Comparing Human and Automated Coding of News Articles on Hydraulic Fracturing in New York and Pennsylvania

ABSTRACT Understanding how the news media portray controversial natural resource issues is an important area of environmental policy research due to the media’s ability to influence public opinion and policymaking. Automated media coding is becoming increasingly used as an alternative or supplement to the human coding of these portrayals. However, the comparability of human versus automated coding of the news media has not been well documented. This research note provides a comparison of an automated method to human coding of newspaper articles on the controversial issue of hydraulic fracturing. It describes the methods and results of coding 1,037 articles from four newspapers in New York and Pennsylvania. This research note compares the automated method to a human coding method and describes the proportion of coverage of New York and Pennsylvania newspapers on economic, environmental, and social topics related to hydraulic fracturing.