Publications and Research
Document Type
Article
Publication Date
2019
Abstract
Researchers typically use social surveys or censuses to examine attitudes and behaviors across nations. While useful for understanding cross-national differences, they are expensive to collect, include only a limited number of issues and countries, and are not very time sensitive. Many countries across the world now have residents who regularly use Twitter and Google, and these internet platforms are increasingly making data on the country-level number of tweets and google searches available for analysis. While there are a lot of challenges with these data, we examine some of the potential benefits. Specifically, our study assesses the extent to which cross-national social media and survey measures related to religious expression are related. Focusing on Africa, where surveys are particularly difficult to administer, and religious expression, which is quite common across the continent, is high, we find that our religion-related measures derived from google searches correspond particularly well with traditional social science measures. We then look at how all three sets of measures explain terrorism, health-related issues, and the number of Christian and Muslim official holidays within the country. We find that the measures derived from Google almost always perform as well, if not better, than the traditional social science measures. We discuss how internet data may be able to offer reliable and time-sensitive measures for examining differences across nations and for better understanding a range of issues in Africa.
Comments
This is the authors' accepted manuscript of an article originally published in Society, available at https://doi.org/10.1007/s12115-019-00359-4