Peterson L, Brown D, Kern T, Bartelstone J
Department of Psychology, University of Missouri-Columbia, USA.
Health Psychol. 1996 Mar;15(2):124-30. doi: 10.1037//0278-6133.15.2.124.
The present article outlines the advantages of the participant event monitoring methodology for the investigation of unpredictable, low-base-rate events in children. Several methods for assessing the quality of participant event monitoring data are advanced with a data set showing participant event monitoring of children's minor injuries by 61 children and their mothers. Child-mother correspondence and debriefing data suggest good accuracy for frequency estimates. Home- and laboratory-based simulations illustrate the participant event monitors' accuracy for major details. Traditional measures of data quality show good overall coder and test-retest reliability, and cross-observer reports show acceptable estimates of validity for objective aspects of the events and the expected lower estimates for the more subjective aspects. Conceptual and pragmatic difficulties of the method are considered, and suggestions for future research are advanced.
本文概述了参与者事件监测方法在调查儿童不可预测的低发生率事件方面的优势。文中提出了几种评估参与者事件监测数据质量的方法，并给出了一个数据集，该数据集展示了61名儿童及其母亲对儿童轻伤的参与者事件监测情况。儿童与母亲的通信及汇报数据表明频率估计具有良好的准确性。基于家庭和实验室的模拟说明了参与者事件监测在主要细节方面的准确性。传统的数据质量衡量指标显示出编码员的整体可靠性以及重测信度良好，而不同观察者之间的报告表明，对于事件客观方面的有效性估计是可接受的，而对于主观性更强的方面，预期估计值较低。文中还考虑了该方法在概念和实际应用方面的困难，并提出了未来研究的建议。
