Manasse H R, Lambert R L
Am J Hosp Pharm. 1980 May;37(5):694-701.
Different types of research are categorized by reviewing their characteristics and describing how they can be used. Controlled research is experimental in nature and, through a priori planning, allows manipulation of the independent variable so causality can be shown. In ex post facto research, the researcher does not have control of the independent variable and so must infer causality by measuring dependent variables within the framework of independent variables that are already established or whose changes have already taken place. Most applied research in pharmacy is of the ex post facto type. The research questions to be answered will help determine the timing of the measurement of dependent variable (i.e., measure past events or make plans to measure future events), the measuring of the dependent variable (primarily with surveys), and whether time can be used as an independent variable. The effect of time can be determined by repeated measurements over a span of time (longitudinal design) or by measurement of several groups at one point in time (cross-sectional design).
不同类型的研究通过审视其特点并描述其用途来进行分类。对照研究本质上是实验性的，通过预先规划，可以对自变量进行操纵，从而证明因果关系。在事后回溯研究中，研究者无法控制自变量，因此必须在已确定的自变量框架内或其变化已经发生的情况下，通过测量因变量来推断因果关系。药学领域的大多数应用研究都属于事后回溯研究类型。有待回答的研究问题将有助于确定因变量测量的时间（即测量过去的事件或制定计划测量未来的事件）、因变量的测量（主要通过调查），以及时间是否可用作自变量。时间的影响可以通过在一段时间内进行重复测量（纵向设计）或在某一时刻对几组进行测量（横断面设计）来确定。
