McDonald B W, Pack A R
Department of Statistics, La Trobe University, Australia.
J Clin Periodontol. 1990 Mar;17(3):153-8. doi: 10.1111/j.1600-051x.1990.tb01079.x.
This paper reviews some approaches to statistical analysis of dental data. Often dental data consist of more than one observation per patient which may not be independent, and consequently some standard statistical methods based on independence of all observations are inappropriate. This paper seeks to make the dental researcher and the consulting statistician aware of some of the efficient approaches that can be used to analyse the data successfully. No specialist statistical knowledge is assumed. Rather the goal is to increase awareness of what can be done. The statistical layman is introduced to some basic concepts, while references are given for the mathematically-minded reader. The data of Pack, Coxhead and McDonald is used as a springboard for this discussion.
本文回顾了一些牙科数据统计分析的方法。牙科数据通常包含每位患者的多个观察结果，这些结果可能并非相互独立，因此一些基于所有观察结果独立性的标准统计方法并不适用。本文旨在让牙科研究人员和咨询统计学家了解一些可以成功用于分析数据的有效方法。无需专业的统计知识。相反，目标是提高对可行方法的认识。向统计门外汉介绍一些基本概念，同时为有数学头脑的读者提供参考文献。帕克、考克斯黑德和麦克唐纳的数据被用作本次讨论的出发点。
