Skip to main content

Learning analytics

Sector Snapshot Getting Started Next Steps Talk and Share

What is Learning Analytics?
Learning Analytics is the process of measuring and collecting data about learners and learning with the aim of improving teaching and learning practice through analysis of the data

Where did Learning Analytics come from? 
Learning Analytics is a new field enabled by the advance of big data sets increasingly sophisticated analytical tools such as visualisation software improved data formats and advances in computing technology. Scientific disciplines (Physics Biology Climate Sciences) have been using analytics since the 1970s and learning is a relative latecomer to this field of inquiry. But over the last 10 years there has been a surge of interest as evidenced by the proliferation of research journals in the field including the Journal of Educational Data Mining and the Journal of Learning Analytics and a host of international conferences such as the International Conference of the Learning Sciences and the Conference on Learning Analytics and Knowledge (Baker and Siemens 2014; SOLAR 2015).

The internet mobile devices and a plethora of Learning Management Systems (LMS) are all leaving “learner-produced data trails” that can provide an insight into the learning process and opportunity for improvement. Higher education reformers who have often lacked the agile evidence-base required to implement change are now looking to analytics as a credible framework for transformation (Long and Siemens 2011).

How does Learning Analytics work? 
A variety of tools and approaches is used in Learning Analytics to provide educators with quantitative intelligence to make informed decisions about student learning. Data is collected from a broad range of sources including behavioural data taken from online learning systems (discussion forums activity completion assessments) and functional data taken from student admissions systems and progress reports (Sharples et al . 2014).

A range of statistical methods is then applied to this data including prediction modelling (used to infer information such as drop-out rates and learner outcomes) social network analysis (which analyses the relationship between networks of individuals groups and organisations) relationship mining (which analyses the links between sets of data patterns such as student success rates) and data for human judgement (data visualisation that enables teachers to give timely feedback to students and integrate results into pedagogical activity) (Baker and Siemens 2014).

The resultant data can then be used to inform activity such as curriculum mapping personalising the learner experience predicting behaviour (such as drop-out rates) designing learning interventions and determining competencies.

The video link below from Dragan Gasevic a leading authority in Learning Analytics (President of SOLAR) describes this emergent field.

See: https://vimeo.com/104688307


Sector Snapshot

Where is Learning Analytics currently being used and how?
In the US several large initiatives which are focussed on both the science of Learning Analytics and its potential impact have been created to drive the field forward. These include the Next Generation Learning Initiative (funded by the Gates Foundation in partnership with EDUCAUSE) which has identified Learning Analytics as a critical area for development (NMC 2012).

Grassroots initiatives include the ‘Purdue Signals’ programme from the University of Purdue which uses data to provide faculty with an early warning system for struggling students and enables them to instigate timely and targeted support.

At the University of Maryland Baltimore County a ‘Check My Activity Tool’ enables learners to compare their Learning activity within an LMS to that of their peers. The aim is to encourage underperforming learners to adapt their behaviour and engage in the similar patterns of learning (improved interactions and resource access) as successful fellow students.

A primary network within the EU is the Learning Analytics Community Exchange (LACE) which has recently published a Manifesto for Learning Analytics in the Workplace. This community is cross sector but includes UK HE.

Approaches to Learning Analytics in UK higher education are wide-ranging and driven by an assortment of motivating factors. Universities with distributed leadership cultures are encouraging a variety of Learning Analytics projects across schools and disciplines (e.g. University of Edinburgh) others are implementing a centralised approach driven primarily by performance management strategies (e.g. Bridgwater College) (Sclater 2014).

Other initiatives include:

  1. The University of Manchester examined patterns of student activity building and utility usage to create a strategy for sustainability and carbon reduction.
  2. The University of Huddersfield analysed the link between library activity data and student attainment. Certain patterns of library usage were used to identify struggling students and arrange extra support.
  3. The University of Derby used analytics to create an evidence-based support programme for Black and Minority Ethnic (BME) students.
  4. The Open University used Learning Analytics to identify issues relating to different types of disability when learning online.

What is clear from this patchwork of initiatives is that the field of Learning Analytics in the UK is in its infancy. Driven by IT departments and Library staff alongside enthusiastic faculty the benefits can be significant. Learning Analytics has yet to be used as a substantial tool for transformational change in higher education (Sclater 2014).

What are the potential benefits of Learning Analytics?
Learning Analytics surfaces previously invisible patterns in the learning process and provides HEIs with insight and intelligence around which to make decisions. It creates potential to improve administration with a more targeted and effective approach to:

  • resource allocation;
  • improve organisational productivity by providing timely feedback that can be actioned quickly;
  • identify and better support learners at-risk of dropping out;
  • provide faculty with intelligence about students learning habits;
  • promote innovation to transform academic organisation and pedagogical approaches.

As George Siemens notes higher education can become “a more intentional intelligent organisation with data evidence and analytics playing the central role in this transition” (Long and Siemens 2011).


Getting Started

How do I get started with Learning Analytics?
A good place to start is by exploring the use of ‘Analytics’ in the HEA journals at http://journals.heacademy.ac.uk/  and the HEA website knowledge HUB by searching for learning analytics.

JISC’s The Code of Practice for Learning Analytics is also a practical place to start. It identifies nine areas that should guide practice. These include:

  1. Responsibility: The collection of data with due regard to legal and ethical issues.
  2. Transparency and consent: Transparency about how and why data is used.
  3. Privacy: Safeguards to protect sensitive data.
  4. Validity: The quality and robustness of the data.
  5. Access: Students should be given access to analytics which use their data.
  6. Enabling positive interventions: Using appropriate means to intervene based on data.
  7. Minimising adverse impacts: Ensuring that analytics do not reinforce bias and social difference.
  8. Stewardship of data: Compliance with relevant legislation including the Data Protection Act.

See: https://www.jisc.ac.uk/guides/code-of-practice-for-learning-analytics

What should I expect if I try this approach?
Whilst there are significant benefits to be gleaned from the field of Learning Analytics there are a host of parallel issues that require thoughtful exploration. These include questions about data ownership the ethics of use and dangers of abuse. Concerns also include the use of data to track student learning habits in online spaces and the implementation of interventions based on this data (Greller and Drachsler 2012).

Learning Analytics focuses on large sets of quantitative data that provide specific intelligence to institutions. What these data sets don’t provide is qualitative insights that when used alongside analytics provide a richer contextual picture of the HE system. Learning Analytics might best be used as part of a balanced approach which encompasses both quantitative and qualitative data to provide a deeper understanding of learning behaviour.


Next Steps

Where can I learn more about Learning Analytics?
Listen to and join the Learning Analytics conversation by following these hashtags and Twitter handles:

#dalmooc

@soLAResearch

@jla_editorial

@laceproject

What HEA resources should I take a look at?

Gordon N (2014) Using pedagogy and learning analytics to manage our students. Higher Education Academy: York 

Slade S (2014) Learning analytics: ethical issues and policy changes. Higher Education Academy: York 

How else can the HEA support my professional development?
The UKPSF provides the framework for recording aspects of professional practice where the innovative use of Learning Analytics could be included. Find out more about UKPSF.

Come to a HEA event to share your experiences with your peers – See https://www.heacademy.ac.uk/events-conferences

In your social media share your experiences of Learning Analytics – you can tweet about it and include the #HEA to share it with those following the tag or perhaps you can submit a guest blog posting through us. See https://www.heacademy.ac.uk/blog  


Talk and Share

#dalmooc  @soLAResearch  @jla_editorial  @laceproject 

The materials published on this page were originally created by the Higher Education Academy.