Intrinsic load comes from the task's own complexity

What it is

Intrinsic load is the processing burden that comes from having to hold and relate several interdependent pieces of information in working memory at once, just to do the task. It's set by the structure of the task or domain itself, not by how it's presented — a cleaner layout or less visual clutter won't shrink it. This is where people mix it up with extraneous load: extraneous load is added by presentation and can in principle be driven to zero, while intrinsic load is the minimum processing the task itself demands — it's still there even under the best possible presentation.

Why it happens

The size of intrinsic load tracks the task's element interactivity: if the pieces of information can be processed one at a time, independently (memorizing a list of unrelated names), interactivity is low and so is the load; if the elements only make sense when considered together (weighing several variables that constrain each other, where changing one forces you to reconsider the rest), interactivity is high and so is the load. Interactivity is a property of the task or domain, not of how it's displayed — the same high-interactivity task carries the same set of interdependent relationships whether it's shown as text, a chart, or spoken aloud.

Studying it

Researchers isolate intrinsic load by manipulating element interactivity while holding presentation constant: two versions of the same content are built with different interactivity (elements that can be learned independently vs. elements that must be understood jointly), with identical format, length, and sequencing, so any difference in learning or task outcome can be attributed to the task itself rather than to layout. The typical independent variable is the degree of element interactivity; dependent variables are learning outcome, accuracy, and completion time. In interface research this is used to tell whether a reported difficulty comes from the task or from how it's shown. The main methodological pitfall: if presentation also differs between conditions, the measured gap can no longer be attributed to intrinsic load alone.

Where it stops holding

Intrinsic load isn't a fixed number that's the same for every user. For a novice, a task may require weighing each element separately — high interactivity. For someone who already holds the relevant schema, several elements can be retrieved and handled as one chunk, so the load actually experienced is lower. This doesn't overturn the definition; it means the "inherent" complexity of a task is inherent relative to the processor's existing knowledge structure — the same task description doesn't carry equal interactivity for everyone.

Applying it

Its main value is diagnostic rather than corrective: when users report a feature is hard, first work out whether the difficulty sits in the task itself or in how it's presented. If stripping visual noise and trimming steps still leaves users struggling to keep track of things, the bottleneck is intrinsic load, and further interface polish won't help — the fix has to touch task scope (fewer variables to juggle at once, breaking the task into a sequence of smaller subtasks) or user capability (training, pre-building the relevant knowledge), not another round of visual design. Verification: keep the interface untouched and vary only the number of variables the task requires handling simultaneously; if completion time and errors move accordingly, the difficulty genuinely traces to the task's element interactivity rather than to presentation.

Related

• Same group: A9.01.2 extraneous load comes from presentation and can be designed away · A9.01.3 germane load builds understanding and shouldn't be driven to zero · A9.01.4 cutting extraneous load is the primary design lever
• Nearby: A9.06 intrinsic, extraneous and germane load (shared-capacity mechanism and theoretical origin of the three-way split)
• Search terms: intrinsic cognitive load · element interactivity · cognitive load theory