Universidad Autónoma de Madrid, Madrid, Spain.
University of Auckland, Auckland, New Zealand.
Behav Res Methods. 2023 Feb;55(2):855-866. doi: 10.3758/s13428-022-01861-0. Epub 2022 Apr 27.
Computer-aided behavior observation is gradually supplanting paper-and-pencil approaches to behavior observation, but there is a dearth of evidence on the relative accuracy of paper-and-pencil versus computer-aided behavior observation formats in the literature. The current study evaluated the accuracy resulting from paper-and-pencil observation and from two computer-aided behavior observation methods: The Observer XT® desktop software and the Big Eye Observer® smartphone application. Twelve postgraduate students without behavior observation experience underwent a behavior observation training protocol. As part of a multi-element design, participants recorded 60 real clinical sessions randomly assigned to one of the three observation methods. All three methods produced high levels of accuracy (paper-and-pencil, .88 ± .01; The Observer XT, .84 ± .01; Big Eye Observer, .84 ± .01). A mixed linear model analysis indicated that paper-and-pencil observation produced marginally superior accuracy values, whereas the accuracy produced by The Observer XT and Big Eye Observer did not differ. The analysis suggests that accuracy of recording was mediated by the number of recordable events in the observation videos. The implications of these findings for research and practice are discussed.
计算机辅助行为观察正在逐渐取代纸笔方法进行行为观察，但文献中缺乏关于纸笔与计算机辅助行为观察格式相对准确性的证据。本研究评估了纸笔观察和两种计算机辅助行为观察方法的准确性：Observer XT®桌面软件和 Big Eye Observer®智能手机应用程序。12 名没有行为观察经验的研究生接受了行为观察培训协议。作为多元素设计的一部分，参与者记录了 60 个真实的临床会议，这些会议随机分配给三种观察方法之一。所有三种方法都产生了很高的准确性（纸笔方法，.88±.01；Observer XT，.84±.01；Big Eye Observer，.84±.01）。混合线性模型分析表明，纸笔观察产生了略高的准确性值，而 Observer XT 和 Big Eye Observer 的准确性没有差异。分析表明，记录的准确性受到观察视频中可记录事件数量的影响。讨论了这些发现对研究和实践的意义。
