A task-irrelevant, highly salient element keeps consuming attentional resources, even once known to be irrelevant

What it is

Even after a user has confirmed through repeated experience that a highly salient element is entirely irrelevant to the current task, and has learned to ignore it, that element still continues to consume some attentional resources — showing up as a slightly lower reaction speed or accuracy on the primary task compared to a condition where the element does not exist at all. "Known to be irrelevant" and "no interference at all" are two different things — the former does not guarantee the latter.

Why it happens

This reveals a ceiling on what top-down suppression can accomplish: suppressing an irrelevant stimulus requires continuously investing cognitive resources to actively dampen its salience signal, and this dampening process itself has a cost — it is not a free switch. In other words, even when suppression succeeds — the user genuinely did not shift attention there, and behavior shows no obvious disruption — the resources spent maintaining that suppression are still drawn from the primary task's resource pool.

This is a different question from whether a larger salience gap is harder to suppress: that entry asks whether a capture signal breaks through suppression; this one asks whether maintaining suppression itself carries a cost even when suppression fully succeeds — and the answer is yes.

Studying it

A common method is an extended version of the additional singleton paradigm: the same participants repeatedly practice the same search task with the same irrelevant distractor until both subjective report and explicit measures show "no longer affected at all" (target search reaction time no longer slows significantly when the distractor is present). At that point, a secondary probe task is inserted, and its performance is measured to see whether it is still worse than in a no-distractor condition — indirectly measuring how much resource maintaining suppression consumes.

Typical dependent variables: primary-task performance (a direct measure, which usually shows no visible difference once well-practiced) and secondary-task performance (an indirect measure that reveals resource consumption masked by seemingly normal primary-task performance).

In interface research, this method is mainly used to evaluate whether a long-present interface element that users have "learned to ignore" (a fixed-position ad slot, a persistent but rarely used toolbar) is still quietly dragging down overall primary-task efficiency, even when usability testing shows users cannot articulate being affected by it at all.

Methodological caveat: this kind of effect is usually small and requires a sufficiently sensitive secondary task and a large enough sample to detect — it cannot be expected to show up directly in a small-sample usability test. This is exactly why this finding is easy for product teams to overlook: it is nearly invisible in routine testing.

Where it stops holding

• The effect size is usually small, and is more easily amplified and observed when the task itself already carries a high load and primary-task resource consumption is near saturation.
• If the irrelevant element's salience is low, the resource it continues to consume is correspondingly smaller — this finding describes a continuous cost, not an all-or-nothing switch.
• Once the element is removed entirely over the long term, the previously occupied resource is released, which shows the cost genuinely comes from the element's ongoing presence, not some irreversible loss of capability.

Applying it

• Do not conclude an element has no effect on efficiency just because users say in interviews or usability tests "I never noticed that area at all" — this needs active measurement rather than relying on subjective self-report, since this effect is precisely the type where "the user isn't aware of it, but the cost is still there."
• A long-present, high-salience interface element confirmed to be useless to most users (a persistent but rarely clicked promotional slot, a decorative motion effect) is worth periodically re-evaluating for removal — checking whether core task metrics show even a small improvement — even without any user complaints.
• For a high-salience but task-irrelevant element that must be kept, prioritize reducing its salience rather than assuming "users are used to it, so it's fine" — habituation only reduces the visible, behavioral disruption; it does not zero out the resource cost.
• Verification: use a dual-task or secondary probe task to measure the resource-occupation difference between conditions with and without the element present, rather than only measuring primary-task performance itself or asking users for their subjective impression.

Related

• Same group: A5.07.1 Sudden onset and motion automatically capture attention · A5.07.2 Captured attention takes time to return · A5.07.3 Frequent capture leads users to actively block out that region · A5.07.4 The larger the salience gap, the harder capture is to suppress · A5.07.5 Bottom-up capture happens early in processing, before content is semantically identified
• Nearby: A5.02 Divided attention and dual-task performance · A5.09 Cognitive tunneling
• Search terms: residual capture cost · additional singleton · top-down suppression cost