01 — From movement to signal
The trajectory complements the information provided by the final click.
A conventional analysis retains the final choice and response time. Pointer tracking records the sequence of horizontal and vertical coordinates with their timestamps. It reveals whether movement starts immediately, slows down, returns toward a competing option or remains still before validation.
Raw data must be cleaned and segmented into movements and pauses. This methodological choice matters: published thresholds range from roughly 100 to 500 ms and sometimes longer. The meaning of a pause therefore depends on the selected threshold, task and interaction context.
02 — Features
Hesitation unfolds in space and time.
Temporal features include initiation latency, speed, acceleration, pause duration and time before click. Spatial features describe traveled distance, straightness, curvature, reversals and direction changes. Contextual information adds answer positions or item difficulty.
Cepeda et al. used cross-validated logistic regression to distinguish certain from uncertain answers. Table 6 reports 89% accuracy and an F1 score of 0.81, combining precision and recall. Traveled distance, deviation and click latency contribute to prediction. These results concern the studied questionnaire and require new validation for another journey.
03 — Interpretation
Connect one trajectory to several possible explanations.
Movements may reflect competition between possible responses. Freeman and Ambady showed that the hand could be temporarily attracted toward a competing category before the final choice. Other work links less efficient movement and longer pauses to cognitive differences in older adults.
The same patterns may also result from distant targets, a trackpad, fatigue, motor ability, animation or an ambiguous layout. Robust interpretation therefore requires a controlled task, meaningful comparisons and preferably several converging signals.
04 — Value for user research
Turn a trace into a testable hypothesis.
In user experience (UX) research, these metrics locate moments where an action loses fluency: options that attract then repel, areas explored without commitment, repeated corrections or unusual delay before clicking.
Those cues can prioritize user testing or qualitative inquiry. Interviews and interface analysis then help explain the observed friction. Collection should always be proportionate, consented, minimized and documented.
05 — From concept to decision
Compare journeys, then investigate what explains the detour.
On a form, I would compare movements around the same choice before and after clarifying its label. The protocol would keep target sizes and positions comparable, separate mouse and trackpad use, and measure corrections alongside final errors.
If detours decrease, an interview or commented replay would help distinguish better comprehension from a motor change. The trace selects the moment to investigate; it does not supply the explanation on its own.