mousePrep Reference

These are the main functions in the mousePrep package for processing mouse-tracking data including data standardization, analysis, outlier detection and aggregation. The functions are organized into categories based on their purpose, but they can be used in any order except when specified otherwise. Each function includes parameters for customization and returns a modified dataset along with an audit log of changes made.

Depending on the file format, one of the standard R functions for reading files into R can be used such as read.table or read.csv After the data is loaded, these functions can be used to clean, preprocess, and engineer features from the raw mouse-tracking data.

Data Standardization

Column Standardization

standardize_cols() — Standardize names/types/units
convert_numeric() — Convert quantitative columns to numeric type

Case filtering

rm_cases() — Drop invalid participants/screens/empty trajectories
rm_cases_time() — Drop cases or participants based on time spent on a question
rm_touch_devices() — Remove touch devices by rule
mouse_class() — Drop cases based on the number of recorded mouse or touch movement data points per screen
mouse_class_col() — Flag cases based on the number of recorded mouse or touch movement data points per screen
rm_non_mouse_events() — Remove non-mouse movement events
rm_resize_events() — Detect and remove resize events. Option to remove trajectory or participant

Feature Engineering

Screen Geometry

screen_dims_calculate() — Compute screen width/height from inner+scroll

Timing & alignment

time_vars_calculate() — Compute Initiation time, response time and move time from timestamps of trajectories

Outlier detection based on timing features

After timing features are computed, outliers can be detected and handled using the following functions.

flag_outliers() — Flag outliers. Option to flag based on data-driven or fixed thresholds
rm_outliers() — Remove outliers based on the flagged observations

Repeated visits for trajectories and questions

multiple_traj() — Flag first, last or longest trajectory when there are multiple trajectories per screen

Merging Other Functional Data

include_clicks() — Merge click time/location with trajectories

Processing via mousetrap

mp_processing_mt() — Use analysis functions in mousetrap package to get to the final dataset

Export & reporting

mp_export_dataset() — Export cleaned dataset in long or wide formats