Co-Principal Investigators: Beck, David; Stone, Sarah; Aulck, Lovenoor; Kennedy, Patrick C.

The use of education technologies is generating large quantities of data at unprecedented speed. These data (e.g., the prose in digital instructional materials, social media data, clickstreams from e-learning platforms) are largely unstructured or noisy, meaning that they are not arranged according to a preset data model and/or have many complicated attributes. The features of these data make it difficult to use conventional analytic tools and techniques. This training program will prepare education researchers to use advanced supervised and unsupervised machine learning and natural language processing methods in conjunction with human coding to facilitate extracting meaningful insights from education big data. The trainers will also address data ethics and professionalism throughout the training. The training is intended for a mixture of education researchers, school district data analysts, and education technology employees who have some background in statistics.

The program will recruit 3 cohorts of 15 to 20 participants per year (50 to 60 in total) for  an intensive 7-month-long training that includes 15 weeks of online webinar-based learning, a 1-week in-person workshop, and continuous virtual mentoring from the project team. In addition, the training team will develop a fully open-access website to host materials from the training for both participants and non-participants looking to extend their data analysis skills.