Numerical reasoning (SHL Interactive)
Adaptive within the combined Verify G+ session · Roughly one minute per question; the session blends domains into a single timeframe
What it tests. Quantitative literacy, selective data extraction, multi-step computation under cognitive load, and mapping findings onto a non-traditional interface.
Worked example. An interactive table shows quarterly revenue, operating expenses and tax rates for three European subsidiaries of a UK-listed conglomerate in euros. You determine the absolute difference in net profit margin between Subsidiary A and Subsidiary C in Q3 of Year 2, convert it into GBP using a currency drop-down (£1 = EUR 1.18), then drag a slider to the exact figure on a visual scale.
Common traps. Misreading chart units (thousands versus millions) and making a scale error on the slider; confusing percentage-point change with relative growth rate; over-calculating to four decimals when the interface only needs an approximate plot.
How to handle it. Spend the first 4 seconds reading only the axis labels and legend before the prompt, master percentage and compounding maths so it is instant, and use an external physical calculator and clear scratch paper rather than the on-screen calculator.
Verbal reasoning (SHL Interactive)
Adaptive within the combined Verify G+ session · Roughly one minute per question
What it tests. High-level comprehension, objective textual analysis and differentiating explicit fact from subjective inference without importing personal bias.
Worked example. A 200-word internal compliance brief on MiFID II unbundling of research fees. You drag three operational conclusions into boxes labelled 'Strictly Substantiated by the Text', 'Contradicted by the Text' or 'Cannot be Determined Without External Information'.
Common traps. Injecting outside commercial knowledge the text does not state; mistaking a probabilistic phrase ('highly likely') for certainty; missing qualifiers like 'mostly', 'traditionally', 'occasionally' or 'exclusively'.
How to handle it. Read the statements or sorting boxes before the passage to prime keyword hunting, flag absolute versus probabilistic terms, and treat the text as an isolated script: if it does not state a fact, that fact does not exist for the test.
Logical reasoning (inductive and deductive)
Adaptive within the combined Verify G+ session · Roughly one minute per question
What it tests. Fluid intelligence: solving novel abstract problems, handling multi-variable logical systems under time pressure and deducing flawless conclusions from limited rules.
Worked example. A 3x3 matrix evolving across rows and columns under two simultaneous rules - a clockwise five-colour rotation and an alternating edge-subtraction rule (each step drops one external edge). You construct the missing bottom-right shape by choosing its base structure, interior colour from a palette and orientation via a rotation wheel.
Common traps. Fixating on one variable while a separate rule governs shading or edge count; over-investing three minutes on a single grid; rushing the drag-and-drop or rotation wheel and misaligning the response.
How to handle it. Deconstruct each matrix into independent dimensions (shape, orientation, border, fill, position), verify a cracked rule both across rows and down columns, and if no pattern crystallises within about 45 seconds, make an educated guess and move on.
Situational Judgement Test (SJT)
Scenario-based · Untimed in practice but to be completed in one block
What it tests. Behavioural alignment with Jefferies' values: collaboration, client focus, integrity and personal accountability. It screens out toxic individualism, blame-shifting and a lack of proactive ownership.
Worked example. An Associate finds a material formatting error in a bound pitchbook already sent to an MD travelling to a client meeting, and tells you to stay quiet. You rank options ranging from calling the MD directly, to staying silent while preparing a corrected version, to collaboratively drafting a correction email with the Associate, to immediately reporting the Associate to the group head.
Common traps. The 'hero' archetype (solving big structural problems alone, creating hidden risk); the corporate sycophant (blindly following an unethical senior instruction); the whistleblower escalation (running to HR over a minor, easily fixable peer issue before resolving it at your level).
How to handle it. Internalise the firm's values, take extreme but communicative ownership (draft a solution and present it for sign-off rather than waiting passively), and choose the calm, methodical, transparent path every time.
Occupational Personality Questionnaire (OPQ32)
100+ items · Untimed
What it tests. Preferred behavioural style across Relationships with People, Thinking Style, and Feelings and Emotions; it builds a profile to see whether your tendencies match a successful banking professional (stress tolerance, attention to detail, achievement orientation, data-driven decisions).
Worked example. A triplet such as 'I find it easy to persuade others to adopt my strategic viewpoint' / 'I thoroughly enjoy analysing complex, data-heavy financial models' / 'I consistently plan and organise my tasks well in advance', where you must choose a Most and a Least.
Common traps. Inconsistency that the backend lie-scale flags; refusing to pick a 'Least' among three positive traits; a flat, neutral profile that reads as lacking drive.
How to handle it. Answer as a consistent professional persona (as you are executing a critical project, not relaxing on a Sunday), lean into analytical, resilient, deadline-comfortable traits, and never contradict an earlier answer across blocks.