Terhorst Lauren, Beck Kelly Battle, McKeon Ashlee B, Graham Kristin M, Ye Feifei, Shiffman Saul
From the Department of Occupational Therapy, School of Health and Rehabilitation Sciences, University of Pittsburgh (LT); Department of Rehabilitation Science and Technology, School of Health and Rehabilitation Sciences, University of Pittsburgh (KBB, KMG); Department of Psychiatry, University of Pittsburgh (ABM); Department of Psychology in Education, School of Education, University of Pittsburgh (FY); and Department of Psychology, University of Pittsburgh (SS), Pennsylvania.
Am J Phys Med Rehabil. 2017 Aug;96(8):596-599. doi: 10.1097/PHM.0000000000000690.
Ecological momentary assessment (EMA) methods collect real-time data in real-world environments, which allow physical medicine and rehabilitation researchers to examine objective outcome data and reduces bias from retrospective recall. The statistical analysis of EMA data is directly related to the research question and the temporal design of the study. Hierarchical linear modeling, which accounts for multiple observations from the same participant, is a particularly useful approach to analyzing EMA data. The objective of this paper was to introduce the process of conducting hierarchical linear modeling analyses with EMA data. This is accomplished using exemplars from recent physical medicine and rehabilitation literature.
生态瞬时评估（EMA）方法在现实世界环境中收集实时数据，这使物理医学与康复研究人员能够检查客观结果数据，并减少回顾性回忆带来的偏差。EMA数据的统计分析与研究问题和研究的时间设计直接相关。考虑到来自同一参与者的多个观察结果的分层线性模型是分析EMA数据的一种特别有用的方法。本文的目的是介绍使用EMA数据进行分层线性模型分析的过程。这通过近期物理医学与康复文献中的示例来完成。
