The Biostatistics Core has a strong emphasis on clinical outcomes and applying data analytical strategies to real world applied pharmaceutical problems. To help serve you better, please keep in mind the following:
Once the data is collected the "die is cast". What is more important than data analysis is data design or instrumentation (in the case of a survey). If you or your student are using a statistician, it is critically important that they be involved early on (before data collection occurs). Many benefits can come from involving the statistician early and it can greatly aid you and your student in preparing for the data analysis. This is the point where you can get input on power and sample size.
Example - 4th year Pharm.D. student
- Student Project: During instrument development if applicable and at least two months before the project is to be presented for data analysis. It is always good to practice to involve the statistician at the beginning of the project rather than at the end!
- Grant Preparation: At least three months before the due date of the grant.
- Analysis for Journal Article: At least two months before the student/faculty member is planning on submitting to the journal.
- Analysis for an Abstract/Professional Presentation: At least one month before the student/faculty member is planning on submission.
Many different software packages can be used for data analysis, including some websites that will run a test for you. In addition, programs like Excel can do a limited number of statistical tests. It is important to use the right tool for the job. Consulting with someone with a background in data analytics can prevent problems from occurring. Data formatting is critical to conducting proper statistical testing. This includes how to format data in an electronic collection table and coding variables so they can be properly interpreted by the software.
For any given research questions, there can be a number of different ways to statistically make inferences or draw conclusions. The statistician can help you to articulate your research question, determine dependent and independent variables and build a data collection scheme consistent with the best statistical tools to answer your questions.