Leveraging Modern Data Analytics in Schools
Data analytics has a fundamental role to play in helping teachers and school administrators make better decisions. One of the big questions likely to be on the agenda for school leaders around the country this year is whether a shift from diagnostic to predictive analytics in schools would be beneficial. With one of the positives being early intervention before issues become evident.
In this article one of Atturra’s Senior Consultants, Ian Quartermaine runs through what the modern analytics journey looks like for a school, and gives an overview of where he sees Australian schools are positioned on this journey.
Four Stages of the Modern Analytics School Journey
1. Descriptive Analysis: Most schools have attempted some form if this. Looking at what happened and what they can learn with hindsight in mind. Their learnings can quite easily be used to make improvements, but the value is relatively low.
2. Diagnostic Analysis: This stage involves looking at why something happened and developing insights from this. This will provide more value, but it is harder to implement.
3. Predictive Analysis: All about being able to predict what will happen based on analytical modelling.
4. Prescriptive Analysis: At the final stage, this involved foresight, and being able to take actions to make certain things happen.
Where are Australian Schools at in the Journey?
Some schools seem to be remaining at the Descriptive or Diagnostic Analysis stages. For example, there are various off-the-shelf applications schools have experimented with, generally Descriptive and (occasionally) Diagnostic in nature. They use specific student data sets from standard testing environments like NAPLAN or ACER. The applications represent what has happened and use some simple heuristics to make connections between data fields as to why?
Schools will also be using descriptive and diagnostic analytics to gather insights about student wellbeing, which will continue to be high on the agenda in 2024. Software applications in this space can obtain data directly from student questionnaires, to provide a rich source of data that schools can tap into.
A handful are experimenting in the Predictive Analysis area. Though, few institutions (besides a handful of tertiary bodies) can claim any success at this more advanced stage. This type of analysis usually requires data modelling in conjunction with Machine Learning. Data is ingested and fed into a model by an engineer who looks for the field values that might reasonably predict the percentage likelihood of an outcome. One example is using student data on Demographics, Attendance, Assessment scores and disability to identify the students who would benefit from a wellbeing assessment.
For any of these types of analytics to be effective, schools first need to look at what they want to predict and what their end goal is. All schools will have a different conversation here, but one thing seems clear. Those who launch into a Data Analytics project without first ascertaining why they are doing it aren’t likely to save time, resources, or money in the process.
About the author:
Ian Quartermaine is a Senior Consultant within the Business Applications team for Atturra. Ian has over 30 years’ experience as an IT leader and manager, including as an adviser to IT managers and school leaders within the Independent Schools sector. He brings experience in school- based applications (school management and learning systems), has managed a national network for Independent Schools in partnership with AISNSW, and negotiated group purchase arrangements. Over these years, he has experienced and given advice on a huge range of school-based business and education applications and processes.
