Individual Submission Summary
Share...

Direct link:

The Effect of New York Times Event Coding Techniques on the Analysis of Protest Data

Sat, August 16, 4:30 to 6:10pm, TBA

Abstract

Newspapers represent the most frequently invoked source of data for the analysis of social movements. We know a great deal about the ways in which routines of news coverage may bias the content of newspaper data. Less is known, however, about how different article retrieval and coding practices influence data collected. For the newspaper most frequently used as a source of data in movements analysis, the New York Times, we compare the depictions of African-American Civil Rights movement activity over time that are produced using the two most common methods of retrieving articles: index versus full-story coding. The overall pattern of findings in models predicting the incidence of civil right protest is consistent when applying data from the alternate collection strategies, but the effects of some control variables change across models. Our results have several implications for social movement research employing New York Times data. While index and full-text access methods may generate patterns of protest that correlate strongly, researchers should employ extreme caution when comparing the results of research from full-text coding with results that rely upon index coding or attempting to combine the collection methodologies to conduct longitudinal analysis.

Authors