McMaster University, Department of Kinesiology, 1280 Main Street West, Hamilton, Ontario L8S 4K1.
Hum Factors. 2010 Jun;52(3):441-55. doi: 10.1177/0018720810370340.
The aim of this study was to determine the effects of video playback speed on posture matching reliability of the distal upper extremity.
Video is frequently used in ergonomic assessments, yet there remains a need to determine the effects of viewing speed on posture observations.
Participants were 7 graduate students experienced with posture-based observational methods. Categorical posture scales were used to evaluate forearm pronation/supination, wrist flexion/extension, wrist radioulnar deviation,and hand activity from workplace video at three playback speeds (quarter, half, and real time). Wrist flexion/extension was also evaluated with a frame-by-frame video method.
Posture counts increased with slower viewing speeds for the wrist and hand, but percentage durations in each posture category were similar for all methods. Posture matching interrater reliability scores increased with slow-motion video playback but remained low even for quarter-time video playback. The highest interrater scores were found in the frame-by-frame analysis of wrist flexion/extension for three posture categories (percentage agreement = 84.9% +/- 1.3%; kappa = 0.54 +/- 0.02).
Although slower video playback speeds increased the number of posture counts for the wrist and hand scales, percentage durations were similar, and reliability scores increased only slightly with slow-motion video playback.
Reviewing video using slow-motion or frame-by-frame methods improves distal upper extremity posture matching reliability. However, ergonomic assessment tools based on percentage duration may not be appreciably enhanced by slowing viewing speed. Thus, the increased viewing time with slower playback should be justified with respect to assessment needs.
本研究旨在确定视频播放速度对上肢远端姿势匹配可靠性的影响。
视频在人体工程学评估中经常使用，但仍需要确定观看速度对姿势观察的影响。
参与者为 7 名具有基于姿势的观察方法经验的研究生。使用分类姿势量表从工作场所视频评估前臂旋前/旋后、腕关节屈伸、腕关节桡尺偏和手部活动，观看速度分别为 1/4、1/2 和实际时间。腕关节屈伸也通过逐帧视频方法进行评估。
观看速度较慢时，腕关节和手部的姿势计数增加，但所有方法的每个姿势类别的百分比持续时间相似。姿势匹配的组内信度评分随慢动作视频播放而增加，但即使是 1/4 时间的视频播放，信度评分仍然较低。在腕关节屈伸的三个姿势类别中，通过逐帧分析获得了最高的组内评分（百分比一致性=84.9% +/- 1.3%；kappa=0.54 +/- 0.02）。
虽然较慢的视频播放速度增加了腕关节和手部比例的姿势计数，但百分比持续时间相似，慢动作视频播放仅略微提高了可靠性评分。
使用慢动作或逐帧方法审查视频可提高远端上肢姿势匹配的可靠性。然而，基于百分比持续时间的人体工程学评估工具可能不会因观看速度的减慢而显著增强。因此，较慢播放速度的增加观看时间应该根据评估需求进行证明。
