Library Learning Analytics: Addressing the Relationship between Professional, Research, and Publication Ethics

The advent of and increasing interest in learning analytics among researchers, practitioners, and administrators alike has academic librarians questioning what roles—if any—they can play in this sociotechnical movement. Briefly, learning analytics attempts to use data mining and analysis practices, including statistical algorithms, machine learning, and artificial intelligence, to investigate students' educational, social, and physical behaviors associated with or indicative of successful learning outcomes. Some cutting-edge approaches to learning analytics even use similar data to evaluate and intervene in professional situations (for example, with faculty, librarians, and advisers). Since student behaviors do not exist in a vacuum, learning analytics also examines, inter alia, the physical and digital resources, educational experiences, and interventions provided by an institution's faculty and staff to determine what effect, if any, they have on learning.

While learning analytics is a fairly new field of study, the academic literature seems to have taken notable interest. Scopus was used to test if this was true. A phrasal search of "learning analytics" in Scopus targeted on anywhere in the document between the years of 2010 (the inception of the field) and 2020 (the last full year of data) returned 12,690 results. Figure 1 demonstrates that the scholarly attention to learning analytics has significantly increased since the field's birth.

Figure 1.

The increase in the number of documents in the Scopus database mentioning "learning analytics" from the field's beginning in 2010 to 2020.

Academic librarianship has also increased its research on learning analytics. Kyle Jones, the author of this editorial, notes that the seeds of learning analytics were planted in the widely influential Value of Academic Libraries report in 2010.¹ Even stronger roots were established by a 2017 white paper by Lynn Silipigni Connaway and her team.² The evidence at the time suggested that so-called library value research had taken a notable turn toward studies of student success—arguably the focus of learning analytics research. Given the increase in learning analytics among academic librarians, a Scopus search was run to establish if a similar trend line existed in the profession. A concatenated phrasal search ("academic librarianship" OR "academic libraries" AND "learning analytics") was entered into the database, again with a broad scope of anywhere in the document. Figure 2 shows that, like learning analytics generally, there is a notable upward trend in the academic literature with 196 returned results—albeit starting in 2013 instead of 2010. portal's Editorial Board continues to see more submissions addressing learning analytics as a topic of interest, and the journal has published four articles in this area dating back to 2014.³

Contested Ethics

Learning analytics is not a neutral technological practice, and the ethical debate it has triggered has and continues to be intellectually vigorous and, at times, contentious—especially among academic library practitioners and scholars. At the heart of this debate is that learning analytics is predicated on accessing data which potentially reveal highly sensitive student behaviors, both digital and physical, and personal information such

Figure 2.

The increase in the number of documents in the Scopus database mentioning "learning analytics" along with "academic librarianship" or "academic libraries" from 2013 to 2020.

as demographics, preferences, and social networks. When interpreted and put into action, these data can lead to the development of student profiles, predictive scores, and targeted interventions by instructors, advisers, and librarians—anyone with access to the information who has power over a student's life. For advocates of learning analytics in libraries, this paradigm shift to make student experiences more transparent and analyzable provides significant potential. Megan Oakleaf and her coauthors argue that "libraries can round out institutional understanding of student learning and success by enriching a data picture that has thus far omitted student-library interactions." They also contend that "learning analytics represents a significant evolution in the ways librarians can use assessment approaches to listen to students, make decisions, and take actions to increase library support and dismantle hurdles that can...