Most commentators agree that over 90% of the data ever created has been generated in less than the past two years, and that trend is only set to increase.
Education-oriented organisations focus on retaining students, delivering effective learning experiences and delivering mandated reports on such activities. Today the bulk of this data derives from activity data - about what is actually done, rather than what someone has stated.
What you can do
Recognise what is possible and plan accordingly
Do not underestimate the importance of knowing and engaging with your online audience. Some organisations have made multi-million investment decisions based on projections from website analytics. Google Analytics is a popular tool which is free to use.
Our effective learning analytics project provides a good practice approach for all aspects of implementing a learning analytics programme. The xAPI specification used in this framework allows data from diverse systems to be integrated for analysis. The principles adopted in this project are applicable to other data analysis activities.
Review the appropriateness of derived metrics
HEFCE undertook an independent review on the role of metrics in research assessment entitled The metric tide.
We explored the implications of their review in our news article.
Be aware of legal implications
Always check legislation such as the Data Protection Act.
Our code of practice for learning analytics is a checklist for proper ethical and legal process when collecting data relating to individuals.
Employ common identifiers to easily combine data sets
Consider how unique identifiers for individuals, course modules, subject areas and other data items can be included in data sources.
Consider longer term uses of data
Our prospect to alumnus work is following emerging data models that capture formal and informal skills throughout a student’s learning life, and how joining up data from different systems can bridge the gap between education and employability.
Our business intelligence project, working with the Higher Education Statistics Agency (HESA), is developing a shared service for UK education.
We have established a shared data centre that will manage very large medical and other data sets to improve big data shared research, emerging data mining and linked data models.
The growing availability of open data presents new possibilities to integrate local institution data with sets that are freely available. For example, social media indicating engagement, trending topics or institution-related activities; Ordnance Survey maps for better visualisations; government data - including demographic information, international data, climate and environmental.
The ubiquity of devices and corresponding data collection apps is driving new ways of collating and exploiting such data in meaningful ways. Martin Hamilton's article A year of living in the future explores this further.
With the incentive of using unique identifiers in data to assist with joined up analysis, linked data technologies are being adopted. This enables software applications to access all pertinent information on a particular topic, ensuring relevant, intelligent filtering not possible through simple keyword searching.
We look forward to the exciting implications for progress in every discipline.