Arithmetic Verification test online: true or false sums
An equation appears with its answer already filled in, and you decide as quickly as you can whether it is right. Over 90 problems, the timing shows how directly simple facts come back to you.
Start Arithmetic VerificationOnce the last problem is answered you see your accuracy and timing, and can keep them with a free account.
How it works
6 practice problems come first, then a 3-second countdown and 90 scored problems. Nothing limits how long you spend on each one, so the whole task runs about 5 minutes.
- 1 · Each problem
Right or wrong
A + sign holds the centre for half a second, then the equation appears. Press T for true or F for false, or click the buttons. On a phone you tap them.
- 2 · Before the count
6 practice problems
A mix of the 3 operations, 3 right and 3 wrong. After each answer the screen tints green or red, so you can check the keys before the scored part.
- 3 · The 90 problems
Small and large
30 additions, 30 subtractions and 30 multiplications, shuffled together. In each set, 15 problems use 2 numbers that add up to 10 or less and 15 use larger pairs.
- 4 · Wrong answers
Off by 1, 2 or 5
Half the problems in each set show a wrong answer. It sits 1, 2 or 5 away from the right one, above or below, and each distance comes up equally often.
The pairs are drawn at random on every attempt: additions from 1 + 1 to 9 + 9, multiplications from 2 × 2 to 9 × 9, and subtractions from numbers up to 18 that leave an answer of at least 1. A thin bar along the top shows your progress. Once the practice ends, nothing tells you whether an answer was right.
Judging an answer that is already written is a standard way to study mental arithmetic in the lab. People see 2 operands, an operation and a proposed result, and decide whether the result holds (Avancini et al., 2015). Because you never produce the answer yourself, the task leans on recognising a fact rather than writing it out.
What it measures
None of the problems is hard with enough time. The task asks how directly the answer is there for you when the clock is running.
Accuracy and speed
- Arithmetic fact fluency: telling a right sum from a wrong one, quickly and reliably.
- Accuracy for addition, subtraction and multiplication, each scored on its own.
- Your average time on the problems you got right.
Patterns in your timing
- Problem size: how much more time the larger problems take you than the small ones.
- Error size: whether you reject an answer that is far off faster than one that is close.
- Reading digits and pressing quickly, which any timed number task also depends on.
Almost everyone needs more time for larger problems. The finding is so regular that a handbook chapter on it is titled What everyone finds (Zbrodoff and Logan, 2005). What varies is the size of the slowdown: a small one suggests most answers come straight from memory, a large one that some are being worked out step by step.
Difficulty memorising arithmetic facts is described as a common and lasting feature of dyscalculia (De Visscher and Noël, 2014). That is why this task is one of the 4 in the dyscalculia check, next to tasks on quantities, written numbers and the number line. A slow or uneven result also has everyday explanations, from years without practice to a tired evening.
A test, not a diagnosis.
One run of 90 problems describes how fact checking went on that day. It cannot confirm or rule out dyscalculia or any other condition.
What you get
The result screen opens as soon as you answer the 90th problem.
On the result screen
- A headline with a short summary, your percentage of correct answers and your average response time.
- A line per operation: a strength from 85% correct, harder from 70%, a challenge below that.
- Your average time on small and on large problems, and how many times slower the large ones were.
- A line on how quickly you caught answers that were far off, and a data panel with both timing effects in milliseconds, accuracy and time per operation, plus a CSV download.
When you keep it
- With a free account the result is stored and opens on any device you sign in from.
- It sits with your other number tasks in your results, so calculation can be read beside number sense.
- Another test is suggested as soon as you finish, to keep building the picture.
How this version is scored
Each answer is logged with the problem, the key you pressed and your time. The result screen is computed from the 90 scored problems.
- Correct answer
- Pressing T when the answer shown is right, or F when it is wrong. The 6 practice problems do not count.
- Response time
- From the moment the equation appears to your key press or tap. Averages use correct answers only and leave out the half-second + sign.
- Problem-size effect
- Your average time on large problems minus your average on small ones, in milliseconds. A problem is small when its 2 numbers add up to 10 or less, and that holds for subtraction too, so 8 − 3 counts as large. Under 150 ms reads as steady speed, 150 to 249 ms as somewhat slower on harder problems, 250 ms or more as a large gap.
- Speed ratio
- Large-problem time divided by small-problem time, shown as, for example, 1.4× slower. The screen places it next to a typical gap of less than 1.2×.
- Split gradient
- Your average time to reject answers that were off by 1, minus your time for answers off by 5. More than 50 ms reads as using number size, above 0 up to 50 ms as weak, 0 or less as absent.
- Headline
- Solid when accuracy is at least 75% and neither timing pattern is in its middle or last band. Some difficulty when either one is in its last band, and mixed in every other case.
Cut-offs, not norms. There are no published adult norms for this exact true-or-false design, so the screen reads your times against fixed cut-offs rather than a percentile. Adult studies of simple arithmetic, such as Campbell and Xue (2001), mostly time answers that people produce themselves, which is a different measure.
No time limit. Each problem waits for you. Quick guessing lowers accuracy and slow checking raises your times, so a steady pace serves both numbers.
Before you start
Timing is half of this result, so a quiet, steady setup matters.
For a cleaner attempt
- On a computer, rest 2 fingers on T and F before the scored part starts.
- On a phone, hold it steady or set it down so both buttons are easy to reach.
- Silence notifications: one interruption adds seconds to a single problem and pulls the averages.
- Work each problem in your head; paper or fingers change what the times mean.
Interpret cautiously when
- You were tired, rushed, unwell or interrupted.
- Timed maths makes you anxious. Anxiety about maths can pull performance down by itself (Ashcraft and Moore, 2009).
- Reading digits or pressing keys quickly is hard for you for reasons that have nothing to do with arithmetic.
- You learned arithmetic in another language or with other symbols, or you took this task recently.
Questions
What does an arithmetic verification task measure?
It measures arithmetic fact fluency: how quickly and accurately you can tell whether a simple sum, difference or product is right. Accuracy is one part. The other is the timing pattern, especially how much the larger problems slow you down, which hints at whether answers are recalled or worked out.
What is a good score on an arithmetic verification test?
There is no published adult norm for this exact task, so the result screen uses fixed cut-offs. It calls an operation a strength from 85% correct, and it reads a problem-size gap under 150 ms as steady speed. Because the problems are simple, look at the timing lines as well as the percentage.
Can I take it on my phone?
Yes. On a touch screen you see True and False buttons without key hints. Hold the phone steady, because your response times are part of the result and a fumbled tap adds to them.
Is this the same task as in the dyscalculia check?
It is one of the check's 4 tasks. Inside the check it starts with a shorter introduction and 4 practice problems, then runs the same 90 problems as here. Taken on its own, it ends on the full result screen described above.
Sources
- Avancini, Soltész and Szűcs (2015): Separating stages of arithmetic verification, an ERP study with a novel paradigm
- Zbrodoff and Logan (2005): What everyone finds, the problem-size effect, in The Handbook of Mathematical Cognition
- De Visscher and Noël (2014): Arithmetic facts storage deficit, the hypersensitivity-to-interference in memory hypothesis
- Campbell and Xue (2001): Cognitive arithmetic across cultures
- Ashcraft and Moore (2009): Mathematics anxiety and the affective drop in performance