Global Guidelines: Ethics in Learning Analytics

In March of 2019, the International Council for Open and Distance Education (ICDE) published a 16-page report on global guidelines regarding ethically-informed practice in learning analytics. The ICDE established a working group, which included seven members from a range of countries. This group met through 2018 and created the global guidelines for ethical best practices in Learning Analytics. This report was authored by Sharon Slade and Alan Tait. This post will include a brief overview of the core components of the report. The full report is archived in LearnTechLib.

The report includes an overview of the history and purpose of Learning Analytics. Learning Analytics as a practice has grown tremendously over the past 10 years and has great value in supporting student success, as well as informing pedagogy, allocating resources, and informing institutional strategy. Learning Analytics also provides the opportunity for educators to use quantitative data to increase opportunities for student learning.

The report includes a number of issues that are important ethical considerations regarding the use and development of Learning Analytics. These issues include: Transparency, Data ownership and control, Accessibility of data, Validity and reliability of data, Institutional responsibility and obligation to act, Communications, Cultural values, Inclusion, Consent, and Student agency and responsibility. A brief overview of the ethical guidelines surrounding these core issues is below.

Data ownership and control

  • The issue of data ownership will be impacted to some extent by relevant national and international legislation (i.e. the General Data Protection Regulation in 2018).
  • Institutions should be aware of and make transparent issues around third party sharing, especially since sharing might include student data.
  • There is lack of clarity around who owns the data (institutions vs. students) and this can make principles surrounding meaningful consent complicated.
  • A suggestion surrounding the aforementioned issue is to have the institution see their role as having temporary stewardship over the data, and not owning the data.
  • For data that may be personal and/or sensitive, students should have some say as to how data can be used and who may be able to access the data.
  • Institutions should grant students the ability to correct and/or add context to their raw data.


  • Institutional transparency might best begin by making clear to students and to other stakeholders the purpose of learning analytics.
  • The core issue of transparency here relates primarily to how student data is collected and analyzed and used to impact students’ learning. Making students and stakeholders more aware of the uses of data brings about challenges, but also allows for greater insight and involvement. However, it is not always possible to be completely transparent, and it is not in the best interest of students to communicate a predicted poor outcome.

Accessibility of data

  • This core issue relates to the determination of who has access to the raw and analyzed data, as well as the ability of the students to access and correct their own data. It could also include making clear which data would be included within a learning analytics application, and what data might be out of scope.
  • Some individuals will have access to some categories of personal student data (i.e., income, academic history, etc.). Within a learning analytics context, data may be accessed on a ‘need-to-know’ basis in order to facilitate the provision of academic and other support services.

Validity and reliability of data

  • The institution needs to ensure that data collected and analyzed is accurate and representative of the issue being measured.
  • Proxy measures should be used with caution.
  • It is important to make sure that datasets are complete and sufficient to enable predictive calculations to be made. When learning analytics are communicated to stakeholders and students, the results should be transparent (where possible) and clearly understood.

Institutional responsibility and obligation to act

  • It is important to reflect on if access to knowing and understanding more about how students learn in turn brings about a moral obligation to act. Due to scarce resources, it is not always possible to provide the interventions that may be necessary. However, it is important for the institution to consider its policy for identifying where support resources are focused.


  • Care should be taken when communicating directly with students on the basis of their analytics. It is important that predictive analytics are truly predictions, and may be useful, it is also important to look for additional context. When communicating with students, it may be most effective to use general support terms (i.e., ‘we’re just checking in with you to see how your studies are going’).
  • Regular communications with staff should be encouraged, in order to help ensure that the staff understand the values, anticipated benefits for students, as well as guidelines for ethical practice.

Cultural values

  • In multicultural contexts, understanding and interpreting data are necessarily more complex. It is important to remember that measures used are likely to differ in different cultures. Care should be taken if purchasing analytics from developers to make sure that the approach is accurate for its purpose and can be adapted if appropriate.


  • With Learning Analytics, there is a danger that certain categories of students will be identified negatively as being particularly at risk. There are ethical issues relating to inclusion and exclusion if the institution communicates a desire to protect its success rate. Learning Analytics should be primarily used to support students.


  • Consent to collect student data should be sought at the point of registration. However, consent at this point is less meaningful for students, who may not know how learning analytics can be used to help them. If consent is sought at registration, it should include transparency and potentially with a later option to withdraw consent.
  • There have been conflicting responses and opinions regarding consent surrounding non-sensitive data. An alternative approach to considering consent in this way includes differentiating between initial consent for the collection of data and specific consent when data are used to intervene in the choices students have and in adapting their learning experiences.
  • Other approaches include providing opportunities for students to provide or withdraw consent where an intervention might significantly alter their experience. Consent should not be considered in binary terms, but should be presented to students as a menu of options.
  • Generally, this principle should be built around a minimum of informed consent.

Student agency and responsibility

  • Where possible, it is recommended that institutions seek to engage students in applications of learning analytics. In this way, students should be treated as equal participants in the uses of their data. Students can be more actively involved in creating and designing the interventions that will support them.

The issues that are discussed in this report may lead to the development of national ethical guidelines surrounding learning analytics, as currently none exist. The report includes an appendix of a range of policies and reports from different parts of the world. The Working Group would appreciate further references to national legislation and other significant reports. These can be sent to: [email protected].

Print Friendly, PDF & Email

Be the first to write a comment.

Your feedback