Learning Analytics, which could reveal a great deal of potentially useful information to several parties, has not been done to any notable extent in Sri Lankan Schools. At present, valuable student data are neither computerized nor available in (non-urban) schools. Thus, computer based analyses/Educational Data Mining methods could not be applicable. In other words, Learning Analytics is pretty much lacking in the domain of Sri Lankan School Education.
This research mainly focuses on finding the ways to adopt Learning Analytics in Sri Lankan schools to assess learning progress, predict performances and track potential issues.
To achieve this goal, three major approaches which are identified would be carried out in parallel (they are interdependent) throughout this project.
There are very little useless information. Thus, the raw data which are used to produce such information are important. Student Data, which could reveal a great deal of information, is not being analyzed properly. An information root, which could be beneficial for all four main stakeholders in the Educational Domain (learners, teachers, administrators and funders) should not be neglected.
At present, general Sri Lankan school learning environments have not been using a considerable IT to automate, to store and to analyze. An initialization of an end to end scenario (data collection to reporting useful information) could encourage them to slowly delve into the electronic learning assistance, which is inarguably better than the current ledger system. An initialization could help them to have a commencement and to have natural evolvement.
A broad research on educational data mining and Learning Analytics would be conducted to build predictive, warning model using past student data. To apply this research outcome, the student data would be collected from number of identified schools, which is geographically limited to Jaffna district initially (as a Pilot effort) and a predictive system would be built.
The electronic data available at present in schools are not comprehensive. This data collection will need a data cleaning task to pre-process the data and to create a student data source to do researches on learning analytics. Taking into account that the schools willing to provide and some of them are ready to feed ledger date in to computers, we assume our focus and time would be mainly on Learning Analytics.
A Student Information System (not necessarily a LMS though) to collect student data which fulfils the basic need to commence integrating Learning Analytics. According to the background study we did, there are no adequate electronic data or such facilities to electronically capture data. To deal with this scenario we have to identify the specific needs in the context of student information system to deploy a suitable (simpler) module in schools.
Thus the main focus would be the Learning Analytics, the literature survey is being done on three major areas, MIS (Management Information System), Data Provenance and Learning Analytics. Learning Analytics system could not operate without intelligent data, thus data collection/capture is an unavoidable prerequisite.
At present, even higher education systems around the world have not integrated Learning Analytics with their e-learning environments to a sufficient extent. Learning Analytics is an emerging field and particularly in Sri Lanka it is at the initial stage. Obviously Sri Lankan School systems are pretty much lagging behind in any kind of student data analyses.
We have visited number of major schools in Jaffna. They are interested in willing to computerize student data up to a feasible level and generate useful information via analyses. But it is not very practical to deploy a system with all three major parts mentioned above. It would be a huge leap and impractical. As our primary focus is on Learning Analytics, the background study is focused on how to deploy a Learning Analytic System in Sri Lankan schools, so that would actually run and evolve with time.