Digit Symbol Substitution Test

Every other test on this site offers you a length. This one does not, and the reason is that the score is a count rather than a rate.

A count only means something next to the clock it was taken against. Sixty substitutions is a good run in ninety seconds and a poor one in three minutes. Offer the choice and the same number on the same card would describe two different performances. Your own history would stop being a history, because run four would no longer be comparable with run one.

The published versions of this task fix the clock for the same reason. Ninety seconds is the window the classic pencil form uses, so that is the window here. It is long enough that the first few slow items stop dominating the total, and short enough that nobody is measuring your patience.

The clock also keeps running through a wrong answer. There is no penalty and no correction. You pick a symbol, the next digit arrives, and the run carries on. Accuracy is reported separately on the card. A fast and careless run and a slower exact one can finish on the same count.

The row of nine pairings at the top never disappears. That is not a kindness, it is the design of the task. Hide the key and the score becomes a measure of how much you memorised, which is a different test with a different number.

What is left when memory is taken out is a small loop that runs once per item. You read a digit. Your eye travels to the key and finds that digit. You read the symbol under it, hold that shape for the moment it takes to look away, and find the same shape on the pad. Then you press.

Each of those steps is quick. The count is high because the loop is short, and it is a useful number because almost nothing else is going on. There is no puzzle, no vocabulary and no trick. Whatever slows the loop down shows up directly in the total.

The pad is laid out in the symbols' own order and the key is not. If the two matched, the answer to every item would be the position of the digit rather than its symbol. You could clear the whole run without looking at the key.

Which symbol belongs to which digit is decided fresh at the start of each run. Take the test three times and you get three different keys.

A fixed key would be memorised, and quickly. People pick up these pairings without trying to, which is well documented and has a name in the literature: incidental learning. By a third attempt most of the key would be in your head. The trip to the top of the screen would stop happening. The count would climb for a reason that has nothing to do with speed.

That climb is real and it is worth something, but it is not what this page reports. Shuffling the key holds the task still between runs. A change in your number is then a change in you, not in what the task was asking.

It has one cost, and it is worth naming. Your first fifteen or twenty items of every run are spent learning a key you have never seen. That is part of the ninety seconds for everyone, on every attempt, which is what makes runs comparable.

The card reports a speed-up. It compares the middle time per item across the last third of your correct answers with the same figure from the first third.

Almost all of that gain is the key. Early on, every item costs you a trip to the top of the screen. By the end of ninety seconds you have used some pairings a dozen times, and those come back without a look. The trip stops happening for the digits you know and keeps happening for the ones you do not.

A small part is the pad. Your hand learns where the nine buttons are, in the same way it learns a keyboard. That is the part this task shares with average typing speed. There the same effect separates hunting for a key from knowing where it is.

A speed-up near zero is not a fault. It usually means the run opened quickly, which leaves less room to improve inside a single ninety seconds. A negative number means the last third was slower than the first, and tiredness is the usual reason.

• A large speed-up with a low count often means a slow start rather than a fast finish.
• A small speed-up with a high count is the profile of someone who arrived already warmed up.
• Accuracy below about 90% usually means the pad is being pressed before the key has been read.

The comparison block on this page draws no bar. There is no ranking, no percentile and no band. That is deliberate and it is worth explaining, because every other digit symbol test online will give you one.

This task has been measured properly for decades, and all of that measurement is of a pencil. The published distributions count symbols a person drew by hand in a fixed window, so the figure contains the speed of handwriting. Here you press one of nine buttons instead. The drawing is gone and a search of the pad has taken its place, and those are different amounts of work.

Nobody has published a distribution for the clicked version. Borrowing the pencil curve would look responsible and would read as a fact, and it would be neither. A median that is out by a third moves an ordinary run from the middle of the range to the bottom quarter. That is a confident wrong statement about a person.

So the page shows your number, your best and your own history, and says the ranking does not exist yet. For a figure that does sit against published data, what counts as a good reaction time is a better place to start. It also explains how much of any such number belongs to your equipment.

One thing about this task does survive the change from pencil to pad, and that is the shape of the age effect. Processing speed by age sets that shape out as percentages rather than counts. It also shows how far the age bands overlap, which is the part that stops a curve from being a ranking.

A high count says you ran a short lookup loop quickly, on this screen, on this day. The key was one you had never seen ninety seconds earlier. That is all it says.

It is not an intelligence score. It is not a measure of how quickly you think in general, and nothing here detects or rules out any condition. Practice will move the number, and what improves is this task. Getting quicker at matching digits to symbols makes you quicker at matching digits to symbols.

The clinical version of this task exists and it looks similar. It is given on paper by someone trained to administer it, alongside other measures. A qualified person reads the result. A page in a browser is not that, and anything that actually worries you belongs with a doctor rather than with a score.