Citing a capacity finding requires checking the original task type

What it is

Every finding of the form "a person can hold N things in mind at once" corresponds to a specific task design underneath it — whether the material was digits or shapes, whether rehearsal was possible, whether the stimulus was still in view when the response was made, how long the delay was. These details determine exactly what N is, and whether that number can be carried over to a different scenario at all. Citing a number as if it were a task-independent universal constant is the common starting point for every way this kind of finding gets misused.

Why it happens

A capacity number can't survive being detached from its task because it is itself the product of a specific set of measurement conditions: changing the presentation format (all at once versus one at a time), whether rehearsal is allowed, whether the material is easy to organize, or whether the stimulus is visible at response time, all changes the resulting number. These differences aren't measurement noise — the task structure itself is what determines the observed capacity value. As a number passes through several hands — from the original paper to a review, from the review to a popular article, from there into a design guideline — each retelling risks dropping the crucial information of "under what conditions was this measured," leaving behind a bare number that, by the time it reaches whoever is using it, can no longer be checked against the boundaries it was originally meant to respect.

Studying it

The most direct way to verify whether a capacity number can be cited for a given scenario is to trace it back to the original study that produced it and confirm several key variables: material type (digits, letters, shapes, semantic content), presentation format (rehearsal allowed, presentation speed), response conditions (answering from memory versus with the stimulus still visible), and the participant population. Only when these variables are close enough to the scenario the number is meant to support does it carry any reference value. If the original study can't be located and only a secondary retelling is available, the right move is to keep tracing further back rather than accepting the number the secondary source hands over.

Where it stops holding

Even after fully verifying the original task type, a capacity number's applicability is still bounded to scenarios sufficiently similar to the original study — a large enough gap in material type, task pacing, or participant population can still leave the number without reference value. This finding only guarantees that a citation with the task type verified is more reliable than one without; it doesn't guarantee the verified number can then be applied unconditionally.

Related

• Same group: A6.10.1 Classic memory-capacity numbers don't apply to counting visible options · A6.10.2 Menu-item limits come from search cost, not memory · A6.10.4 The classic capacity number came from an absolute-judgment task on a single dimension, structurally unlike most interface scenarios · A6.10.5 Popular science flattened the number into a universal design rule, detached from its original conditions · A6.10.6 Design limits on password length or color count that borrow this number lack empirical support · A6.10.7 Deciding whether a design limit should invoke memory capacity requires first confirming the user is really doing unprompted recall
• Nearby: A6.02 Working memory capacity · Q1.15 Sample-size rules of thumb and their misuse
• Search terms: citation verification · task-dependent capacity · secondary source distortion