Objectives

• Be able to explain the basic principles of research design.
• Be able to review any given study and critique the experimental design.
• Know the difference between box-related and event-related designs and which you would like to use to investigate your hypothesis.
• Understand basic ethical concerns as they apply to neuroimaging studies.

Elements of an Experiment

Ideas

• Observations
• Case studies
• Curiosity
• Other studies
  • Issues
  • Translation
  • Replication
  • Extension
• Method development
• Moderators
• Mediators
• Testing Theory

Operationalization

Rendering the original abstract idea into something testable.

• Make the idea measurable.
• Points to variables of interest.
• Leads to a testable hypothesis.

Variables

• Independent (manipulated)
• Dependent (measured)

Protocol

• Select the methods that best fit the operational definition and variables.
• Guide this selection with a theory.

The HRF

• HRF - hemodynamic response function
• Model representing a theoretical fluctuation in hemodynamic activity.
• Activity in the CNS is statistically compared to the HRF to determine if and how well it matches. This is how we "light-up" the brain.

Basic types of HRF Designs

Blocked Design

• ​Hemodynamic activity is given a binary model.
  • i.e. activity is either on or off.
• Good for:
  • Task related investigation
  • Isolating very significant associations

• HRF is modeled on events with a random occurrence and intensity.
• Good for:
  • Multimodal studies.
  • Estimating the exact shape and timing of the true hemodynamic response

Why Ethics?

Tuskegee Syphilis Experiment

1932-1972

Declaration of Helsinki

1964

Ethics in Research

• Respect for persons
  • Individuals have a right to choose to participate or withdraw.
  • Individuals have a right to information.
• Beneficence
  • Risks should be minimized.
  • Benefits > Risks
• Justice
  • Risks are equally distributed throughout society.

Ethics in Research

• IRB
  • Oversees research approval process.
  • Research protocol
  • Consent form
• Confidentiality of data
  • All participant data is considered health data and protected by law (HIPAA).
  • Must be de-identified before made public.

Ethics in Research

• Incidental findings
  • Sometimes we happen upon a medically actionable situation.
  • Duty to inform and warn but CANNOT diagnose.
• Pregnancy
  • While there are no known risks, we cannot say there is no risk.
  • Local IRB gives guidelines on how to screen for pregnancy.