Data Mining for Analytical Nuggets
To prove anything in education you need data. Furthermore, when so much of education is subjective rather an objective, more qualitative than quantitative, data needs to be carefully explored, sifted, refined and analysed. Once done, you have those nuggets of evidence that can justify actions, policy decisions and students’ futures.
According to the International Society for Education Data Mining, it is ‘an emerging discipline, concerned with developing methods for exploring the unique and increasingly large-scale data that come from educational settings, and using those methods to better understand students, and the settings in which they learn’. Sources of data range from summative test and exam results to administrative information on individual students. Issues of time, sequences and context are important, leading to judgements about progress.
But how can data mining help us in online learning?
In ‘An examination of online learning effectiveness using data mining’ (Shukora, Tasira, Van der Meijdenb, 2014) the authors concluded, “online collaborative learning environment was found to have significant influence upon students’ learning achievements…they can perform significantly better upon learning in an online collaborative learning environment”. Data mining allowed the researchers to not only analyse the performance of students but, perhaps more importantly, explore how students preferred to learn. We know that a happy, confident student is more likely to achieve therefore the data needs to diagnose the perspective of the student’s profile, not just the summative data.
According to McKinsey Global Institute, the education sector ranks as one of the economy’s top ten in terms of the amount of data stored. The question is, what do we do with it?
As The Guardian points out (26th March 2014) data mining takes the form of ‘analytics which make use of the digital traces left by learners in online learning environments in order to improve teaching and learning and the environments in which these take place.’ The article makes the point that students (and their teachers) need to understand the ‘how’ of learning equally as much as the ‘what’. Yet in our highly accountable system where it’s the ‘what’ that gets measured in exams, the ‘how’ gets overlooked.
This “learning to learn” methodology can be built into online materials easily. As Education Manager for nimble®, I support schools and colleges in their use of our authoring programme and LMS (Learner Management System), to create materials that assess this metacognitive approach to learning. You need a range of assessment methods to do this, formative ones particularly. We’ve built a range of surveys into our platform for this reason to encourage students to review how they’re learning as they use each resource, which leads to better quality reflection in the learning process.
The LMS can gather data from these assessments to create ‘learner profiles’ so teacher and student can see what methods work best. These analytics will inform the teacher of whole group preferences too. In this way the skills of learning, which are equally as relevant to employers, can be enhanced; skills such as problem solving, collaboration, reflection and enquiry.
This data mining process needs to be cyclical because it doesn’t end, student progress is on-going so the data needs to trigger information about how the student can constantly improve. Hackman and Woolley maintain analytics are cognitive, technical and social in nature, as this process illustrates:
• Select the information
• Capture evidence
• Aggregate and report findings
• Predict outcomes (such as level of performance)
• Use data to inform ways in which the student can improve further
• Refine ideas – suggest methodologies to improve performance, try them out
• Share with others (students, teachers, groups)
Much of this can be done through the LMS (or your Moodle, VLE etc) but the key factor to bear in mind is to share the information with students, regularly.
One school that uses our platform extensively, Shireland Collegiate Academy, assign these competences to online learning activities. Students are expected to catalogue the competences they cover so that on ‘catch-up’ days they can ensure they address any competences that have been overlooked. This method places the responsibility for their learning firmly on the students’ shoulders, helping them to develop the same competences that will be relevant in the workplace. It’s not surprising the academy is considered to be Outstanding by OFSTED and achievement has been raised significantly.
Online learning could be accused of being little more than an endless process of clicking the mouse but what teachers must bear in mind is each click has the potential to provide information, on a personal and a group level. All those clicks provide data rich for mining and capable of providing the nuggets to inform subsequent decisions.