About Computational Social Science

While there is no singular, consensual definition of computational social science (CSS), it can be broadly understood as using computational techniques to analyze social science data or create social simulations, whether that data be large scale data that requires high speed computing capacities (i.e., “big data”), data that is of more modest scale but requires computationally-intensive processes (i.e., analyzing sets of texts, sounds, impacts, social networks, sensors, or sensory touch data), data collected online through scraping or interaction (e.g., online experiments), and data that is amenable to tools from machine learning or similar modalities, among other kinds of data.

Understanding, deploying, and advancing these methods requires more advanced skills in data retrieval and management and programming than is typically required in social science PhD programs. Distinct from engineering disciplines, analyzing these data from a social science perspective also requires attention to the social origins and meaning of these data, the samples or structured populations from which data may be drawn, the reasonableness of extrapolation from these data to larger populations or other social settings, research ethics, and epistemology among other concerns