Yesterday media around the country reported that one in four Australian school students drop out without completing Year 12. The stories were based on a report from the Mitchell Institute, who published the report “Educational opportunity in Australia 2015: Who succeeds and who misses out”, and a detailed factsheet about completion in senior schools. There were some key demographic factors that drove the differences in drop out rates - location, indigenous status, language background and socio economic status, and they also identified a number of key other factors - “poor grades in core subjects, low attendance, and disengagement in the classroom, including behavioural problems”. These are key predictors of the likelihood of individual students dropping out.
The principal actionable finding that they identify is summarised in a single sentence:
 | ||
| What is needed for all students is more personalised teaching and learning approaches | ||
 | ||
Which fits exactly with the theme of so much of the work that we do in education - whether it is through our broad events like the Next Level Learning workshops around the country next month, or the individual activities that my colleagues participate in with individual schools - like being student mentors, or hosting school career-planning days at our offices.
The other thing that is critical is the ability for a school, or school system, to be able to accurately predict which students are at risk of dropping out, to enable better intervention strategies. The work that we’re doing with advanced predictive analytics is core here, because it is possible to identify the key indicators, and easily analyse a student cohort in real time for their likelihood to drop out. Assuming that you are collecting the relevant data points.
I wrote about this back in July, as we’ve been working with Tacoma Public School District as part of their ongoing initiative to prevent student dropout for school students. The district has delivered a dramatic turnaround - eight years ago five of the high schools were described as “dropout factories”, and five years ago just 55% of students graduated on time, compared to 81% nationally. But last year that had been boosted to 78%, ensuring that the district is recognised nationally for its educational achievements. The district is now developing the next level of data-driven improvement with the help of Azure Machine Learning. As Shaun Taylor, the CIO for Tacoma Public Schools puts it:
| ||
| With Azure Machine Learning, we proved that we have the right tool to get us where we want to go in terms of predicting student success. It’s a tool our educators will be able to use to start tackling the problem of student disengagement. | ||
| ||
Read more about using Azure Machine Learning to reduce student dropout