51 Study Design
Study design in experimental research is fundamentally different from study design in clinical research for several reasons. First, the question to be answered in a clinical trial usually comes from well-documented effects of, for example, a new treatment method. After finishing the preclinical studies, and possibly even a phase 1, phase 2, or phase 3 clinical trial, a randomized controlled trial is designed and conducted. Most of the prerequisites for the design of a randomized controlled trial are lacking in experimental research, and this will have direct consequences for the study design in experimental work. Second, the aim of experimental studies is to detect the mechanism behind an effect, or understand a causal relationship, rather than to investigate the effect itself as in clinical studies.
For these reasons, designing and conducting experimental studies can be perceived as difficult, or convincing the animal ethics committee can be seen as impossible. This section on statistics aims to suggest improvements in the study design, helping you to increase the scientific level of your work. This first chapter will discuss some of the specific aspects of study design in experimental studies, and the rest of this section will go into more detail concerning the statistical aspects of experimental research.
51.1 Hypothesis
The study you are about to start will usually begin with a question. This question has arisen from either literature, previous work you and your colleagues performed, or has come up during a discussion. A question, however, is not the best starting point for the design of a proper study, as it will leave too many open ends in the design. Instead of starting with a question, starting with a hypothesis will create a solid basis for your design. Formulation of a hypothesis helps the scientist to think about aspects of the issue at hand and avoids performing unnecessary studies that have to be repeated afterwards in a different manner. The hypothesis actually consists of two parts: the null hypothesis and the alternative hypothesis. The null hypothesis is the default hypothesis, assuming a steady state or no relationship between two variables. The alternative hypothesis states that there is an effect of treatment or a relation between two variables. In a specific experiment, the null hypothesis is tested against the alternative hypothesis. Formulation of a good hypothesis reflects the thought given to a study design, but will also determine specific aspects of the design and analysis. For example, when the hypothesis states that the effect of treatment A will be greater than the effect of treatment B, one-sided testing can be performed in the analysis, thus increasing the value of a specific experiment.
However, in experimental research, hypotheses are not always available due to a lack of data. Starting from a thought or a question rather than from proper data from previous experiments, studies are sometimes necessary to create a solid and well-defined hypothesis. In such cases, results from literature, preliminary work, and thorough knowledge of pathophysiology are the starting points for the design of a hypothesis-finding study.
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
