Deductive logic (SwitchChallenge / scales sx)
Adaptive levels of increasing complexity · Exactly 6 minutes
What it tests. Working memory, spatial-lexical processing speed and deductive logic under extreme time pressure.
Worked example. With operator 2413, the first slot tells you which top-row position moves to the bottom's first slot; an operator starting with 4 means the far-right symbol moves to the far left. Stacked operators require deducing the net effect.
Common traps. Eye-balling all four shapes at once (cognitive fatigue by level 5), missing that an inversion followed by an inversion can cancel out, and getting stuck 90 seconds on one puzzle, which tanks the volume score.
How to handle it. Isolate one asset: find which operator routes Position 1 to its new home to eliminate two or three options instantly. Clear single-layer puzzles in 8-12 seconds, allowing up to 25 for the complex variations.
Situational judgement (ChatAssess)
Typically 15 to 20 distinct scenarios · About 20 to 30 minutes total (untimed overall but dynamically responsive)
What it tests. Situational judgement, commercial alignment, communication style and relationship management against Schroders' risk appetite and client-first ethos.
Worked example. A colleague flags a mistake in a report; the strong response involves immediate transparent validation of the error followed by an actionable resolution plan, rather than a quiet fix.
Common traps. The people-pleaser paradox (delaying a hard decision to keep everyone happy), over-reliance on escalation, and using the same informal tone with a Senior Portfolio Manager as with a fellow intern.
How to handle it. Adopt the Schroders mindset: systematic risk management, analytical precision, integrity and collaboration. Prioritise the client outcome where it does not breach compliance, and reject passive 'wait and see' options.
Personality (ADEPT-15 Workstyles)
About 100 statements · No time limit, typically 15 to 20 minutes
What it tests. Fifteen personality traits across leadership style, task management, emotional resilience and interpersonal dynamics, mapped to cultural fit.
Worked example. Choosing between 'I enjoy taking the lead during ambiguous projects' and 'I pay meticulous attention to minor data points' builds a behavioural blueprint the engine cross-checks for consistency.
Common traps. Extreme alternation to look 'even' (which fails the consistency check), the lone-wolf archetype, and strongly agreeing with unrealistic perfection ('I have never made an analytical error under pressure').
How to handle it. Align with Excellence, Innovation, Integrity, Teamwork and Passion: structure, conscientiousness, data-driven openness and calm under volatility. Be methodical but swift; overthinking a pairing introduces contradictions.
Numerical reasoning (scales numerical, finance) - some streams
36 questions · 12 minutes (about 20 seconds per question)
What it tests. Data location across tabs, rapid estimation, percentages and chart analysis under time pressure.
Worked example. For a statement on 'Operating Margin 2024', isolate the metric, navigate straight to the correct tab and estimate rather than computing to the decimal.
Common traps. Searching all six tabs for one metric, and blind end-of-test guessing into the negative-marking penalty.
How to handle it. Read the statement first, isolate the core metric, jump to the matching tab and estimate. Do not expect to finish all 36; protect accuracy.
Verbal reasoning (scales verbal) - some streams
49 questions · 12 minutes
What it tests. High-speed scanning, inference versus explicit fact and comprehension of dense corporate text.
Worked example. For a statement on 'MiFID II disclosure thresholds', flag the keyword, jump to the relevant paragraph via the tab headers and check for an explicit semantic match or contradiction.
Common traps. Reading tabs sequentially start to finish, and importing outside market knowledge the text does not state.
How to handle it. Read the question stem first, flag the unique keyword and jump straight to the relevant paragraph; never wrestle one statement past about 20 seconds.