March 7, 2012

Choosing Appropriate Methods

• Biostatistics vs. lab research
• Good Study Design and Analysis Plans as Features of Ethical Research with Humans
• Working with statisticians

Consider:

• Does the study design enable the investigator to address the aims and hypotheses?
• What might go wrong? What alternative strategies would you turn to if something went wrong?
• What are the expected outcomes and why are they important?
• How will you assess statistical significance?
• How many subjects will you need? How did you derive that number?
• What are the controls? Have you addressed potential sources of bias?

Basic science approach:

1. Approach
2. Overview of methods
3. Essential reagents needed
4. Critical equipment required
5. Number of subjects/animals and how they were derived
6. Statistical analysis
7. Controls
8. Replicates
9. Detailed expectations
10. How results will be interpreted
11. Time required to complete
12. Expected outcomes
13. Potential problems and alternative strategies (if working hypothesis proves invalid)
14. Timeline
15. Future directions

Clinical science approach:

1. Study design and protocol overview
2. Study environment
3. Participant recruitment
4. Inclusion criteria
5. Exclusion criteria
6. Eligibility screening, enrollment, and randomization
7. Description of intervention
8. Interim monitoring
9. Data collection schedule and procedures
10. Data analysis plan
11. Sample size consideration
12. Missing data
13. Data management
14. Team communication and project management
15. Timeline
16. Limitations

• Age distribution of NIH principal investigators