Jacobson A F, Hamilton P, Galloway J
University of Texas, Arlington School of Nursing.
West J Nurs Res. 1993 Aug;15(4):483-94. doi: 10.1177/019394599301500407.
Secondary analysis of existing data offers many advantages to the nurse researcher. Data from large-scale studies may be reanalyzed and refined by secondary analysts with a fresh perspective, thus enhancing the original study's contribution to scientific knowledge. High-quality data can be obtained for comparatively little expenditure of time and money. The secondary analyst, however, must exercise care in evaluating and analyzing a data set to maximize the internal and external validity of the reanalysis. Because the secondary analyst's lack of involvement in data collection procedures may decrease insight into the original study's limitations, vigilant skepticism should accompany all phases of the research process in secondary data analysis, just as it should in all other research. Miller (1982) advised, "Begin by assuming the worst and seek out the same kinds of information about sample selection procedures, sample size, response rates, field procedures, and coding conventions that you would insist on if you were collecting your own data" (p. 722). By systematically evaluating potential data sets according to rigorous predetermined criteria, the nurse researcher can minimize the possible pitfalls inherent in secondary analysis. On the other hand, investigators who use secondary sources appropriately can make significant contributions to nursing science at less cost than that engendered by traditional research methods.
对现有数据进行二次分析为护理研究者带来诸多优势。大规模研究的数据可由二次分析人员以全新视角重新分析和完善，从而增强原始研究对科学知识的贡献。能以相对较少的时间和资金投入获取高质量数据。然而，二次分析人员在评估和分析数据集时必须谨慎行事，以最大限度提高重新分析的内部和外部效度。由于二次分析人员未参与数据收集过程，可能会减少对原始研究局限性的洞察，因此在二次数据分析的研究过程各阶段都应保持警惕性怀疑，这与所有其他研究一样。米勒（1982年）建议：“一开始先假设情况最差，然后去寻找与样本选择程序、样本量、回复率、实地调查程序和编码惯例等相关的信息，就如同你自己收集数据时会坚持要求的那样”（第722页）。通过根据严格的预先确定标准系统评估潜在数据集，护理研究者可以将二次分析中固有的潜在陷阱降至最低。另一方面，恰当地使用二手资料的研究者能够以低于传统研究方法的成本为护理科学做出重大贡献。
