April 2013
Sun Mon Tue Wed Thu Fri Sat
  1 2 3 4 5 6
7 8 9 10 11 12 13
14 15 16 17 18 19 20
21 22 23 24 25 26 27
28 29 30        

Authors' Committee


Matt Blackwell (Gov)


Martin Andersen (HealthPol)
Kevin Bartz (Stats)
Deirdre Bloome (Social Policy)
John Graves (HealthPol)
Rich Nielsen (Gov)
Maya Sen (Gov)
Gary King (Gov)

Weekly Research Workshop Sponsors

Alberto Abadie, Lee Fleming, Adam Glynn, Guido Imbens, Gary King, Arthur Spirling, Jamie Robins, Don Rubin, Chris Winship

Weekly Workshop Schedule

Recent Comments

Recent Entries



SMR Blog
Brad DeLong
Cognitive Daily
Complexity & Social Networks
Developing Intelligence
The Education Wonks
Empirical Legal Studies
Free Exchange
Health Care Economist
Junk Charts
Language Log
Law & Econ Prof Blog
Machine Learning (Theory)
Marginal Revolution
Mixing Memory
Mystery Pollster
New Economist
Political Arithmetik
Political Science Methods
Pure Pedantry
Science & Law Blog
Simon Jackman
Social Science++
Statistical modeling, causal inference, and social science



Powered by
Movable Type 4.24-en

« App Stats: Vadhan on "Privacy Tools for Sharing Research Data" | Main

28 April 2013

App Stats: Roberts, Stewart, and Tingley on "Topic models for open ended survey responses with applications to experiments"

We hope you can join us this Wednesday, May 1, 2013 for the Applied Statistics Workshop. Molly Roberts, Brandon Stewart, and Dustin Tingley, all from the Department of Government at Harvard University, will give a presentation entitled "Topic models for open ended survey responses with applications to experiments". A light lunch will be served at 12 pm and the talk will begin at 12.15.

"Topic models for open ended survey responses with applications to experiments"
Molly Roberts, Brandon Stewart, and Dustin Tingley
Government Department, Harvard University
CGIS K354 (1737 Cambridge St.)
Wednesday, May 1st, 2013 12.00 pm


Despite broad use of surveys and survey experiments by political science, the vast majority of survey analysis deals with responses to options along a scale or from pre-established categories. Yet, in most areas of life individuals communicate either by writing or by speaking, a fact reflected in earlier debates about open and closed ended survey questions. Despite good reasons to collect and analyze open ended data, it is relatively rare in the discipline and almost exclusively done through a process involving human coding of survey responses. We present an alternative, semi-automated approach, the Structural Topic Model (STM) (Roberts et al. 2013), that draws on recent developments in machine learning based analysis of textual data. A crucial contribution of the method is that it incorporates information about the text, such as the author's gender, country of origin, treatment status, or when something was written. This paper focuses on how the STM is extremely helpful for descriptive, exploratory, or inferential purposes for survey researchers and experimentalists. The STM makes analyzing open ended responses easier, more revealing, and capable of being used to estimate treatment effects. We illustrate these innovations with several experiments.

Posted by Konstantin Kashin at April 28, 2013 11:25 PM