A number memory test measures your digit span — how long a string of numbers you can hold in short-term memory after a single glance. A number flashes, disappears, and you type it back; each round it grows by one digit. The point where you finally slip is your span, and it's one of the cleanest windows into how your working memory actually works.
🔢 Take the number memory test →
The famous answer is about seven. In 1956, psychologist George Miller described the capacity of short-term memory as "the magical number seven, plus or minus two" — meaning most people manage somewhere between five and nine digits from a single look. On a growing-number test, reaching a span of seven puts you right at average. Interestingly, later research argues the number of truly independent "chunks" we hold may be closer to four — which is exactly why grouping digits helps so much.
Span shifts a little with age and rises reliably with practice, so the most honest benchmark is your own trend. Adding one digit to your steady span is a genuinely big jump in short-term capacity.
Short-term memory is a small, fast, leaky store. It can keep only a handful of separate items alive at once before older ones fade or get crowded out by new ones. That's why a string of individual digits gets hard around seven — you're bumping against the ceiling of how many separate things the system will hold. The way to beat that ceiling isn't to hold more items; it's to make each item bigger.
🔢 Take the number memory test →
About seven — Miller's "seven plus or minus two," so a range of roughly five to nine from a single glance.
Seven is average, eight is above average, and nine or ten is good — usually with chunking. Past ten almost always means deliberate grouping.
Short-term memory holds only a few separate items before older ones fade. Chunking beats the limit by making each item bigger.
Group digits like a phone number, keep a steady rhythm, anchor pairs to dates you know, and practise regularly.
Source. Miller, G. A. (1956), "The magical number seven, plus or minus two," Psychological Review, 63(2), 81–97. Later chunk-capacity estimates draw on Cowan, N. (2001), Behavioral and Brain Sciences, 24(1).