Statistical methods can benefit scientists throughout the entire research
process. In the early phases of medical device or drug development, extremely
efficient experimentation should be used to more quickly lead researchers to
optimal formulations and designs. From these experiments, predictive models
guide the research.
Statistical methods such as sample size determination and design protocol should
be employed in the clinical testing phase in order to minimize risk while
considering costs. The nature of the data being collected (continuous,
semi-continuous, qualitative) must be considered for the design and appropriate
analysis of the clinical study.