Abstract

With computers being used in all areas of life, vast amounts of digital data are generated continuously, much of it language data. Such big data is attractive to researchers for its volume, variety, and velocity, holding the promise of overcoming limitations of smaller-scale analyses. At the same time, extracting reliable and meaningful information from such data remains a constant methodological challenge because of the unknown quality of the data (veracity) and the conceptual challenges, computational techniques, and statistical methods required to interpret it and support scientific insight. For SLA, especially the online learning and assessment platforms increasingly used for foreign language teaching and learning offer unprecedented opportunities for big data exploration of SLA research questions, while also making it possible to turn SLA interventions into real-life applications. In this chapter, we review the key methodological challenges arising for big data SLA research and identify natural language processing and statistical methods that can help overcome challenges. We review empirical case studies that demonstrate the value of big data for SLA research and highlight areas where big data is likely to benefit future research and practice.