The collaborative effort of Pedro Lencastre, Samip Bhurtel, Anis Yazidi, Gustavo B. M. e Mello, Sergiy Denysov, and Pedro G. Lind has yielded the EyeT4Empathy dataset, encapsulating a diverse array of eye movement recordings and empathy evaluations from 60 participants. This comprehensive compilation intersects visual information processing, gaze typing, and empathy assessment.

Participants undertook two distinct tasks: one involving free-form exploration of structureless images and the other utilizing gaze typing exercises through an eye-gaze interface. These recordings, partitioned into two datasets, not only capture gaze positions but also integrate measurements of pupil diameter.

Additionally, this dataset delves into empathy evaluation, particularly focusing on individuals with movement impairments. Participants underwent empathy assessment twice via a comprehensive questionnaire comprising forty questions, expanding the assessment scope to encompass cognitive and affective empathy.

The EyeT4Empathy dataset provides an extensive avenue for multifaceted analysis, shedding light on statistical aspects of eye-gaze trajectories during free-viewing tasks and exploring how empathy manifests through eye-related metrics. This resource, accessible via Scientific Data, empowers researchers and practitioners to delve into the intricate relationships between visual cognition, eye movements, and empathetic responses.

Dataset Details:

• Measurements:

  • Gaze Position

• Technology Type:

  • Eye Tracking Device

• Factor Types:

  • Luminosity
  • Screen Distance

• Sample Characteristics:

  • Organism: Human Eye
  • Environment: Laboratory Facility
  • Location: Municipality of Oslo

Link to the Paper

Additional Dataset Sources:

1. Eye Tracker Data

   • Source: Figshare (2022)
   • Contains 504 CSV files within the provided zip file.

2. Raw Data

   • Source: Figshare (2022)
   • Includes 60 TSV files.

3. Questionnaires

   • Source: Figshare (2022)
   • Consists of questionnaire files in both PDF and another format, aiding in comprehensive understanding.

The dataset, accessible on FigShare under a Creative Commons license, encompasses 502 files documenting eye-tracking movements. Each file follows a naming convention that signifies various experimental parameters, including the type of experiment, stimuli, participant IDs, and trial numbers.

Additionally, raw eye-tracker data extracted from the device is named following a structured pattern, aiding in participant identification based on experiment type and participant numbers.

For a deeper understanding of the dataset columns, accompanying documents titled columns_explained.pdf and coordinate_system.pdf provide comprehensive insights into column details and coordinate systems utilized within the dataset.

Certainly! To provide a concise and clear understanding, I'll outline the column descriptions for each dataset within the EyeT4Empathy compilation.

Eye Tracker Data Columns Description:

The columns within each CSV file of the Eye Tracker Data contain the following information:

1. Timestamps:

   • Recording timestamp, computer timestamp, eyetracker timestamp.

2. Recording Details:

   • Project name, export date, participant name, recording name, recording date, start time, duration, resolution, software version, monitor latency.

3. Calibration and Validation Metrics:

   • Calibration accuracy, precision, and validation metrics measured in millimeters, degrees, and pixels.

4. Gaze Points and Eye Movements:

   • Coordinates for gaze points (X, Y), eye positions, pupil diameters, gaze directions, fixation points, and movement types.

5. Stimulus Information:

   • Details of presented stimuli and media, including names, dimensions, positions, types, and original media sizes.

6. Other Information:

   • Mouse positions, client area positions, viewport positions and sizes, full page dimensions, and event-related data.

Raw Data Columns Description:

The columns within each TSV file of the Raw Data include the following information:

1. Timestamps and Recording Details:

   • Recording timestamp, computer timestamp, eyetracker timestamp, project name, export date, recording name, recording date, start time, duration, resolution, software version, monitor latency.

2. Calibration and Validation Metrics:

   • Various metrics for calibration accuracy, precision, and validation in millimeters, degrees, and pixels.

3. Gaze Points, Eye Movements, and Stimulus Information:

   • Coordinates for gaze points (X, Y), eye positions, pupil diameters, gaze directions, fixation points, movement types, presented stimuli, media details, and sizes.

4. Other Information:

   • Mouse positions, client area positions, viewport positions and sizes, full page dimensions, and additional event-related data.

These descriptions provide an overview of the types of information contained within the columns of the Eye Tracker Data and Raw Data files in the EyeT4Empathy dataset. Detailed information for each specific column might be available in accompanying documentation such as columns_explained.pdf or coordinate_system.pdf within the dataset folder.

Experimental Methodology:

Experimental Design and Protocol:
The experiment was conducted at the AI Lab of Oslo Metropolitan University in Norway.

Participant Details:
Sixty participants, aged 20 to 40, engaged in both gaze typing and visual information foraging interventions. Predominantly students, the group comprised four nurses who were also part-time students and 56 bachelor's and master's students from the University of Oslo, NTNU, and Oslo Metropolitan University. Data collection occurred at Oslo Metropolitan University, resulting in approximately 10 eye-gaze trajectory datasets per participant, totalling 502 datasets. Datasets with inadequate data points or excessive NaN values were excluded. Each participant completed the empathy level questionnaire twice, before and after their respective intervention.