Founded in 1948, the Summer Institute in Survey Research Techniques is designed specifically to meet the needs of professionals and graduate students seeking to deepen their expertise in survey methodology and data collection. Offered through the Michigan Program in Survey and Data Science within the Institute for Social Research at the University of Michigan, the program provides a rigorous and flexible curriculum that blends theoretical foundations with practical application — entirely online.

Classes are open for registration.
You do not have to be affiliated with the University to attend.
Registration and payment are required a minimum of two weeks prior to the start of the class.

July 20-30, 2026 (M T TH: Live instruction; W: Video instruction)
10:00am-11:30am
Natural Language Processing with R
Presented by Robyn Ferg
Course Fee: $1,200

In this two-week course, students will learn a variety of natural language processing methods for analyzing and extracting meaning from text data. The course will start with an introduction to text data, including text preprocessing and exploratory methods. The topics that follow will include machine learning models used for topic modeling, clustering, classification, sentiment analysis, and word embeddings. Students will also be introduced to web scraping. Considerations to both long and short texts of various subject matter. Class examples will be demonstrated primarily in R. This course assumes a bachelors-level background in Statistics or related field and knowledge of R or Python; no prior knowledge of text analysis is assumed.

Robyn Ferg is a senior statistician at Westat. Her doctoral and postdoctoral research focused on developing methods for extracting insights from social media data. She has taught graduate and short courses at the Joint Program in Survey Methodology (University of Maryland) and given talks on text analysis at the Census Bureau, University of Maryland, University of Michigan, Michigan State University, and several national and international conferences. She has published original research on this topic in several peer reviewed journals. She has a PhD in Statistics from the University of Michigan.