Statistical Methodologies Useful for the Analysis of Data from Risk‐assessment Studies

作者：Gary G. Koch, James D. Beck · 发表于：Journal of Public Health Dentistry · 年份：1992 · DOI：10.1111/j.1752-7325.1992.tb02260.x · 被引用次数：29 · 研究领域：Advanced Statistical Methods and Models、Reliability and Agreement in Measurement、Advanced Statistical Process Monitoring

In dentistry, the vast majority of studies attempting to predict who is at high risk for getting a disease or condition, or attempting to identify risk factors for a specific disease or condition, have focused either on only one risk factor at a time, or have measured multiple potential risk factors, but analyzed their effects in isolation. Since researchers tend to agree that most dental conditions have a multiple etiology, it is necessary to develop models that consider simultaneously the effect of a number of potential risk factors on the disease or condition of interest if we are to have any understanding of the relative impact of potential risk factors. Many existing statistical techniques will aid dental researchers in identifying risk factors. However, the selection of an appropriate analytic technique depends on a number of conditions. The strategy for this paper is to discuss a wide range of possible statistical techniques that may be applied to the problem of deriving a model for identification of multiple risk factors for dental diseases and conditions. We have approached this task by presenting a number of dental research problems needing an appropriate analytic technique. Next, basic issues that must be considered in choosing an appropriate analytic strategy are discussed. These issues include features of the study design, the data structure of the variables being measured, and the types of assumptions that are applicable to provide valid inferences about the tar...