A public dataset of people with aphasia and cognitively healthy people during an eye-tracking-based auditory comprehension test

Description:This dataset contains eye-tracking recordings from individuals with Broca’s aphasia and matched control participants collected during the administration of the Test for Reception of Grammar (TROG-2, Brazilian Portuguese adaptation). Data were acquired using a Tobii PCEye Mini device at 60 Hz and include raw gaze coordinates, timestamps, stimulus identifiers, and participant responses. Each trial consisted of a visual display with four images accompanied by an auditory sentence, requiring participants to select the picture matching the spoken prompt. The dataset includes: Raw gaze data (CSV files with timestamp, X/Y coordinates, stimulus identifier, and key press). Participants metadata (age, sex, handedness, education, diagnosis, time post-onset, calibration error; anonymized). AOI definitions for stimuli and example scripts for fixation and accuracy measures. README, Data Dictionary, and Stimuli documentation clarifying file organization and licensing restrictions (TROG-2 copyrighted material not redistributed). The dataset enables the computation of accuracy, fixation-based measures (time to first fixation, proportion of gaze on target AOI, number of fixations), and reaction times. Potential applications include benchmarking methods in computational linguistics, investigating language comprehension in aphasia, and developing assistive technologies based on gaze behavior. Ethics: The study was approved by the institutional ethics committeeesearch Ethics Committee under protocol no. 5.071.239 in October 2021, and all participants provided informed consent.数据集说明：本数据集收录了布洛卡失语症（Broca’s aphasia）患者与匹配对照组被试在完成语法接受测试第二版（Test for Reception of Grammar, TROG-2，巴西葡萄牙语改编版）时的眼动追踪记录。数据通过Tobii PCEye Mini设备以60赫兹的采样率采集，涵盖原始注视坐标、时间戳、刺激物标识符与被试反应数据。 每个试次均包含带有四张图片的视觉呈现与一句听觉语句，要求被试挑选出与口头提示匹配的图片。本数据集包含以下内容： - 原始注视数据：包含时间戳、X/Y坐标、刺激物标识符与按键记录的CSV格式文件； - 被试元数据：年龄、性别、利手、受教育程度、诊断结果、发病后时长、校准误差，所有数据均已完成匿名化处理； - 刺激物感兴趣区（Area of Interest, AOI）定义与用于计算注视指标及准确率的示例脚本； - README文档、数据字典与刺激物说明文件，用于阐明文件组织结构与许可限制（TROG-2的受版权保护材料不进行重新分发）。 本数据集可用于计算准确率、基于注视的分析指标（首次注视时间、目标感兴趣区注视占比、注视次数）与反应时。 其潜在应用场景包括计算语言学领域的方法基准测试、探究失语症患者的语言理解能力，以及开发基于注视行为的辅助技术。 伦理声明：本研究于2021年10月经机构伦理委员会以编号5.071.239的研究方案获批，所有被试均已签署知情同意书。