Phillips L R
West J Nurs Res. 1989 Apr;11(2):181-95. doi: 10.1177/019394598901100204.
As Hinshaw (1981) has noted, the process of designing research might best be conceptualized as a series of informed decisions that involve constant trade-offs and compromise. With each advantage gained by a particular theoretical or methodological decision, the researcher inherits a certain number of threats to the study's internal and external validity and a potential number of error sources. Because of the nature of the research questions of interest to nurses and because of situational and contextual constraints in the clinical setting, nurse researchers often are required to select samples that contain small or large proportions of elderly subjects. The purpose of this article has been to raise some of the theoretical and methodological issues associated with the selection of both age-heterogeneous samples and age-homogeneous samples. Although in research all possible error sources cannot be controlled, it is possible to reduce the amount of error by a clear understanding of the implications of choices regarding age-related issues and inclusion of design features that reduce the amount of error introduced when research subjects who are relatively young are studied along with research subjects who are elderly.
正如欣肖（1981年）所指出的，研究设计过程最好被概念化为一系列明智的决策，这些决策涉及持续的权衡和妥协。通过特定的理论或方法决策获得每一项优势时，研究者都会面临一定数量对研究内部和外部效度的威胁以及潜在的误差来源。由于护士感兴趣的研究问题的性质，以及临床环境中的情境和背景限制，护士研究者常常需要选择包含不同比例老年受试者的样本。本文的目的是提出一些与选择年龄异质性样本和年龄同质性样本相关的理论和方法问题。虽然在研究中无法控制所有可能的误差来源，但通过清楚了解与年龄相关问题的选择所带来的影响，并纳入一些设计特征，以减少在研究相对年轻的研究对象和老年研究对象时引入的误差量，就有可能减少误差。
