Attention, Scanning & Perception

How people read screens

Users scan, they don't read (Nielsen's early web studies; Krug). They satisfice: click the first plausible thing.
F-pattern (NN/g eyetracking, 2006 and 2017 update): on text-heavy, unformatted pages, gaze concentrates in two horizontal bands near the top then a vertical strip down the left. It's a symptom of poor formatting — counter it with front-loaded headings, bullets, bold keywords, meaningful first words. Z-pattern applies to sparse layouts (hero pages).
Layer-cake pattern: with good headings/subheadings, users scan headings and skip body — write headings that carry the meaning alone.
First words matter most: front-load links, headings, and list items with information-carrying words ("Pricing for teams" not "Learn more about…").

Selective attention & its failures

Banner blindness (Benway & Lane 1998; repeatedly replicated by NN/g): users learn to ignore anything that looks like an ad — including your own important content if it's styled like a promo (big, boxed, colorful, right-rail).
Inattentional blindness (Simons & Chabris 1999, "invisible gorilla"): users focused on a task literally don't see unexpected elements — don't put critical info outside the task's visual path; interstitials placed mid-task get dismissed unread.
Change blindness: changes outside the focus of attention go unnoticed; if the system updates something elsewhere on screen (price total, side panel), signal it with motion or highlight (safe use of common fate / animation).
Tunnel vision under stress: error states and time pressure narrow attention — error recovery UI must sit exactly where the user is looking (inline, next to the field), not in a distant toast.

Visual salience (what wins attention)

Pre-attentive features are processed in <250 ms: color contrast, size, orientation, motion (Treisman's feature-integration work; Healey & Enns 2012 survey). Motion is the strongest magnet and the most annoying — reserve it for what genuinely needs attention. One salient element per screen; salience is zero-sum.

Try it — salience is zero-sum. Accent more of the buttons and watch the one that matters stop mattering.

Habituation & alert fatigue

Habituation — diminished response to a repeated stimulus — is one of the most basic and best-established learning phenomena (Rankin et al. 2009 consensus paper). UI consequences: any warning, badge, banner, or confirmation shown routinely stops being seen; users develop automatic dismissal gestures (clicking through security warnings, swiping away permission prompts). Eye-tracking and fMRI work on security warnings shows attention to a warning drops sharply after only a few exposures, and that varying a warning's appearance ("polymorphic" warnings) slows habituation (Anderson et al., CHI 2015; Sunshine et al. 2009 on SSL warning click-through). Design rules: budget interruptions ruthlessly — every low-value alert taxes the credibility of future ones; make rare/dangerous confirmations look and behave differently from routine ones (different layout, require typing a name, not just the same OK button in the same place); never use a warning where a constraint or undo would work (habituated confirmations don't stop slips — see Mental Models, Affordances & Learnability).

Faces and gaze

Humans preferentially attend to faces (rich literature in vision science), and images of people looking at content can cue attention toward it — treat gaze-direction claims from marketing blogs cautiously; the reliable part is that faces attract fixation, so use human imagery deliberately, not as decoration near critical UI it will compete with.

Sources

Nielsen Norman Group: "F-Shaped Pattern of Reading" (2006; 2017 update), "Banner Blindness Revisited," "Text Scanning Patterns" (nngroup.com).
Benway, J. P. & Lane, D. M. (1998). "Banner blindness." ITG Newsletter.
Simons, D. J. & Chabris, C. F. (1999). "Gorillas in our midst." Perception, 28(9).
Treisman, A. & Gelade, G. (1980). "A feature-integration theory of attention." Cognitive Psychology, 12(1); Healey, C. & Enns, J. (2012). "Attention and visual memory in visualization." IEEE TVCG.
Rankin, C. H. et al. (2009). "Habituation revisited." Neurobiology of Learning and Memory, 92(2).
Anderson, B. B. et al. (2015). "How polymorphic warnings reduce habituation in the brain." Proc. CHI '15; Sunshine et al. (2009). "Crying wolf: An empirical study of SSL warning effectiveness." USENIX Security.
Krug, S. Don't Make Me Think (3rd ed. 2014) — scanning & satisficing.