Images contain information absent in text, and this extra information presents opportunities and challenges. It is an opportunity because one image can document variables with which text sources (newspaper articles, speeches or legislative documents) struggle or on datasets too large to feasibly code manually. It has been a challenge because of the technical difficulty of identifying the objects and concepts encoded in an image, requiring researchers to rely on manual coding. Because human coders are slow, expensive and have different interpretations of the same images, studies using images have historically been small. The falling cost of computing, coupled with the availability of large datasets, means these techniques will become mainstream shortly. Our research focuses on hastening that process, and the purpose of this blog post is to provide a very high-level overview of images as data

Read More