Study Design

There are several types of research study designs, each with its inherent strengths and flaws. The study design used to answer a particular research question depends on the nature of the question and the availability of resources. 

Elements of a Study Design

In all types of research, a clear hypothesis is necessary in order to choose the study design. Once the purpose of the investigation is understood and the hypothesis and objectives have been formulated, it is necessary to identify the study design to be used for the project in order to rigorously rest the hypothesis.   

  • Experimental/ Interventional Study Design- typically randomized clinical trial   
    • Control Groups  
    • Random Assignment  
    • Blinding and Placebo Control (singe blind, double blind, triple)   
  • Observational Study Design  
    • Population health   
    • Cohort Studies   
    • Case Control   
    • Cross sectional   

It is paramount that the entire research team be part of the overall clinical research protocol development process.

There are also four different primary phases of a clinical trial involving drugs. They are:

  • Phase I - first study of a drug in humans, most commonly healthy adults although it might also include patients with no other clear treatment options for example in oncology research; safety and dose ranging tolerability; basic metabolism or mechanism studies (generally small studies of 20 or fewer subjects).  
  • Phase II - first study of drug in target populations (patients); safety, dose ranging in patients; preliminary efficacy assessment (generally 100-300 subjects)  
  • Phase III - larger study to prove efficacy and increase experience with safety (generally large studies ranging from 100 to >1000 subjects   
  • Phase IV - post-marketing studies; vary in size; often to provide ongoing safety data.  

For studies involving devices the linked schema depicts the overall study lifespan from preclinical/pilot/pivotal to post market studies. For reference go to Medical Device Clinical Trials: Regulatory Pathways & Study Types Explained.

Most often the assistance of a statistician will be required for the selection of an experimental design, the development of the corresponding analysis plan and the subsequent interpretation of parameter estimates and test results.  Penn has many offices that can support your statistical needs. The Center for Human Phenomic Science (CHPS) has the Study Design and Biostatistical Core (SDAB). The Center for Clinical Epidemiology and Biostatistics has The Department of Epidemiology and Biostatistics (DBEI), and the Biostatistics Analysis Center (BAC). The Children’s Hospital of Pennsylvania (CHOP) has the Biostatistics and Data Management Core (BDMC).  Additionally, the Penn Medicine Abramson Cancer Center has the following Biostatistics and Bioinformatics Core (BBC).