TrajTracker Analyze expects the data format exported from TrajTracker Experiment paradigms.

It will preprocess the data, including:​

• Convert the data into easy-to-use matlab structures.

• Resample the trajectory data into a fixed sampling rate, for easier analysis

• Smooth the trajectory data

• Calculate several measures per time point along the trajectory: direction, implied endpoints, speed, acceleration.

• A set of perl-based scripts to easily handle the raw data files (filter, copy/move)

• ... and more

TrajTracker Analyze supports several kinds of plots for basic visualization of the data, e.g.:​

• Average-trajectories per target number: geometrical plots (x coord by y coord), temporal plots (x coordinate by time), momentary-finger-direction plots.

• Single-trial raw trajectories

• Show any data per time point

TrajTracker Analyze has a powerful mechanism for analyzing the data with regressions.​​ 

• Regressions can be run per trial or per time-point within the trajectory.

• Extremely easy definition of new regression models.

  • The dependent variable and the predictors can be set to a large variety of predefined measures - per-trial measures (e.g., the target number, the movement time) and within-trajectory measures (e.g., momentary coordinates or speed).

  • User-defined measures can be easily plugged into the regression model

• Several group-level analyses of regression results (significance of predictors, comparison between predictors)

• Highly configurable plots of the regression results.

For example,

• Look for acceleration/deceleration points to detect changes of mind.

• Calculate the first-finger-deviation time as an index for the time when the stimulus information became available.

• Parse each trajectory into a sequence of clockwise and counter-clockwise curves, and extract several types of information about these curves