Research Staff Professional Development

The Office of Institutional Effectiveness and Advancement (OIEA) is committed to the professional development of all office staff. This page provides an outline of the basic professional development for the research and assistant research analysts of the office.  This development focuses on the statistical software R and on statistical learning/predictive modeling.



  1. Introduction to R Programming, offered through edX
  2. Use the R package "swirl to learn R in R." Use "install.packages('swirl')" from within RStudio to install.
  3. DataCamp - online courses to learn R, python, and data analysis.  Inexpensive subscription required.
  4. Principles of Machine Learning, offered through edX
  5. Statistical Learning
    1. Online Course in Statistical Learning, offered through Stanford University
    2. The course above closely follows the freely available An Introduction to Statistical Learning

Statistical (Machine) Learning - More Advanced

  1. The Elements of Statistical Learning, a freely available text
  2. Applied Predictive Modeling - from the author of R package "caret."


  1. SQLBolt - interactive tutorials to learn base SQL.

Programming and Data Wrangling Techniques - More Advanced

  1. dplyr cheat sheet - from the people who do RStudio
  2. The Art of R Programming: A Tour of Statistical Software Design, a text focusing on software development with R

Data Visualization - More Advanced

  1. Introduction to R graphics with ggplot2
  2. ggplot2 cheat sheet - from the people who do RStudio
  3. Building web applications with Shiny

Causal Inference

  1. An Overview of matching methods
  2. Why you shouldn't use propensity score matching - we tend to prefer Mahalanobis and coarsened exact matching.
  3. MatchIt package - can do both propensity score matching and non-propensity methods such as coarsened exact matching and Mahalanobis distance matching.
  4. Controlling for covariates in randomized control trials (or post-matching) for a continuous outcome.
  5. Controlling for covariates in randomized control trials for a binary outcome.