Smooth pursuit is a voluntary eye movement that allows a viewer to keep their gaze fixated on a moving object. Also known as smooth pursuit eye movement (SPEM), this process involves matching eye velocity to the velocity of a target to reduce retinal blur. For marketers and UX researchers, understanding this mechanism is essential when analyzing how users interact with moving content via eye-tracking technology.
What is Smooth Pursuit?
Smooth pursuit is one of two primary ways humans voluntarily shift their gaze, the other being saccadic eye movement. While saccades are rapid, jerky jumps between points, smooth pursuit is a continuous, slow movement. Its main purpose is to maintain a sharp image of a moving target on the fovea, the part of the eye responsible for central, high-resolution vision.
Most individuals cannot initiate smooth pursuit without a moving visual signal. Attempting to move the eyes smoothly across a static scene typically results in a series of saccades rather than a fluid motion.
Why Smooth Pursuit Matters in Research
For professionals using eye-tracking tools to study user behavior, smooth pursuit provides data on how effectively a user can follow a stimulus.
- Attention Tracking: It confirms that a user is actively "locked on" to a specific moving element, such as a video ad or a sliding banner.
- Measurement Reliability: High-quality pursuit indicates that the visual system is accurately processing motion, which is crucial for valid UX testing results.
- Cognitive Insight: Because smooth pursuit reflects mental states, researchers use it as a "window to the mind" to see how users anticipate movement.
- User Impairment Detection: Deficits in pursuit can signal that a test subject is fatigued, aged, or under the influence of substances, which might skew data.
How Smooth Pursuit Works
The pursuit system functions through a continuous feedback loop that aims to eliminate "retinal slip," which is the movement of an image across the retina.
- Open-Loop Phase: This is the initial response to a moving object. It is "ballistic," meaning the visual system hasn't had time to correct the movement yet. [The eye movement response starts only 100 to 150 ms after motion onset] (ScienceDirect).
- Closed-Loop Phase: This stage lasts until the movement stops. Here, the brain uses ongoing visual feedback to match the eye's angular velocity to the target's velocity.
- Catch-up Saccades: If the target moves too fast, the eyes fall behind. The brain then triggers a rapid "catch-up" jump to re-foveate the object. [Eye tracking ability begins to deteriorate above velocities of 30 degrees per second] (ScienceDirect).
Best Practices for Recording Eye Movements
When conducting eye-tracking studies for website or app performance, follow these standards to ensure data accuracy.
- Select the right hardware: Use a search coil for extreme medical precision or an infrared eye tracker for a non-invasive, practical UX setup.
- Control target velocity: Ensure moving stimuli do not exceed 30°/s. Beyond this speed, you will record catch-up saccades rather than smooth pursuit.
- Hold the head stationary: Use a chin rest or bite board when possible. While some modern trackers handle head movement, the highest precision occurs when the head is fixed.
- Account for age effects: Be aware that older participants are more likely to struggle with pursuit tasks; allow them two or three practice trials to acclimate to the stimulus.
- Check for asymmetry: Most humans are naturally [better at horizontal than vertical smooth pursuit] (Wikipedia). Take this into account when designing horizontal vs. vertical scrolling tests.
Common Mistakes in Tracking Analysis
Mistake: Treating all eye movements over a video as "pursuit." Fix: Use software to filter out saccades. Saccades have much higher initial acceleration and peak velocity than smooth pursuit.
Mistake: Assuming a lag in tracking is always a technical hardware error. Fix: Check the participant's profile. Preterm infants, people with schizophrenia, or those with PTSD may show naturally [reduced velocity of smooth pursuit during ongoing tracking] (Wikipedia).
Mistake: Running tests in total darkness without a clear visual target. Fix: Ensure the stimulus is bright and distinct. While pursuit is possible via touch or following one's own hand in the dark, it is highly inefficient for standard UX testing.
Smooth Pursuit vs. Saccades
| Feature | Smooth Pursuit | Saccades |
|---|---|---|
| Primary Goal | Match target velocity | Match target position |
| Stimulus | Retinal velocity error (slip) | Retinal position error |
| Movement Type | Continuous and slow | Rapid and jumpy |
| Voluntary? | Yes (but requires a target) | Yes |
| Latency | [90 to 150 ms] (Wikipedia) | 200 to 250 ms |
FAQ
Can humans perform smooth pursuit without a moving object? Generally, no. Most people cannot move their eyes smoothly without a visual signal or a proprioceptive signal (like following their own finger in the dark). Attempting to do so without a target results in saccades. However, [infants can attain adult levels of smooth pursuit by 14 weeks of age] (ScienceDirect) for predictably moving targets.
How does alcohol affect smooth pursuit? Alcohol and other depressants cause a "Lack of Smooth Pursuit." This is a scorable clue used in standard field sobriety tests. It prevents the eyes from following a stimulus smoothly, causing them to "jerk" or fall behind.
What is a catch-up saccade? A catch-up saccade is a quick eye jump that occurs when the smooth pursuit system fails to keep up with a target. This typically happens when a target moves faster than 30°/s or if the viewer's pursuit system is impaired.
Is smooth pursuit identical to the Optokinetic Nystagmus (OKN)? No. OKN is an automatic, reflexive response to large-scale motion (like looking out a moving train window). Smooth pursuit is the voluntary tracking of a specific, discrete object.
Can you improve smooth pursuit through practice? Experimental evidence shows "Smooth Pursuit Adaptation" is possible. If a participant is repeatedly exposed to targets that change velocity predictably, they can learn to [upregulate or downregulate their pursuit gain] (ScienceDirect) to minimize errors.
