Interactive method selector¶
Select the observed data support, scientific target, clustering and uncertainty claim. This is a navigation aid, not a statistical estimator. Results indicate a suitable starting point, not an automatic or unique correct model.
Default: aligned common grid, gaze variation.
Interpret the fitted components against the recorded units and design.
Read the method and exampleScientific limitations are not optional
A recommended route does not imply calibrated confidence intervals, a causal effect, or identification of an outcome-dependent missing-gaze mechanism. Whole-function W3 inference and field MNAR interpretation remain restricted by issues #289 and #296. For unusual combinations (e.g. repeated binary trajectories on asynchronous coordinates), return to the method comparison rather than forcing a single estimator.
Next: Input schemas · Workflow comparison · Runnable tutorials