I have taught several different topics to students with varied backgrounds:

• STAT 212: A first course in biostatistics for undergraduates
• STAT 430: Data analysis for senior undergraduates and masters
• STAT 434: Survival analysis for graduates and advanced undergraduates
• STAT 530: Bioinformatics for masters and PhDs
• STAT 578: Topics courses for statistics PhDs

I've noticed that many of my students have trouble understanding how the topics they learn are connected. To address this issue, I am interested in how to teach statistics by helping students to construct concepts and methods for themselves.

One common disconnect is between statistics and data analysis. My students tend to have trouble analyzing data to answer a substantive question, even after they have learned the necessary statistical methods. I am interested in showing how conceptual problems in data analysis actually lead naturally to the development of statistical ideas. Here are slides from my latest attempt, from a two-hour short course on statistics that I teach each summer as part of the Mayo Clinic and Illinois Alliance Computational Genomics Course:

Another common disconnect is between different statistical methods. My students have trouble seeing that standard procedures are all based on the same core set of statistical and probabilistic principles. I am interested in how to teach students to start from these core principles to create and understand standard procedures.

Finally, I am interested in how to teach PhD students to do research. I encourage all of my students to read this and this, and so far have been fortunate to advise:

NameCurrent position
Will BiscarriJP Morgan
Robin TuNorthrop Grumman
Yihe WangFacebook
Huiqin XinFacebook
Rachel ZhouWaymo