Is there a way to test working memory?

Yes. Clinical and research tests include digit span, the n-back task and the Symbol Digit Modalities Test. In job assessments, working memory is usually tested indirectly through timed tasks like symbol coding with arithmetic, where you must hold several values and a rule in mind at once.

What are some effective short-term memory exercises for a test like this?

Drill the exact format: decode symbols, calculate, compare. Add covered-legend trials, in which you read and then hide the legend, and backwards digit span. Short daily sessions of 10–20 minutes over a week help more than one long session.

Is working memory the same as IQ?

No. Working memory is one component of cognitive ability and relates to reasoning performance, but it isn't the same thing. In the eligo battery it is reported as its own module, "Processing speed", next to the reasoning modules.

What is a symbol digit test?

A test in which a key pairs symbols with digits and you translate as many symbols as possible in a fixed time. The classic clinical form, the SDMT, uses nine symbols, the digits 1–9 and 90 seconds. Employment versions such as eligo's add arithmetic and a larger/smaller decision, and reshuffle the key on every trial.

How many trials can I expect per block?

It depends entirely on your speed. Blocks last 30 or 50 seconds with no per-trial timer, so at two to three seconds per trial you would complete roughly 10–20 trials per block.

Can I use a calculator or paper?

The task is too fast for either to help, and calculations stay within single digits. Use paper only if your invitation allows it, and only for the other modules.

Working Memory Test: Symbol Coding Practice

A working memory test in a job assessment is usually a short, heavily timed task: you hold a few pieces of information in mind, transform them and respond, over and over, against the clock. The most common hiring version is symbol coding with mental arithmetic. You look up what each symbol stands for, add, multiply or subtract, and click the larger or smaller result. In an autumn 2026 eligo PERLS® DUO battery, this "processing speed" module ran as 12 blocks of 30 or 50 seconds each, about eight minutes in total. This guide explains the format and gives you 30 original drills with an answer key.

Key facts

Publisher: eligo PERLS® DUO, Processing speed module
Format: symbol-to-number coding, arithmetic and larger/smaller comparison
Sections: 4 untimed samples, then 6 conditions × 2 timed blocks = 12 blocks
Time: 0:30 or 0:50 per block, no per-trial timer, about 8 min in total
Answer format: click the left or right button
Scoring: percentile against a comparison group; error weighting not published
Adaptive: no; the legend reshuffles on every trial

Working memory tests, processing speed and why employers use them

Working memory is the mental workspace where you hold and manipulate information for a few seconds, such as a phone number you're dialling or a running total. Processing speed is how fast you can take in simple information and act on it. Employment tests rarely measure either in a pure form. They combine both: you must find two or four symbols in a legend, keep their values in mind, calculate, and remember whether you're looking for the larger or the smaller value.

Clinical versions are well studied. In the Symbol Digit Modalities Test, a key pairs the digits 1–9 with nine geometric symbols, and you write or say the matching digit for as many symbols as you can in 90 seconds. The n-back task, introduced by Wayne Kirchner in 1958, asks you to signal whenever the current item matches the one shown n steps earlier. Digit span asks you to repeat ever-longer strings of digits. If you searched for a free online working memory test for medical reasons such as ADHD, see a clinician. This page is about job assessments.

What the eligo symbol speed task looks like

In the 2026 run described here, the module came last, after logic, charts and matrices.

Screen: an instruction line, "Please select the larger value." or "Please select the smaller value.", with the key word in bold. Below it are two grey arrow-shaped buttons with "or" between them. The left button points left, the right one points right.
Expressions: each button holds a single symbol or a two-symbol expression with `+`, `*` (multiply) or `-`.
Legend: a box maps nine filled dark-grey symbols to the digits 1–9. The symbols are a triangle ▲, an inverted triangle ▼, a right triangle ◣, a diamond ◆, a square ■, a hexagon ⬢, a circle ●, an X-cross ✖ and a plus-cross ✚. The mapping is reshuffled on every single trial, so memorising it is impossible.
Timer: a countdown with a small clock graphic top right. A large "5" countdown ring appears before a block. You answer as many trials as you can until the block ends.
Samples: four untimed sample screens with arrow navigation come first.

Condition	Left vs right	Rule seen	Block
A	symbol vs symbol	larger	0:30
B	a + b vs c	smaller	0:30
C	a * b vs c	larger	0:30
D	a + b vs c + d	smaller	0:50
E	a b vs c d	larger	0:50
F	a − b vs c − d	smaller	0:50

The larger/smaller word alternated between conditions in this run. Treat that pattern as observed, not guaranteed, and read the instruction every time.

Design features you should expect:

Near ties. Many pairs differ by exactly one (for example 28 vs 27, 45 vs 44, 15 vs 16), so estimating fails.
Look-alike symbols. The three triangles ▲ ▼ ◣, the two crosses ✖ ✚, and the diamond ◆ next to the square ■.
Drawn size means nothing. One symbol may be drawn smaller than another while carrying the higher value.
Small operators. The `-` sign is easy to misread as `*` at speed, and that single slip flips an answer.

Worked example

Select the smaller value. `● - ✖` or `▼ - ■` Legend: ■1 ◆2 ✖3 ▲4 ⬢5 ◣6 ●7 ✚8 ▼9

Read the bold word: smaller.
Read the operators: both are subtractions.
Look up left: ● = 7, ✖ = 3, so 7 − 3 = 4.
Look up right: ▼ = 9, ■ = 1, so 9 − 1 = 8.
4 < 8, so click left.

The whole cycle should take about two to three seconds. Don't say the symbol names to yourself; go straight from shape to digit.

Practice drills

Each line is one trial with its own legend, just as on screen. Cover the answer key, set a timer and aim for the targets in the table below. On paper, write L or R for each line. Expressions use `*` and `-` as in the real task.

Drill	Target for 4 trials	Real-test equivalent
A–C	under 10 s	0:30 blocks
D–F	under 14 s	0:50 blocks
G (switching)	under 20 s for 6 trials	changing rules

Drill A: single vs single — select the larger value

A1. `✚` or `▲` — legend: ◆1 ✖2 ⬢3 ◣4 ▼5 ■6 ✚7 ▲8 ●9 A2. `✖` or `◣` — legend: ■1 ●2 ◆3 ▲4 ⬢5 ▼6 ✚7 ✖8 ◣9 A3. `✖` or `▲` — legend: ◆1 ▲2 ✖3 ✚4 ▼5 ●6 ◣7 ■8 ⬢9 A4. `◆` or `⬢` — legend: ●1 ⬢2 ■3 ◆4 ✚5 ▲6 ◣7 ✖8 ▼9

Drill B: sum vs single — select the smaller value

B1. `✚ + ■` or `◆` — legend: ▲1 ⬢2 ✚3 ●4 ■5 ▼6 ◣7 ✖8 ◆9 B2. `▼ + ◣` or `⬢` — legend: ▼1 ⬢2 ◣3 ✚4 ●5 ◆6 ▲7 ■8 ✖9 B3. `⬢ + ◣` or `▼` — legend: ⬢1 ◣2 ■3 ▲4 ▼5 ●6 ◆7 ✖8 ✚9 B4. `✖ + ■` or `✚` — legend: ◆1 ▼2 ✖3 ●4 ■5 ◣6 ⬢7 ▲8 ✚9

Drill C: product vs single — select the larger value

C1. `✖ ▼` or `◆` — legend: ▼1 ▲2 ■3 ●4 ✚5 ✖6 ⬢7 ◣8 ◆9 C2. `✖ ✚` or `■` — legend: ✚1 ▲2 ◣3 ■4 ✖5 ▼6 ◆7 ●8 ⬢9 C3. `▲ ◆` or `▼` — legend: ▲1 ◆2 ■3 ●4 ▼5 ◣6 ⬢7 ✖8 ✚9 C4. `⬢ ●` or `▲` — legend: ●1 ✚2 ◣3 ▲4 ⬢5 ✖6 ■7 ▼8 ◆9

Drill D: sum vs sum — select the smaller value

D1. `● + ◣` or `▲ + ◆` — legend: ▼1 ■2 ▲3 ◣4 ✖5 ✚6 ●7 ⬢8 ◆9 D2. `◆ + ■` or `▼ + ▲` — legend: ◣1 ▲2 ■3 ●4 ⬢5 ✖6 ▼7 ◆8 ✚9 D3. `◆ + ✖` or `■ + ●` — legend: ◣1 ▲2 ⬢3 ■4 ✖5 ●6 ◆7 ✚8 ▼9 D4. `▲ + ●` or `▼ + ✖` — legend: ◆1 ■2 ▲3 ▼4 ✚5 ⬢6 ◣7 ●8 ✖9

Drill E: product vs product — select the larger value

E1. `■ ◣` or `✖ ▼` — legend: ⬢1 ◆2 ✖3 ■4 ▲5 ✚6 ◣7 ▼8 ●9 E2. `✚ ▼` or `◣ ▲` — legend: ⬢1 ■2 ◣3 ✚4 ▼5 ◆6 ●7 ▲8 ✖9 E3. `▼ ◆` or `✚ ◣` — legend: ✚1 ▼2 ●3 ■4 ◆5 ⬢6 ◣7 ▲8 ✖9 E4. `⬢ ▼` or `◣ ✚` — legend: ◆1 ✖2 ■3 ▲4 ⬢5 ✚6 ◣7 ●8 ▼9

Drill F: difference vs difference — select the smaller value

F1. `● - ▼` or `✖ - ■` — legend: ✚1 ▼2 ◣3 ▲4 ■5 ⬢6 ●7 ◆8 ✖9 F2. `⬢ - ◣` or `◆ - ■` — legend: ✖1 ■2 ●3 ◆4 ◣5 ✚6 ▼7 ⬢8 ▲9 F3. `▲ - ◆` or `✖ - ●` — legend: ◆1 ■2 ⬢3 ●4 ◣5 ✚6 ▲7 ▼8 ✖9 F4. `✚ - ▲` or `✖ - ◣` — legend: ◆1 ◣2 ▼3 ✖4 ▲5 ✚6 ●7 ⬢8 ■9

Drill G: switching — the rule changes on every line

G1. larger: `⬢ ◆` or `◣ ●` — legend: ◆1 ◣2 ■3 ●4 ⬢5 ✚6 ▼7 ✖8 ▲9 G2. smaller: `▲ + ✚` or `■ + ◣` — legend: ●1 ⬢2 ✚3 ▼4 ■5 ◣6 ▲7 ◆8 ✖9 G3. smaller: `▲ + ⬢` or `✖` — legend: ▼1 ▲2 ◆3 ✚4 ✖5 ⬢6 ●7 ◣8 ■9 G4. larger: `⬢ + ▲` or `◣ + ●` — legend: ◆1 ▼2 ●3 ▲4 ✚5 ◣6 ⬢7 ■8 ✖9 G5. smaller: `◆ + ▼` or `■` — legend: ✖1 ◆2 ▼3 ■4 ◣5 ⬢6 ▲7 ●8 ✚9 G6. larger: `⬢ ■` or `✖ ✚` — legend: ●1 ▲2 ◣3 ✚4 ■5 ⬢6 ✖7 ◆8 ▼9

Answer key

A: A1 right (7 vs 8) · A2 right (8 vs 9) · A3 left (3 vs 2) · A4 left (4 vs 2)
B: B1 left (8 vs 9) · B2 right (4 vs 2) · B3 left (3 vs 5) · B4 left (8 vs 9)
C: C1 right (6 vs 9) · C2 left (5 vs 4) · C3 right (2 vs 5) · C4 left (5 vs 4)
D: D1 left (11 vs 12) · D2 right (11 vs 9) · D3 right (12 vs 10) · D4 left (11 vs 13)
E: E1 left (28 vs 24) · E2 right (20 vs 24) · E3 left (10 vs 7) · E4 left (45 vs 42)
F: F1 right (5 vs 4) · F2 right (3 vs 2) · F3 right (6 vs 5) · F4 left (1 vs 2)
G: G1 right (5 vs 8) · G2 left (10 vs 11) · G3 right (8 vs 5) · G4 left (11 vs 9) · G5 right (5 vs 4) · G6 left (30 vs 28)

Working-memory variants (harder than the real task)

These go beyond the eligo format. They train the "hold it in your head" part.

Covered legend: Read the legend of any drill line for 5 seconds, cover it, then solve the trial from memory. Start with drill A and work up to D.
Two-back on drills: Read only the results of drill B aloud in order. After each one, say whether it is larger than the result two trials earlier.
Digit span: Ask someone to read you digit strings (start with 5 digits, add one each round) and repeat them backwards. Stop when you fail twice at one length, note your longest correct length and try to raise it by one during the week.

Mental arithmetic you need to be automatic

Because each digit appears only once in a legend, the two digits in one expression are always different. Everything stays within 1–9, so the arithmetic is small. It has to be instant.

Products of two different digits. Know them as well as your own phone number. Watch the pairs that are equal or nearly equal:

Product	Ways to make it	Near neighbours
12	2 × 6, 3 × 4	14 (2 × 7), 10 (2 × 5)
18	2 × 9, 3 × 6	20 (4 × 5), 21 (3 × 7)
24	3 × 8, 4 × 6	21 (3 × 7), 27 (3 × 9), 28 (4 × 7)
36	4 × 9	35 (5 × 7), 40 (5 × 8)
42	6 × 7	40 (5 × 8), 45 (5 × 9)
48	6 × 8	45 (5 × 9), 54 (6 × 9), 56 (7 × 8)

Sums run from 3 to 17. Compare sums of two pairs without adding when you can. For example, 4 + 9 vs 5 + 7: 4 is one less than 5 and 9 is two more than 7, so the left side is larger by 1.
Differences: in the trials seen, the first symbol always held the larger digit, so every result was between 1 and 8. Two differences are often only one apart (3 vs 4, 2 vs 3), so calculate both rather than guessing.

Traps and how to avoid them

Rule flip. Start every trial by reading the bold word: "larger" or "smaller". Picking the larger side on a "smaller" block is the most expensive error, because you repeat it for the whole block.
Operator misread. Check `*` vs `-` vs `+` before looking up any symbol.
Stale legend. The legend changes every trial. Never reuse a value from the previous screen.
Triangle confusion. ▲ (point up), ▼ (point down) and ◣ (right angle at the bottom left) are different symbols. Learn them once and look carefully.
Speed over accuracy. It isn't published whether wrong clicks are penalised. Since you can't know, aim for errorless runs first and then shave time.
Mouse travel. Keep the pointer between the two buttons so that both are one short move away.

How to prepare: a one-week plan

Day	Practice	Time
1	Times tables to 9 × 9 until instant; flash cards for the near-tie products above	15 min
2	Drills A–C on paper, three passes; note your time per 4 trials	15 min
3	Drills D–F, three passes; then make your own lines with a shuffled legend	20 min
4	Drill G (switching) plus all drills in random order	20 min
5	Working-memory variants 1–2	15 min
6	Full simulation: 12 blocks (A–F twice) with 30 s / 50 s timers	10 min
7	Light review and rest	5 min

Set-up on the day: use a mouse rather than a trackpad, put the browser in full screen at 100 % zoom, and work in a quiet room. Don't do this module straight after a heavy meal or late at night. The task rewards alertness more than knowledge.

Sources

執筆：

Adrian、専門家

心理学修士、7年の受験歴。

最終更新：Oct 1, 2026