Contemporary technology has created unprecedented opportunities to create radical improvements in learning, educational achievement, and social mobility, but also conditions under which information about learners and workers is collected continuously and often invisibly. The use of such evidence to pursue research must proceed in ways that respect the privacy, dignity, and discretion of the people they describe.

At the same time, the potential of new data sources to improve educational attainment and social mobility remains under-realized. In addition to technical and coordination challenges, researchers, administrators, and employers are facing complex questions about how to use these data responsibly. Pathways researchers contribute to ongoing national and international efforts to specify responsible behavior in the active utilization of data to improve learning and progress at school and at work.


fairness and ethics in computational assessment

How can AI inform academic evaluation?

This project leverages a corpus of over 800,000 college applications to determine how education data science might inform norms and practices of fairness and ethics in educational evaluation.

responsible use

ITHAKA/Stanford assemblies

What is responsible use?

This project entailed a series of assemblies co-hosted by ITHAKA S+R and Stanford University to develop shared understandings about responsible use of student data in higher education. Outputs of those convenings in 2014 and 2016 have contributed to ongoing international discussion about data-use practice and governance in a wide variety of sectors. Find documentation of the assemblies here.

responsible use


Setting the table: Responsible use of student data in higher education

What does it mean to use student data responsibly?

EDUCAUSE Review, 2018

Martin Kurzweil and Mitchell L. Stevens

Visit external site to read this publication

An ethically ambitious higher education data science

How do we pivot from rule compliance to ethical proaction?

Research and Practice in Assessment, 2014

Mitchell L. Stevens

Visit external site to read this publication