track your error types — most candidates lose the majority of their marks to two personal recurring traps, and killing those two is worth more than a hundred generic questions.

Persona Prep generates SHL-style numerical items with seven rotating data frames, full worked solutions including why each trap tempts, adaptive timed runs, and a stanine rank (1–9) against other candidates. Two aptitude exercises are free, no card required.