Memory recall tests: what you are asked to hold

The skill

What is a memory recall test?

A memory recall test in employment assessment presents a small set of information for a short, fixed period, removes it, and then asks about it. The question can be about what was shown, where it was, what order it came in, or which of several near-identical options matches what you saw.

Saville publishes the clearest example in Cognivy's launch scope. Its Swift Global Aptitude is a 12-minute combination test measuring number reasoning, abstract reasoning and memory recall, and Saville describes the recall component as assessing the ability to correctly remember figures and their position in a grid of cards. Saville states that Swift Global was designed for completion on a mobile device, uses minimised language content and is highly visual.

That description sets the honest scope of this guide. It is about remembering figures and positions in a short, visual, work-relevant assessment. It is not about clinical memory testing, school entrance exams or consumer brain-training products, and none of those formats predict what a Swift Global recall section asks.

Try it before anything else

Free memory recall practice questions

Six exercises covering the question types above, one question each. A page cannot withdraw a display the way the live test does, so run each one on yourself: study the data line for about 20 seconds, cover it, then answer from memory without looking back. The solution shows the encoding that makes the question answerable and names why every wrong option is on the list.

What employers are looking at

What do memory recall tests measure?

Where recall is assessed, it is usually as one component of a broad cognitive measure rather than as a standalone hurdle. Saville describes Swift Global as an interactive measure of the ability to think abstractly, reason quickly and problem-solve independent of previous knowledge, with recall as one of three components alongside number and abstract patterns.

Two abilities separate here, and only one of them is memory capacity. Encoding: in the few seconds available, did you convert what you saw into something structured — a pattern, a shape, a short phrase — rather than a list of unrelated details. Retrieval precision: can you distinguish what you actually saw from an option that is almost identical, which is what the distractors are built to test.

Reading your result

How memory recall tests are scored

Recall is scored as part of a combination test rather than as a paper of its own, and that changes what the number can tell you. Everything below comes from what Saville publishes about Swift Global reporting — there is no benchmark table here, because no such figure exists.

• One total score, with a recall sub-score underneath. Saville reports one total score plus individual sub-scores for its combination tests in a single report, so an employer can see the recall component separately from the overall result. Practise as if the recall line will be read on its own, because it can be.
• No pass mark is published. Saville does not publish a pass mark for Swift Global or for its recall component. If a recruiter has disclosed a target for your application, that figure is the authority, and nothing on this page can substitute for it.
• Recall is minutes inside a twelve-minute test. Swift Global Aptitude covers number reasoning, abstract reasoning and memory recall in 12 minutes together, so the recall section is a matter of minutes rather than a paper of its own. In a single overall figure a weak memory section can be invisible, which is why the separate sub-score is the line to watch.
• Difficulty rises by design, not in response to you. Saville states that its fixed-length format ensures a fairer experience for candidates with a gradual increase in difficulty. Later recall questions are designed to be harder rather than selected by your performance, so a hard run near the end of the section is the format working as built, not feedback on your score.
• Score your practice by cause, not by total. When you practise, record recall accuracy separately from the other components, and log whether each miss came from encoding, retention or a choice between near-identical options. Cognivy makes no claim that memory training raises a test score, because no provider in this scope publishes evidence on practice effects. The honest use of practice is fixing the routine, not chasing a number.

Question formats

What do the questions look like?

Recall items differ mainly in what they ask you to have stored, and that decides how you should encode in the first place.

• Position recall in a grid. Symbols or figures laid out in a grid, with a question about what was where. Saville describes its recall component as remembering figures and their position in a grid of cards.
• Visual recognition. A set of figures shown, then one figure offered with the question of whether it was in the set. Familiarity is not enough here, because the distractors are drawn to look familiar.
• Order recall. A sequence presented once, with a question about what came before or after a named item. Position within the sequence has to be encoded, not just membership of it.
• Sequence reconstruction. The whole order to be rebuilt rather than one relationship queried. Chunking the sequence into two or three groups is what makes this manageable.
• Short record recall. A reference, a date and an amount shown together, with one field asked about afterwards. The fields interfere with each other, which is the point.
• Recognition among similar distractors. Options that share the same characters in a different order, or the same symbol in a different place. This is where the difference between remembering the set and remembering the arrangement shows up.

Launch coverage

Which providers use memory recall tests?

One launch family publishes a memory component, and Cognivy's coverage stays inside it rather than borrowing formats from outside employment assessment.

• Memory Recall, within Swift Global Aptitude. Saville lists Swift Global Aptitude as a 12-minute combination test measuring number reasoning, abstract reasoning and memory recall, and describes the recall component as assessing the ability to correctly remember figures and their position in a grid of cards. Saville states that it was designed for completion on a mobile device, uses minimised language content and is highly visual, and offers a free practice test for Swift Global Aptitude under timed conditions.

Provider names identify the assessment format. Cognivy is independent and is not affiliated with or endorsed by these providers.

What varies, and what does not

Timing, delivery and scoring

Only one launch provider publishes a memory component, so the figures below are specific to it. Everything else about timing comes from your own assessment.

• Saville Swift Global Aptitude: 12 minutes for three components. Recall shares the twelve minutes with number reasoning and abstract reasoning, so the memory section is a matter of minutes rather than a paper of its own.
• Display time is set by the test, not by you. Recall items control how long the material is visible, which is what separates them from every other format in this library. There is no option to re-read, so the encoding decision has to be made immediately.
• Saville uses a fixed-length format. Saville states that its fixed-length format ensures a fairer experience for candidates with a gradual increase in difficulty, so later recall items are designed to be harder rather than selected by your performance.
• Designed for a phone. Saville states that Swift Global was designed for completion on a mobile device and is highly visual with minimised language content. Screen size changes how much of a display you take in at once, so rehearse on the kind of device you will actually use.

A repeatable approach

A method that survives the clock

Almost everything that decides a recall item happens before the display disappears. These five steps are about those seconds.

1. 1

   Decide what is being asked for before you look. Membership, position or order. If the assessment has told you it will ask about position, encoding the set of symbols is wasted effort, and the two encodings are genuinely different.

2. 2

   Chunk into two or three groups. Read a grid as rows, a sequence as pairs, a record as fields. Three chunks are three things to hold rather than nine, and the structure that groups them is already there in the material.

3. 3

   Attach a verbal label. Say the row to yourself: star, arrow, circle. A verbal label gives you a second copy of the material in a different form, so recall does not depend on the picture alone.

4. 4

   Encode position with the symbol, not separately. Top-left star rather than star and, somewhere, top-left. Position questions fail when the two are stored as separate facts that then have to be reconnected.

5. 5

   Answer from your encoding, not from recognition. If an option feels familiar but does not match what you encoded, trust the encoding. Distractors in this format are built to feel familiar, which is precisely why familiarity is not evidence.

What slows progress

Common mistakes, and the fix for each

• Storing details as an unrelated list. Fix: chunk into two or three groups and label each one. Nine separate facts exceed what most people hold for a few seconds; three labelled groups do not.
• Confusing familiarity with recall. Fix: answer from what you encoded rather than from what feels right. Distractors are constructed from the same characters and symbols precisely so that everything on offer feels familiar.
• Encoding for the wrong question. Fix: check what the item type asks for. Membership, position and order need different encodings, and preparing the wrong one costs the mark even when your memory worked perfectly.
• Changing strategy between items. Fix: choose one encoding routine and keep it. Switching mid-test means each new item starts from scratch, and there is no time in a few-second display to experiment.
• Practising recognition only. Fix: practise position and order deliberately. Recognising that you have seen something is a smaller demand than reproducing where it sat, so practice built only on recognition leaves the harder half unrehearsed.

In practice

How do you encode a grid you will only see once?

Decide what the question will ask before you look. Membership, position and order need genuinely different encodings, and a few seconds of display leave no room to prepare all three. Saville describes its recall component as remembering figures and their position in a grid of cards, and for a position question, encoding a bare list of which symbols appeared is preparation for the wrong question.

Then read the grid as rows, and store position with the symbol rather than beside it: top-left star rather than star and, somewhere, top-left. Sample question 1 turns entirely on this — each of its four answer options names a symbol that appears twice in the grid, so remembering which symbols were present answers nothing, while three said rows make the question straightforward.

Give the rows words as you go. A short verbal phrase — star, arrow, circle — is a second copy of the material in a different form, and it can be rehearsed silently across the gap between display and question, which a fading visual impression cannot.

In practice

Chunking: what actually works under exam conditions?

Chunking is not a trick; it is using organisation the material already has. A grid arrives as rows, a sequence breaks into pairs, a record is a set of labelled fields. Two or three chunks are two or three things to hold across the gap, where nine unrelated details are not — and a repeat, like the doubled name in sample question 3, is one stored fact covering two positions.

Under exam conditions the routine matters more than the theory. Fix one chunking per material type before the test — rows for grids, pairs for sequences, fields in a set order for records — and keep it for every question, because a few-second display leaves no time to experiment. Start encoding the first chunk immediately rather than taking the display in as a whole: the material is withdrawn on a timer, and a chunk you have encoded is still available when a general impression is not.

Sample question 2 shows the payoff. Five codes chunked as K7 M2, then K2 M7, then K9 turn a what-came-before question into reading the middle pair, and the same move is what makes reconstruction manageable when the whole order has to be rebuilt rather than one relationship queried.

In practice

How do you tell real recall from familiarity?

Familiarity is the feeling of having seen an option before, and in this format it is worthless, because you have seen every option before. Distractors are constructed from the same characters, symbols and digits as the material — the four door codes in sample question 4 are the same four digits rearranged, and all four sat on the studied list — so nothing on offer looks new.

Real recall is retrieving the encoding you made: the said row, the labelled pair, the number bound to its field. When an option feels right but does not match the phrase you rehearsed, trust the phrase. Then answer promptly once you have retrieved it, because deliberating between two options lets the encoding decay, and a near-identical distractor is built to look better the longer you stare at it.

In review, the two are separable errors. A miss where you never encoded the detail needs a better routine in the seconds you are given; a miss where you encoded it and then chose the familiar option anyway needs the discipline of answering from the encoding. The fix for one does nothing for the other, which is why logging the cause of each miss beats counting the total.

A clear route

How to prepare in the days you have

Preparation here is about routine rather than about expanding capacity. Cognivy makes no claim that memory training raises a test score; what you can work on directly is what you do in the seconds you are given.

1. 1

   First, fix one encoding routine. Rows for grids, pairs for sequences, fields in a set order for records. Write it down and use nothing else, so that no item begins with a decision.

2. 2

   Next, practise chunking untimed. Look at a grid, look away, say the three rows aloud. Untimed rehearsal builds the routine; adding the clock before the routine exists just produces guessing.

3. 3

   Then practise the specific question types. Position, order and recognition each need their own sessions, because each rewards a different encoding. Mixing them early hides which one you are weak at.

4. 4

   Shorten the viewing time gradually. Only after accuracy is stable. Cutting display time before the encoding routine is automatic trains guessing rather than recall.

5. 5

   Rehearse on the right device. Saville states that Swift Global was designed for completion on a mobile device and is highly visual. If your assessment is mobile-first, practising on a large monitor is rehearsing a different task.

Cognivy uses your assessment date to choose the route rather than asking you to predict a study schedule. When you sit down to practise, you choose the session length that fits that day.

Pace

Encoding fast without losing accuracy

There is no speed decision during a recall item; the display controls the clock. The four habits below decide what you get out of the time you are given.

• Start encoding immediately. Do not spend the first second taking the display in as a whole. Begin with the first chunk and work across, because the display is withdrawn on a timer and a chunk you have encoded is still available when a general impression is not.
• Say it, do not just see it. A short verbal label can be rehearsed silently across the gap between display and question, which a visual impression cannot.
• Use structure that is already there. Rows, columns, repeats, symmetry. Any structure in the material is free organisation, and a repeat is one item to store instead of two.
• Answer quickly once you have retrieved it. Deliberating between two options lets the encoding decay, and a near-identical distractor is built to look better the longer you stare at it.

Keep exploring

Related test types and how they differ

Memory recall borders three test types that ask for a similar kind of care with detail.

Direct answers

Memory recall test FAQs