Your eyes do not scan smoothly. They make rapid jumps called saccades with brief pauses called fixations. Nearly all visual processing happens during fixations — but the depth of that processing depends heavily on how long the fixation lasts. The brain suppresses visual input during saccades.
Drag the slider to change fixation duration and see how the processing window and circle change.
Duration250ms
Duration
250ms
Processing depth
Medium
Saccades/sec
3–5
Blind during jump
~90%
A fixation is not binary — different durations unlock different levels of information extraction. Drag the slider to see which cognitive operations complete within your chosen duration.
Duration250ms
Basic features (colour, contrast, orientation)0–50ms
Even during a fixation, the eye is never truly still. Tiny involuntary movements called microsaccades (0.1–1°) occur continuously — and they reveal cognitive and emotional state.
What they do
Prevent Troxler fading — without microsaccades, static images fade from perception within seconds. They refresh the retinal signal continuously.
High rate = stress
Elevated microsaccade rate signals anxiety, cognitive load, and emotional arousal. The eye "searches" more when the brain is activated.
Low rate = flow
Deep focus and flow states reduce microsaccade frequency — the eye holds more steadily on the fixation point during peak concentration.
Clinical use
Microsaccade analysis used in research for ADHD, autism, schizophrenia, and as a covert attention probe — no instruction needed.
Experts show longer average fixations than novices on identical stimuli — but fewer total fixations. This efficiency gap is the primary gaze signature of domain expertise.
02 · Core Mechanics
Eye Tracking Signals
Eye tracking captures four primary signals — each measuring a different aspect of attention, cognition, and arousal. Together they provide a complete window into attentional state, cognitive load, emotional arousal, and fatigue — simultaneously and non-invasively.
100–600ms
Duration range
~250ms
Reading avg
~400ms
Scene viewing
85%
Time spent fixating
Fixation radius and colour in heatmap software encode dwell time — a large dark-red circle means long, deep processing. This is the data behind every eye tracking heatmap.
20–200ms
Duration
100–700°/s
Peak velocity
~90%
Visual suppression
3–5/sec
Frequency
Forward saccade
Advances through content
Typical 5–15 character span
Guided by peripheral preview
Longer in expert readers
Regression
Backward re-reading movement
Signals comprehension failure
Dyslexic readers: 2–3× more
Also used to re-check key info
3–4mm
Baseline diameter
+8–40%
Load-induced change
<200ms
Response latency
Kahneman 1973
Foundational study
Pupil dilation is one of the few physiological signals that cannot be voluntarily controlled — it is an honest signal of cognitive and emotional state that the person cannot fake or suppress.
15–20/min
Normal rate
3–8/min
During reading
25–40/min
Fatigue onset
PERCLOS
Drowsiness metric
Focused state
Blink rate drops to 3–8/min during focused reading or task engagement. The brain suppresses blinking to maintain visual continuity.
PERCLOS metric
Percentage of eyelid closure over 80% in a 1-minute window. The gold standard for drowsiness detection — now mandated in EU vehicles.
Together these four signals — fixation, saccade, pupil, and blink — form a complete window into attentional state, cognitive load, emotional arousal, and fatigue. No other non-invasive measure captures all four simultaneously.
03 · Core Mechanics
Smooth Pursuit
Smooth pursuit is the eye movement used to track a moving object — a ball, a car, a cursor. Unlike saccades it is continuous and velocity-matched. It is one of the most diagnostically valuable eye movements in neurology.
Smooth pursuit is the eye movement used to track a moving object — a ball, a passing car, a moving cursor. Unlike saccades, smooth pursuit is continuous and velocity-matched to the target. It only activates for moving targets.
Normal pursuit
Gain ~0.9–1.0 (matches target speed)
Smooth, continuous movement
Speed range: 1–100°/sec
Requires intentional tracking
Cannot track imaginary objects
Impaired pursuit
Gain <0.8 — eye falls behind
Catch-up saccades fill the gap
Seen in: schizophrenia, Parkinson's, cerebellar disorders
Degrades with age and fatigue
Diagnostic biomarker (SPEM test)
Smooth pursuit gain — the ratio of eye velocity to target velocity — is one of the most robust biomarkers for schizophrenia. Impaired SPEM (Smooth Pursuit Eye Movement) is present in 80% of schizophrenia patients and 40% of their unaffected relatives.
Schizophrenia
Reduced pursuit gain (<0.7) is present in 80% of patients. One of the most replicated biomarkers in psychiatric genetics — also present in unaffected relatives (endophenotype).
Parkinson's disease
Smooth pursuit speed is reduced and gain drops. Intrusive saccades break the smooth movement. Detectable in early-stage and pre-symptomatic individuals.
Cerebellar disorders
Cerebellar ataxia disrupts the gain adaptation mechanism — the system cannot learn to match target velocity. Manifests as consistent overshoot or undershoot.
Normal ageing
Pursuit gain declines gradually from age 50 onwards — catch-up saccades become more frequent. This is normal and does not indicate pathology unless decline is rapid.
Elite athletes show superior smooth pursuit performance — higher gain, earlier initiation, and faster velocity matching. This is trainable.
Elite athlete
Pursuit initiates 80–100ms before target moves (predictive)
Gain maintained at 0.95+ even at high speeds
Smooth pursuit combined with anticipatory saccades
Cricket, baseball, tennis: tracks ball to bat contact
Novice athlete
Pursuit is reactive — initiates after target moves
Gain drops above 30°/sec — loses the ball
More catch-up saccades, less predictive
Trainable with specific gaze training protocols
In cricket, the ball is physically too fast to track with smooth pursuit after pitching — batsmen use predictive saccades to place gaze ahead of the ball, then use brief pursuit for confirmation before shot execution.
Smooth pursuit gain — the ratio of eye velocity to target velocity — is one of the most robust biomarkers for schizophrenia, present in 80% of patients.
04 · Core Mechanics
Vergence & Depth
Vergence is the only eye movement where both eyes move in opposite directions simultaneously. Both eyes converge inward for near objects, diverge for far ones. The vergence-accommodation reflex coordinates this with focus.
Vergence is the only eye movement where both eyes move in opposite directions simultaneously. When converging, both eyes rotate inward to fixate a near object. When diverging, they rotate outward for far objects. This coordinates with accommodation (focus) — the vergence-accommodation reflex.
Viewing distance5/10
Far (2m+)Near (20cm)
The vergence-accommodation conflict is the primary cause of VR-induced eyestrain. Screens present objects at different vergence distances than focus distances — a mismatch the visual system finds fatiguing over time.
In VR headsets, the display is physically 2–3cm from the eyes but renders objects appearing at 0.5m–infinity. The eyes must converge for the rendered depth while accommodating at the fixed screen distance — a conflict that doesn't exist in natural vision.
Vergence-accommodation conflict
Accommodation (lens focus) and vergence (eye angle) are neurologically linked. In natural vision they always match. VR breaks this coupling — causing headaches, eyestrain, and nausea.
Varifocal displays
Next-generation VR headsets use eye tracking to measure vergence in real time and dynamically adjust the focal plane of the display to match — eliminating the conflict.
Cybersickness predictor
Users with higher vergence flexibility show lower cybersickness rates. Eye tracking vergence data predicts VR comfort within the first 2 minutes of use.
Gaze-contingent rendering
Eye tracking vergence + fixation used together to render full quality only where the eye is looking — reducing compute load by 60–70% without perceived quality loss.
Convergence insufficiency
Inability to maintain comfortable binocular vision at near distances. Presents as diplopia (double vision), headaches, and reading difficulty. Treatable with vergence training exercises.
Post-concussion syndrome
Convergence insufficiency is one of the most common sequelae of concussion — present in up to 50% of cases. Eye tracking vergence testing is used in concussion rehabilitation assessment.
Autism spectrum
Some ASD individuals show atypical vergence responses during face viewing — contributing to social gaze differences beyond the attentional avoidance typically reported.
Strabismus monitoring
Eye tracking provides objective, continuous measurement of eye alignment — used in strabismus treatment monitoring as an alternative to manual cover tests.
The vergence-accommodation conflict is the primary cause of VR-induced eyestrain — screens present objects at different vergence distances than focus distances.
05 · Core Mechanics
Anti-Saccade & Inhibitory Control
In the anti-saccade task, a stimulus appears and you must look away from it — in the opposite direction. This requires the prefrontal cortex to inhibit the automatic reflexive saccade. Error rate is one of the most clinically useful measures in eye tracking research.
In the anti-saccade task, a stimulus appears and you must look away from it — in the opposite direction. This requires the prefrontal cortex to inhibit the automatic reflexive saccade. Error rate on this task is one of the most clinically useful measures in eye tracking research.
What it measures
Inhibitory control — the ability to suppress a prepotent (automatic) response. Requires prefrontal cortex activation to override the reflex to look toward a new stimulus.
Error = inhibitory failure
An anti-saccade error — looking toward the stimulus instead of away — is not a mistake. It is the reflexive saccade breaking through inhibitory control.
Normal error rate
Healthy young adults: 10–20% error rate. This increases with age, sleep deprivation, cognitive load, alcohol, and multiple clinical conditions.
Correction latency
Even when an error occurs, healthy participants correct it within 200ms. Slow or absent correction is itself a diagnostic signal.
Alzheimer's disease
Error rate increases from ~15% in healthy ageing to 40–60% in early Alzheimer's. Error rate correlates with MMSE score (r = −0.72). One of the most sensitive early biomarkers.
Schizophrenia
Error rate typically 30–50% vs 15% in controls. Reflects prefrontal dysfunction and impaired executive control. Present in first-episode patients and unmedicated relatives.
ADHD
Elevated error rate and longer correction latency. Reflects impaired response inhibition — a core executive function deficit in ADHD. Used in diagnostic research and treatment monitoring.
Sleep deprivation
Error rate doubles after 24 hours of sleep deprivation. Anti-saccade performance is one of the most sensitive measures of sleep-loss-induced cognitive impairment.
Parkinson's disease
Elevated error rate plus hypometric corrective saccades. Combined anti-saccade + smooth pursuit battery distinguishes Parkinson's from healthy ageing with high sensitivity.
Bipolar disorder
Error rate is elevated during manic episodes and normalises during euthymia — making anti-saccade a potential state marker for mood episode monitoring.
The anti-saccade task is one of the most used paradigms in psychiatric eye tracking research precisely because it is fast (5 minutes), non-verbal, requires no reading or cultural knowledge, and has extremely well-characterised norms across age and clinical groups.
The anti-saccade error rate doubles after 24 hours of sleep deprivation and is elevated in Alzheimer's, schizophrenia, ADHD, and Parkinson's — making it one of the most clinically sensitive eye movement paradigms.
03 · Attention Systems
Bottom-up Attention
Bottom-up attention is automatic and stimulus-driven. Certain visual features capture gaze before any conscious decision to look — this happens in under 150ms, faster than you can think.
1Your visual system continuously scans the entire scene in parallel through peripheral vision
2Features — faces, contrast, motion, text — are detected automatically by low-level visual neurons
3These create a "saliency map" — a priority map of where to direct the next fixation
4The eye is sent to the highest-saliency location within 150–200ms — before any conscious decision
5This repeats 3–5 times per second. You experience it as "noticing something interesting"
Bottom-up vs Top-down: when you have a clear goal, top-down overrides bottom-up. Without a goal, your eyes follow the saliency map. Try Section 04 to see a task completely change where you look.
A face in any scene receives the first fixation — even when the task has nothing to do with people. This is hardwired from infancy and cannot be suppressed by instruction.
04 · Attention Systems
Top-down Attention
Top-down attention is goal-directed. Your task, expertise, and mental state rewire where you look on the exact same image. Yarbus (1967): identical stimulus, radically different gaze.
Same scene — four different tasks — four completely different gaze patterns.
Free viewing: fixations spread evenly — all regions receive attention.
Yarbus (1967): identical paintings produced radically different scan paths with different task instructions. This was the first proof that cognition directly shapes where we look.
Yarbus (1967) showed the same painting produced completely different scan paths with different task instructions — the first proof that cognition directly shapes gaze.
05 · Attention Systems
Inattentional Blindness
You can fixate directly on something and still not consciously perceive it. When cognitive load is high, the brain suppresses unexpected stimuli — even when the eyes land on them. Fixation ≠ perception.
Task loadMedium
Awareness probability
58%
Eyes fixed on it?
Yes
Classic result
~50%
Miss rate
42%
Safety: drivers can fixate a hazard and still fail to register it when cognitively loaded. Hands-free calls impair driving despite eyes remaining on the road.
Simons & Chabris (1999): ~50% of participants missed a gorilla walking through a scene while counting basketball passes — even though their eyes fixated it.
06 · Attention Systems
Selective Attention
Selective attention is the mechanism by which the brain chooses what to process deeply and what to suppress — the visual equivalent of the cocktail party effect. Eye tracking makes this selection visible.
Attended stream
Stream A
Suppression
Active
Switch cost
~250ms
Theory
Broadbent 1958
Early selection (Broadbent): attended channel processed fully, others filtered. Late selection (Deutsch): all processed, selection at awareness stage. Eye tracking supports both.
07 · Attention Systems
Change Blindness
Change blindness is the failure to notice changes to a visual scene — even large ones — when they occur during a saccade, blink, or brief visual interruption. Eye tracking reveals exactly what was fixated when the change happened.
Detection rate
~50%
Change size
Large
Saccade blind?
Yes
Simons & Levin
1998
Simons & Levin (1998): people failed to notice when a conversation partner was replaced by a different person during a brief visual interruption — change blindness extends to real-world events.
08 · Attention Systems
Visual Search
Visual search reveals how the eye finds a target among distractors. Eye tracking exposes search strategies, efficiency, and error types. Expert searchers show faster TTFF and fewer false fixations than novices.
Fixations to find
8–12
Search time
2.4s
Miss rate
18%
Strategy
Serial
Pop-out search (unique colour/shape) is parallel — all distractors are processed simultaneously and the target jumps out. Conjunction search (colour AND shape) is serial — each item is checked one by one.
09 · Cognitive Processes
Cognitive Load
The pupil dilates reliably with mental effort, independent of lighting — a phenomenon known as pupillometry, established by Kahneman in the 1960s. Drag difficulty to see how pupil size and gaze behaviour change.
Task difficulty3 / 10
Pupil diameter
+18%
Fixation scatter
Low
Dwell per AOI
Long
Blink rate
Normal
At extreme cognitive load (9–10/10), fixations shorten and scatter — the system is overwhelmed. This is the eye tracking signature of overload, distinct from deep engagement.
10 · Cognitive Processes
Working Memory
Working memory load is directly visible in gaze patterns. As working memory fills, fixation patterns fragment, re-fixation rates increase, and dwell time on new information drops. Eye tracking measures capacity without verbal report.
WM items
2 / 4
Regression rate
Low
Dwell time
Long
New info proc.
Full
When working memory is full, gaze becomes stuck — participants re-fixate already-seen information rather than processing new items. This fixation regression is a reliable indicator of capacity saturation.
11 · Cognitive Processes
Mental Fatigue
Mental fatigue is detectable in gaze long before subjective reports. Fixation stability degrades, saccade velocity drops, blink rate increases, and microsaccades become more frequent. Used in pilot monitoring and driver safety.
Hours awake
8h
Fixation stability
High
Blink rate
15/min
PERCLOS
<5%
After 17 hours awake, eye movement impairment equals 0.05% BAC. After 24 hours, 0.10% — above legal driving limits in most countries. Eye tracking detects this impairment before the person does.
12 · Cognitive Processes
Decision Making
Eye tracking reveals decision making in real time. People fixate more on options under active consideration. The pattern of fixation shifts predicts choice — sometimes before the person consciously knows what they will decide.
Decision time
0.8s
Pre-choice fixation
Option A
Fixation switching
Low
Gaze cascade
Active
The gaze cascade effect (Shimojo et al., 2003): in the final 1–2 seconds before a preference choice, fixation drifts toward the chosen option. This drift may actually cause the preference, not just reflect it.
13 · Cognitive Processes
Mind Wandering
Mind wandering — when attention drifts to unrelated thoughts — is detectable in gaze data. Fixations become unfocused, pupil diameter drops, blink rate increases. Used in e-learning to detect when learners have mentally checked out.
Comprehension
High
Re-fixation rate
Low
Pupil size
Normal
Detection
47% waking
People mind wander approximately 47% of waking hours (Killingsworth & Gilbert, 2010). During reading, mind wandering produces longer fixations and fewer regressions — readers look without processing.
14 · Cognitive Processes
Anticipatory Gaze
Experts do not look at where things are — they look at where things will be. Anticipatory gaze reflects predictive cognition: the eye arrives at a target location before the event occurs. This distinguishes expert from novice across many domains.
Gaze lead time
300ms
Fixation target
Wrist
Expert advantage
vs novice
Prediction accuracy
87%
Elite cricket batsmen fixate the bowler's wrist 300ms before ball release — not the ball. This anticipatory gaze allows shot preparation to begin before the ball is bowled.
15 · Emotional & Social
The Happiness Effect
Fredrickson's Broaden-and-Build Theory (2001): positive affect literally widens the attentional spotlight. Anxiety narrows it to the perceived threat. This is measurable in fixation data — drag the mood slider to see it.
Teams in a positive emotional climate process more information and generate more creative solutions — measurable in group eye tracking studies as broader collective scan paths.
16 · Emotional & Social
Memory & Gaze
Where you have looked before influences what you remember. Items that receive longer fixations are better encoded. Eye tracking reveals the intimate connection between visual attention and memory consolidation.
Avg recall
55%
Effect of dwell
~7%/100ms
Re-fixation bonus
+15%
Gaze avoidance
Detectable
Each additional 100ms of fixation increases recall probability by approximately 5–8% (Rayner et al., 2009). Re-fixations (looking back at an item) show even stronger memory encoding than single long fixations.
17 · Emotional & Social
Social Gaze
Social gaze — where we look during social interaction — is one of the most information-rich channels of human communication. We follow each other's gaze automatically, and this joint attention is foundational to social cognition and language development.
Gaze contact
50–60%
Following latency
100ms
Joint fixation
Synchronised
Social signal
Strong
Gaze following is present in 6-month-old infants and remains automatic in adults. When someone looks at an object, observers reflexively shift gaze to the same location — even when the person is looking offscreen.
18 · Emotional & Social
Emotion Recognition
The eye region is fixated most intensely during emotion recognition. We extract emotional meaning from eyes first, then the mouth. Eye tracking reveals how clinical groups and cultures process emotional faces differently.
Eye region fix.
68%
Mouth region
18%
TTFF (eyes)
180ms
Accuracy
94%
People with reduced fixation on the eye region — a pattern seen in autism — show impaired emotion recognition even on controlled laboratory tasks, independent of general intelligence.
19 · Emotional & Social
Trust & Deception
Gaze patterns change during deception and trust violations. Eye tracking reveals the cognitive load of constructing false information — though it is not a reliable lie detector.
Eye contact
Normal
Blink rate
15/min
Pupil dilation
Baseline
Cognitive load
Low
Contrary to popular belief, liars do not consistently avert gaze. Trained deceivers often increase eye contact during deception. Pupil dilation and blink suppression are more reliable indicators than gaze direction alone.
20 · Emotional & Social
Cross-cultural Gaze
People from different cultures process visual scenes differently. East Asian viewers distribute fixations more broadly across scenes while Western viewers fixate central objects more intensely. These differences are detectable within the first 500ms of viewing.
Background fix.
25% vs 65%
Object fix.
75% vs 35%
Cultural diff.
p<0.001
Masuda & Nisbett
2001
Masuda & Nisbett (2001): East Asian participants fixated background elements 60% more than Western participants on the same scenes — reflecting deeper differences in holistic vs analytic cognition.
21 · Learning & Education
Expertise & Training
Expert gaze is not innate — it is trained. Years of experience build attentional schemas that guide the eye directly to diagnostically important regions. This gap between novice and expert is measurable, and eye tracking can accelerate its development.
Expert fixations
4–5
Novice fixations
10–12
Expert avg dwell
380ms
TTFF key zone
0.5s
Chess grandmasters fixate 4–5 key squares; beginners fixate 10–12. Radiologists reach critical diagnostic zones within 500ms — novices take 3–4 seconds and often miss them entirely.
22 · Learning & Education
Reading Patterns
Web readers do not read word-by-word. Eye tracking reveals predictable patterns depending on page layout, content type, and user goal. Each pattern tells a designer something different about how content is processed.
Pattern type
F-pattern
Content read
~28%
Fixation density
Left-heavy
Context
News / blog
The F-pattern emerges on poorly structured pages — users read the first line fully, scan a second partial line, then scan just the left margin. Good formatting with headings produces layer-cake or spotted patterns instead.
23 · Learning & Education
Multimedia Learning
In multimedia learning, how the eye moves between text and images determines how well material is integrated. The split-attention effect is directly visible in gaze switching patterns between separated text and diagrams.
Fixation switches
Low
Comprehension
High
Cognitive load
Low
Mayer principle
Contiguity
Students with integrated text-diagram designs show 40% fewer cross-element fixation switches than split designs — each switch is cognitive cost as the learner rebuilds the mental model.
24 · Learning & Education
Collaborative Learning
Joint attention — when two people fixate the same point simultaneously — predicts successful knowledge construction in collaborative tasks. Dual eye tracking reveals whether partners are truly co-attending or processing independently.
Joint fixation
68%
Prediction
Learning
Sync. rate
High
Dual tracking
2-person
Studies using dual eye tracking show that periods of joint fixation — both people looking at the same thing — predict learning outcome independently of verbal interaction.
25 · Learning & Education
Test Anxiety
Under exam pressure, fixation patterns reveal anxiety that self-report cannot. Test-anxious students show more re-reading of questions, longer dwell on incorrect answer options, and more fixations near time limits.
Re-reads
2.1×
Answer switching
8%
Question dwell
2.1s
Correct answers
88%
Test-anxious students spend 35% more time re-reading question stems without changing their answer — a gaze signature of confidence failure distinct from genuine content difficulty.
26 · Learning & Education
Handwriting & Motor Learning
During handwriting acquisition, novices fixate the pen tip — monitoring what they are doing now. Experts fixate ahead of the pen — guiding future movements. This anticipatory shift is a marker of motor learning completion.
Gaze position
On pen tip
Lead distance
0 chars
Writing fluency
Low
Motor planning
Reactive
The transition from pen-tip to ahead-of-pen fixation occurs at a consistent point in handwriting acquisition and predicts future writing fluency more reliably than accuracy measures alone.
27 · Sport & Performance
Quiet Eye
The quiet eye is the final fixation before a motor action — a golf putt, a basketball free throw, a rifle shot. Its duration predicts outcome. Longer quiet eye = better performance. One of the most replicated findings in sport psychology.
Quiet eye dur.
800ms
Onset before act.
600ms
Location
Target
Outcome
Made
Elite golfers maintain quiet eye for 2–3× longer than near-elite players. Under pressure, quiet eye shortens — and this shortening predicts performance decrements before they occur.
28 · Sport & Performance
Performance Under Pressure
Under competitive pressure, quiet eye duration shrinks and gaze becomes more erratic — explaining performance degradation in high-stakes situations (choking). Eye tracking is used in elite training to restore quiet eye under simulated pressure.
Quiet eye dur.
600ms
GK fixation
18%
Outcome
Scored
Pressure effect
None
Penalty shootout research: players who scored showed longer quiet eye and fixated the goalkeeper for less time. Players who missed showed shorter quiet eye and more goalkeeper fixation — attending to the wrong stimulus under pressure.
29 · Sport & Performance
Tactical Decision Making
In team sports, tactical decision quality depends on how broadly and early players scan the field before receiving a pass. Eye tracking reveals gaze behaviour in the critical window before the ball arrives.
Pre-receive scans
4–6
Scan lead time
0.8s
Pass accuracy
87%
Decision time
0.4s
Elite footballers scan the field 0.5–1.0 seconds before receiving the ball — amateurs look only after receipt. This pre-receive scanning is the single strongest predictor of subsequent pass quality.
30 · Sport & Performance
Surgical Skill
Surgical trainees show scattered, reactive fixation patterns — responding to events rather than anticipating them. Expert surgeons fixate ahead of the instrument and allocate attention efficiently. Eye tracking provides objective skill metrics for training certification.
Field fixation
70%
Instrument fix.
12%
Anticipatory
Yes
Error rate
Low
Expert surgeons fixate the operative field for 70% of procedure time; trainees fixate their instrument for 55% — monitoring what they are doing rather than where they are going.
31 · HCI & Technology
Gaze-based Interfaces
Eye tracking is increasingly used as an input modality — selecting, scrolling, typing, and navigating by gaze alone. Critical for accessibility (ALS, locked-in syndrome) and emerging in consumer AR/VR interfaces.
WPM (gaze type)
15–25
Dwell threshold
500ms
Error rate
8%
User group
ALS/motor
Stephen Hawking communicated using a gaze interface for the final decade of his life. Modern AAC devices allow non-verbal users to produce 15–25 words per minute using eye tracking alone.
32 · HCI & Technology
Attention-aware Computing
Attention-aware computing systems detect what users are attending to in real time and adapt accordingly — pausing videos when gaze leaves the screen, adjusting complexity based on pupil-measured load.
Gaze on content
94%
System response
Active
Notifications
Suppressed
Adaptation
None needed
Gaze-contingent systems — where display content changes based on where the user is looking — reduce reading time by 15–25% and comprehension errors by 30% compared to standard display conditions.
33 · HCI & Technology
Coding & Programming
Programmers fixate differently on familiar vs unfamiliar code. Bug-fixing shows a distinct search-and-verify fixation pattern. Eye tracking is used to study how developers read code and design better IDEs and code review tools.
Identifier focus
60%
Syntax focus
15%
Comprehension
High
Scan pattern
Structured
Expert programmers spend 60% of code-reading time on identifiers (variable and function names); novices spend equal time on both identifiers and syntax. Meaningful naming is cognitive efficiency, measurable in gaze.
34 · Applied Research
AOI Metrics
Areas of Interest (AOIs) are defined regions analysed for four core metrics. Each metric answers a different question about how attention interacts with a design element. Select an AOI zone to see its values and interpretation.
Sign fixation rate
40%
Landmark fixation
78%
TTFF (exit sign)
3.8s
Floor fixation
22%
Hospital wayfinding studies show 60% of directional signage goes unfixated by first-time visitors, who instead fixate landmarks, people, and floor features. Signs placed at decision points receive 3× more fixations.
35 · HCI & Technology
UX & Advertising
Eye tracking is foundational to UX research and advertising effectiveness testing. It reveals what users actually look at — versus what they say they look at — and measures attention to interface elements and ads in real time.
Banner fixation
4%
CTA fixation
62%
Ad recall
8%
Content read
~35%
Banner blindness: users have learned to ignore rectangular display ads — only 4% of page viewers fixate banner ad zones. Above-the-fold CTA buttons receive 10× more fixations than identical buttons placed below the fold.
36 · Applied Research
Wayfinding & Architecture
Eye tracking reveals how people navigate built environments — what signage they notice, what architectural features guide their path, and how experts and novices differ in spatial search strategies.
Sign fixation
Low
Decision points
4
Path efficiency
Medium
Landmark fixation
High
First-time visitors in large buildings fixate architectural landmarks and contrast features first — signage is only sought when landmarks fail. Effective wayfinding design uses both visual landmarks and well-placed signs at decision points.
37 · Applied Research
Museum & Gallery
Museum and gallery visitor studies use eye tracking to reveal which artworks receive attention, for how long, in what sequence, and whether label text is read. Used to optimise layout, label placement, and visitor experience.
Avg dwell/artwork
12s
Label readership
18%
TTFF first artwork
0.4s
Total visit time
3.2min
The average museum visitor spends less than 15 seconds in front of any single artwork. Fixations concentrate on faces and high-contrast areas — most of the painted surface is never fixated.
38 · Applied Research
Virtual Reality
In VR, gaze naturalness — whether a person looks around as they would in real life — is a key measure of presence and immersion. Abnormal gaze patterns indicate the environment is not convincing. Eye tracking also enables foveated rendering.
Gaze naturalness
94%
Presence score
8.2/10
GPU saving
65%
Sickness onset
None
Foveated rendering — rendering high resolution only where the eye is currently fixating — reduces GPU load by 60–70% without perceptible quality loss. Eye tracking makes this possible in real time.
39 · Clinical & Neuroscience
Clinical Signals
Gaze patterns are increasingly used as objective biomarkers for neurological and psychiatric conditions. Select a condition to see its characteristic eye movement signature — what it looks like in data, and how it is being applied clinically.
ADHD: highly unstable fixations — shorter dwell, more intrusive saccades to non-relevant areas, difficulty sustaining fixation on target regions. High fixation count with low efficiency is a key marker.
Fixation Stability
Low
Face Fixation
Reduced
Saccade Quality
Irregular
Diagnostic Use
Research
40 · Clinical & Neuroscience
Alzheimer's Detection
Eye movements change years before cognitive symptoms of Alzheimer's appear. Smooth pursuit gain drops, saccadic latency increases, and fixation stability degrades. Eye tracking is being developed as an early, non-invasive biomarker for neurodegeneration.
Pursuit gain
0.92
Saccade latency
180ms
Fixation stability
High
AS error rate
12%
Eye tracking detects Alzheimer's-related changes 5–10 years before cognitive symptoms. The combination of reduced smooth pursuit gain, increased anti-saccade errors, and degraded fixation stability forms a sensitive biomarker profile.
41 · Clinical & Neuroscience
Dyslexia
Dyslexic readers show longer fixations, more regressions (backward refixations), and shorter saccade lengths during reading. These patterns are visible even on single words — distinct from general comprehension difficulty.
Fixation duration
200ms
Regressions/line
1.2
Saccade length
8 chars
Reading rate
280 wpm
Dyslexic readers make 2–3× more regressions per line of text than typical readers. This reflects decoding failure forcing re-processing — not a cause of reading difficulty but an objective symptom.
42 · Clinical & Neuroscience
Sleep Deprivation
Sleep deprivation degrades eye movement control before any subjective feeling of impairment. Fixation stability drops, pupil becomes sluggish, microsaccade frequency increases, and anti-saccade error rate rises.
Hours awake
8h
Fixation stability
High
AS error rate
15%
BAC equivalent
Baseline
After 17 hours awake, eye movement impairment equals 0.05% BAC. After 24 hours, 0.10% — above the legal driving limit in most countries. Eye tracking detects this impairment before the person does.
43 · Medical
Radiology Training
Where did the radiologist look before missing a finding? Eye tracking reconstructs the visual search and distinguishes between search errors (never looked there) and recognition errors (looked but failed to identify).
Finding TTFF
0.5s
Search errors
5%
Coverage
94%
Decision time
45s
60% of missed radiological diagnoses involve a search error — the abnormality was never fixated. 40% are recognition errors — fixated but not identified. Eye tracking distinguishes these two failure modes precisely.
44 · Medical
Nursing & Triage
Experienced nurses detect patient deterioration faster because they fixate vital signs and clinical cues earlier and more efficiently than trainees. Eye tracking has been used to study clinical intuition and design better patient monitoring environments.
Patient face fix.
40%
Monitor fixation
28%
TTFF deterioration
1.2s
Early detection
86%
Expert nurses spend 40% of monitoring time fixating the patient's face and breathing pattern — trainees spend the same time fixating equipment. Clinical intuition is partly a gaze strategy difference.
