Updated April 2026 · 9 min read · Saville Assessment · Single-domain tests
| Provider | Saville Assessment |
|---|---|
| Test family | Saville Standalone (Verify Aptitude Tests) |
| Format | Single-domain tests, fixed questions, ~20-25 minutes each |
| Available domains | Numerical Analysis, Verbal Analysis, Diagrammatic Analysis, plus Comprehension variants and Mechanical / Spatial |
| Used by | British Council, Lloyds Banking Group, Centrica, Severn Trent |
Saville Standalone tests are the single-domain cousins of the Swift family. Where Swift bundles three reasoning modes into one adaptive test, Standalone tests dedicate a full session to numerical, verbal, or diagrammatic reasoning. Used when an employer wants a more thorough evaluation of one specific cognitive domain.
~24 questions in 24 minutes. Tables, charts, financial reports — multi-step calculations involving percentages, ratios, weighted averages, and currency conversion. Calculator provided. Heavy on real-world business context.
Sample Numerical Analysis question
A retail group reports the following quarterly revenue (in £ millions): Q1 = 2.40, Q2 = 2.76, Q3 = 2.54, Q4 = 2.97. By what percentage did Q4 revenue exceed the average of the first three quarters? (Round to one decimal place.)
Approach: Average of Q1-Q3 = (2.40 + 2.76 + 2.54) / 3 = 7.70 / 3 = 2.5667. Percentage increase = (2.97 − 2.5667) / 2.5667 × 100 = 15.7%. Answer: 15.7%.
Watch-out: Standalone Numerical favours candidates who set up the calculation cleanly before reaching for the calculator. Misreading "exceed the average" as "exceed Q3" is the most common error — it costs you the question and roughly 60 seconds of recovery time.
~30 questions in 24 minutes. True / False / Cannot Say on business and policy passages. The Cannot Say distinctions are subtle — Saville's verbal tests are notable for the precision required between "False" (passage contradicts) and "Cannot Say" (passage doesn't address).
Sample Verbal Analysis question
Passage: "The 2024 employee engagement survey reported that 67% of staff felt their managers communicated openly. Among managers themselves, however, only 41% rated their own communication skills as 'effective' or 'highly effective'. The company is now rolling out a manager training programme focused on feedback delivery."
Statement: Managers undervalue their own communication compared to how staff perceive it.
Approach: The passage gives two figures (staff: 67% felt managers communicated openly; managers: 41% rated themselves effective). These measure different things — staff's perception of openness vs managers' self-rating of effectiveness — so the comparison the statement makes can't be drawn directly from the passage. Answer: Cannot Say.
Trap: it's tempting to mark "True" because 67% > 41% feels like managers undervaluing themselves. But the two percentages measure different constructs. Saville rewards strict passage adherence — only what is stated counts.
~24 questions in 24 minutes. Process flow diagrams with input symbols transformed by operators — your task is to determine the output or identify the operator. Used for analytical and engineering roles.
Sample Diagrammatic Analysis question
Three operators are defined on shape sequences: α reverses the order, β rotates each shape 90° clockwise, γ swaps the first and last shape. The input is [▲, ■, ●, ◆]. The output of (α then γ) is shown as [◆, ●, ■, ▲]. Which single operator produces the same output from the original input?
Approach: Apply α to [▲, ■, ●, ◆] → [◆, ●, ■, ▲]. That's already the final output, which means γ (the swap-first-last) doesn't change it further (because the first and last are already swapped). So the single operator that produces [◆, ●, ■, ▲] from [▲, ■, ●, ◆] is α (reverse).
Pattern to learn: when a two-step operator chain produces a "clean reversal" output, one of the steps is almost always a no-op on that particular input. Test α alone before testing composite sequences.
Numerical Analysis (24 min, 24 questions = 60 seconds per question). Read the chart legend and axis labels before the question text — most candidates lose 15 seconds on each item by re-reading the chart. Pre-write the formula in shorthand ("Q4 vs avg Q1-3") before touching the calculator; this stops you from anchoring on the wrong reference point. If a question would take more than 90 seconds, flag it and return — Standalone allows revisiting, which Swift does not.
Verbal Analysis (24 min, 30 questions = 48 seconds per question). Read the passage first, then each statement. For every statement, ask in order: (1) does the passage directly state this? If yes → True. (2) Does the passage directly contradict this? If yes → False. (3) If neither → Cannot Say. The most common error is upgrading "Cannot Say" to "True" because the statement sounds reasonable. Saville's scoring punishes false positives more than blanks — when in doubt, go with Cannot Say.
Diagrammatic Analysis (24 min, 24 questions = 60 seconds per question). Sketch the input and output side by side. Most operators fall into three families: reorder (reverse, swap, rotate position), transform (rotate shape, mirror, recolour), or filter (remove specific elements). Work backwards from the output — what minimum change converts input to output? Then identify which single operator or pair achieves it.
Common across all three: the on-screen timer is reliable — trust it rather than your wristwatch. Saville Standalone does not penalise unanswered questions but does penalise random guessing because of negative marking on some role norm groups. Check with the employer whether negative marking applies before deciding to blank questions vs. guess.
Operational-level versions for customer service, admin, and apprenticeship roles. Same format as Analysis variants but with simpler calculations and shorter passages.
For technical and engineering roles. Mechanical tests cover levers, pulleys, gears, springs. Spatial tests cover 2D-to-3D rotation, mental folding, and spatial pattern recognition.
| Feature | Saville Standalone | Saville Swift |
|---|---|---|
| Format | Fixed (non-adaptive) | Adaptive |
| Duration | 24 min per domain | ~25 min total (combined) |
| Skip allowed | Yes, return to skipped items | No, sequential |
| Best for | Deep evaluation of one domain | Fast triangulation across domains |
Sten scores and percentiles relative to a norm group (graduate, professional, etc.). Cutoffs vary by employer and role. Saville's reports include sub-scale breakdowns — for numerical, you might see sub-scores for "data interpretation" vs "calculation accuracy" vs "speed."
Domain-specific deep work. Unlike Swift where you balance three modes, Standalone tests reward focused preparation in one domain. If you have Numerical Analysis, drill numerical hard for two weeks.
Use the skip-and-return feature. Standalone tests typically allow flagging items and returning later. Don't dwell — flag anything taking over 90 seconds and come back.
Sub-scale awareness. Saville reports include sub-scale scores. If your weak area is "calculation accuracy" rather than "data interpretation," you can target practice on the right component.
Numerical reasoning on Standalone tests is almost always table-based: two or three small tables of financial, sales, or operational data, followed by a question that requires a multi-step calculation and a unit conversion.
Q. A retail chain sells three product lines. Units sold last quarter were 660 (Line A), 1,140 (Line B) and 310 (Line C). Average selling price was £1.00, £1.00 and £1.00 respectively. Total revenue to the nearest £ was:
A) £1,780 B) £1,950 C) £2,048 D) £2,110
A. Sum the units: 660 + 1,140 + 310 = 2,110. Answer: D.
The actual Standalone question adds distractors: prices in pence rather than pounds, mixed currencies, unit ambiguity (per pack vs per item). Candidates who rush the unit check pick C or B despite nailing the arithmetic.
Standard Standalone Verify numerical assessments give 18 questions in 18 minutes — about 60 seconds per question. That sounds generous but each question has 3–5 numbers to read, a calculation (often multi-step), and a unit conversion.
If you're past 75 seconds and still unsure, flag and move on — you can't recover four lost minutes from one stubborn question.
If you want a shortcut: TestSolve reads each test question on your screen and sends the answer to your phone in about 5 seconds. Free first solve, no signup. Pricing.