User Experience and Learning in Video-Based and Computer-Vision-Based CPR (CVCPR) Training Dataset

[Dataset Description] This dataset contains user experience and learning response data collected from a randomized between-subjects experiment comparing two CPR (cardiopulmonary resuscitation) education approaches: conventional video-based CPR education and computer-vision-based CPR training. A total of 60 adults aged 19–29 participated in the experiment and were randomly assigned to two independent conditions, with 30 participants in each condition. The control group received CPR education through a conventional instructional video, while the experimental group used a computer-vision-based CPR training system using real-time skeleton detection. The system detected participants’ body movements through a camera and provided real-time visual feedback during CPR practice, including feedback related to compression posture, depth, and pace. The dataset includes participant demographics and pre-experiment responses concerning prior technology experience, interest in CPR, CPR confidence, procedural knowledge, previous training difficulty, and prior CPR training experience. Following the assigned training condition, participants completed a 10-item knowledge acquisition assessment and 7-point Likert-scale measures covering Cognitive Load (CL), Self-Efficacy (SE), Academic Interest (AI), Motivation (MO), Immersion (IM), Learnability (LE), and Preference (PR). These measures represent cognitive, motivational, and experiential dimensions of the CPR learning experience. The dataset also includes responses to five open-ended questions addressing participants’ overall impressions, perceived advantages and disadvantages of the training method, willingness and perceived ability to perform CPR in a future emergency, and suggestions for improvement. These qualitative responses complement the quantitative measures by capturing participants’ perceptions of feedback, interaction, realism, usability, and hands-on learning. Overall, the dataset supports comparative research on video-based and computer-vision-based CPR education, particularly in relation to learning outcomes, cognitive load, self-efficacy, motivation, immersion, user experience, and interactive emergency training. Keywords: CPR Education; Computer Vision; Skeleton Detection; Video-Based Learning; User Experience; Real-Time Feedback; Self-Efficacy; Cognitive Load; Immersion; Interactive Learning; Emergency Training【数据集说明】 本数据集采集自一项随机分组被试间实验，该实验对比了两种心肺复苏术（cardiopulmonary resuscitation, CPR）培训方案：传统基于视频的CPR培训，与基于计算机视觉（Computer Vision）的CPR实操培训。 本次实验共招募60名19至29岁的成年人，将其随机分为两个独立组别，每组各30名参与者。对照组采用传统教学视频开展CPR培训，实验组则使用搭载实时骨骼检测模块的计算机视觉CPR培训系统：该系统通过摄像头捕捉参与者的肢体动作，并在CPR实操过程中提供实时视觉反馈，涵盖按压姿势、按压深度及按压节奏的相关反馈。 数据集包含参与者的人口统计学信息，以及实验前调研数据，调研内容涉及既往技术使用经验、对CPR的兴趣程度、CPR操作信心、程序性知识、既往培训难度感知及既往CPR培训经历。 在完成指定培训方案后，参与者需完成两项测评：其一为包含10道题目的知识习得评估；其二为涵盖7个维度的7点李克特量表测评，分别为认知负荷（Cognitive Load, CL）、自我效能感（Self-Efficacy, SE）、学业兴趣（Academic Interest, AI）、动机（Motivation, MO）、沉浸感（Immersion, IM）、易学性（Learnability, LE）与偏好度（Preference, PR）。上述测评维度覆盖了CPR学习体验的认知、动机与体验三大层面。 数据集同时包含5道开放式问题的作答数据，问题涉及参与者对培训的整体印象、对本次培训方式的优缺点感知、未来紧急场景下实施CPR的意愿与感知能力，以及改进建议。这些质性作答数据补充了量化测评的不足，可捕捉参与者对培训反馈、交互设计、场景真实性、易用性及实操学习的主观感知。 总体而言，本数据集可用于对比研究基于视频与基于计算机视觉的CPR培训方案，尤其适用于探究学习成果、认知负荷、自我效能感、动机、沉浸感、用户体验及交互式应急培训相关的研究方向。 关键词：心肺复苏术培训（CPR Education）；计算机视觉（Computer Vision）；骨骼检测（Skeleton Detection）；视频化学习（Video-Based Learning）；用户体验（User Experience）；实时反馈（Real-Time Feedback）；自我效能感（Self-Efficacy）；认知负荷（Cognitive Load）；沉浸感（Immersion）；交互式学习（Interactive Learning）；应急培训（Emergency Training）