“Reading between the Lines: A Text-as-Data Approach to Studying Myanmar's Parliament”
College Fellow & Assistant Professor
Weinberg College of Arts & Sciences
SEPTEMBER 29, 2017
206 INGRAHAM HALL
This paper presents an approach to bridge qualitative and quantitative methods in political science, using expert assessments of participation in Myanmar’s parliament derived from textual sources to inform a quantitative analysis. Accessing diverse expertise, particularly in challenging developing or authoritarian contexts, requires expanding the pool of “experts” and leveraging information not just in interview contexts but also through written works as a way to systematically apply previous knowledge to current modeling challenges. This paper investigates participation in Myanmar’s parliament and demonstrates a method for characterizing expert opinions from newspaper, ethnographic, and other qualitative sources using text-as-data analysis. Specifically, the paper elicits expert opinions from 4,126 English-language newspaper articles about Myanmar’s parliament published from 2010-2013 across 12 national and international sources. I treat the editorial boards of these sources as “experts” and use sentiment analysis to characterize their opinions and uncertainty about how certain variables such as party affiliation and ethnicity impact participation in Myanmar’s parliament. I evaluate the opinions extracted from these textual sources against opinions from Myanmar experts gathered through an online survey. The paper then demonstrates the text-as-data method for eliciting expert opinions with additional qualitative sources of scholarship about Myanmar’s parliament, and discusses how this approach generalizes in order to bridge the qualitative/quantitative divide and leverage new materials for studying politics in authoritarian contexts throughout Southeast Asia.