Multiple Hypothesis Testing and Its Applications
STA4515 (Graduate Only)
- This is one of the "New 4500 level courses that are 0.25FCEs and last 6 weeks" offered in
the Department of Statistical Sciences.
- Rationale and Objective
- A central issue in many current big-data scientific studies is how to assess statistical
significance while taking into account the inherent large-scale multiple hypothesis testing.
- This 6-week graduate course will first go over the fundamental elements of single and multiple
- It will then move on to more advanced topics such as incorporating prior information to improve
power with specific applications to whole genome genetic association studies.
- It will also discuss the fallacies of p-value and alternative measures of statistical evidence and
- It will present both analytical and empirical arguments.
It expects participating students write a research report on suggested or self-selected topics
related to multiple hypothesis testing.
- Audience: graduate students with solid training in statistics or biostatistics. Although
application examples involve statistical genetics, knowledge in statistical genetics is not required.
Examples will be presented in a self-sufficient way.
- Time: Thursdays 10-1pm
- Location: SS1084.
- Instructor: Lei Sun (email@example.com)
- Office hour: TBA
- Format of instruction: Lectures.
- Evaluation: in-class presentation and take-home (research-focused) project. Details to be announced during the first lecture.
COURSE SYLLABUS AND LECTURE NOTES
- The old-fashioned blackboard and chalk sticks will be used during the lectures combined with notes in .pdf format and handouts.