Ethics
Use of Student Data
Slade and Prinsloo (2013) state that the "collection of data and their use face a number of ethical challenges, including location and interpretation of data; informed consent, privacy, and de-identification
of data; and classification and management of data" (p. 1510).
The above image is an illustration of predictive analytics at play. Care should be taken when communicating directly with students on the basis of their analytics, as these are merely computer prompts based on "students like you". Netflix uses this type of prompt when the service suggests what type of show you would like to watch next, based on your viewing history, or Amazon will suggest similar products by cleverly asserting that "People bought these similar items". Slade and Tait remind us that "Students are more than the sum of their visible data."
Student agency gives students voice and often, choice, in how they learn. When the process of learning is automated through the use of adaptive learning technologies, does the student have real choice, or the illusion of choice? The students make a choice by submitting their own responses; however, it is then the program that is ubiquitously "learning" and collecting data from the student that determines their path, without taking into consideration other important factors, as we learned here.
As a consequence of taking student choice, it can be argued that the technology may be taking a student's voice as well. This is concerning because, as Prinsloo (2017) argues, “student data is not something separate from students’ identities, their histories, their beings. ... data is an integral, albeit informational part of students’ being. Data is therefore not something a student owns but rather is" (p. 7). Students who are not involved in the implementation of learning analytics as it pertains to their education are effectively being denied access to part of who they are.
Slade and Tait (2019) offer a consideration that the institution does not own the student data that it holds, but has temporary stewardship.
Machine Learning and Human Bias
Machine-learning-based adaptive platforms are the most sophisticated scientific method to achieve a truly adaptive state. Machine learning uses learning algorithms that create other algorithms, which in turn create adaptive sequences and predictive analytics that can continuously collect data and use it to move a student through a guided learning path.
But, just because the computer program uses data doesn’t mean that it is neutral. Consider the human bias in the creation of adaptive algorithms for machine learning by watching this brief video: