annavanharmelen/microsaccade-bias-shift-vs-maintain-analysis

Purpose and execution model

MATLAB analysis repository for a microsaccade-bias experiment. It processes behavioral and eye-tracking recordings produced by the separate Python experiment repository. There is no single application entry point; analysts run MATLAB scripts in a prescribed sequence and edit participant/data-location mapping as needed.

Subject and data-location boundary

getSubjParam.m is the central subject/session configuration boundary: it maps randomized participant/session identifiers to the repository owner’s data organization. Analysis scripts depend on this mapping rather than defining an independent data-ingestion layer.

Relevant paths:

• getSubjParam.m

• setBehaviourParam.m

• getBehaviour.m

Behavioral analysis pipeline

getBehaviour.m is the main behavioral-analysis entry point. It uses behavioral parameter setup and produces statistics/visualization through supporting scripts, separating behavioral results from the eye-tracking pipeline.

Relevant paths:

• getBehaviour.m

• setBehaviourParam.m

• behaviour_stats.m

Eye-data epoching stage

epochEyeData.m is the first required eye-tracking stage. It converts/segments raw eye-tracking data into analysis epochs that downstream gaze-position and saccade computations consume. This is the principal transition from acquired recordings to trial/epoch-level analysis data.

Relevant paths:

• epochEyeData.m

• frevede_nanBlinks_1eye.m

• frevede_pixel2dva.m

Participant-level gaze-position bias

get_GazePositionBias.m computes average gaze bias for each participant after epoching. Supporting transforms convert gaze position into one- or two-dimensional shift/bias representations and provide eye-data summaries.

Relevant paths:

• get_GazePositionBias.m

• PBlab_gazepos2shift_1D.m

• PBlab_gazepos2shift_2D.m

Participant-level saccade bias

get_SaccadeBias.m derives per-participant saccade bias from the epoched eye data. It is parallel to gaze-position analysis but represents a distinct metric and output stream.

Relevant paths:

• get_SaccadeBias.m

• Eye_stats.m

• frevede_allbyall_correlations_new.m

Group-level aggregation and inference

Grand-average scripts consume participant-level bias outputs. Their stated design weights every participant equally rather than weighting by completed trial count. Statistical support includes correlation and 1-D cluster-statistics utilities, with plotting helpers for group-result presentation.

Relevant paths:

• GA_GazePositionBias.m

• GA_SaccadeBias.m

• frevede_ftclusterstat1D.m

Runtime and external dependencies

The runtime is MATLAB. Several eye-tracking and statistical stages depend on the FieldTrip MATLAB toolbox, originally developed against FieldTrip version 2020.10.23; FieldTrip likely provides core statistical/data-structure functionality for functions such as frevede_ftclusterstat1D.m. Input data and experimental semantics originate from the external Microsaccade-bias Python experiment repository.

Relevant paths:

• frevede_ftclusterstat1D.m

• frevede_errorbarplot.m

• create_colour_map.m