We believe that one-to-one human interaction with a tutor is the secret to improved confidence and attainment in Maths BUT we also know that even the best tutor or teacher in the world needs data to help them adapt and progress. And that is why we have built an online platform to capture this data, why we are in the process of testing an assessment solution and why we monitor closely every interaction between pupil and tutor in our online classroom.
Improving the learning experience and pupil outcomes
The top priority we’ve set for our research team looking at the data is to use it to improve the learning experience and pupil outcomes. It’s a win-win. The more we can do this, the more likely we are to be able to build long-term, meaningful relationships with our schools.
Techniques we’re using and what they tell us
The amount of data we gather is enormous. So we now use state-of-the-art machine learning techniques and statistics to find hidden patterns in the data and extract information and insights about the most efficient teaching practices.
Before long we might be able to tell:
- How gender influences the pupil experience, outcome and confidence in maths;
- What impact any background noise or poor headphone use at school might have on the session’s outcome;
- How pupil’s engagement influences their effort, confidence and the number of steps to success they complete in a session.
Evidence-based approach avoid prejudices
The evidence-based approach to these questions allows us to avoid many prejudices. Moreover, unsupervised learning models can help us to find insights we couldn’t have come up with based only on the expert knowledge.
We are now able to use historical data to predict a school’s likely experience with Third Space. This model consists of a set of rules weighted according to their impact and our confidence in the numbers. These rules take into account such Key Performance Indicators as:
- Average number of tutors per pupil
- Percentage of girls on our programme
- Category of pupil feedback
- School profile data such as their Pupil Premium funding
- Pupil’s baseline attainment level
Analysing the feedback pupils give us
Text analysis of pupil comments at the end of each one-to-one Maths session allow us to quickly identify the most important issues and to act in time to improve the learning experience. The data helps us to find how different feedback types correlate with the pupil success. For example when pupils leave comments that demonstrate reflection and self-assessment, such as a request for more challenging tasks or a request for the next topic to be studied, this has a strong positive correlation with pupil success.
Aggregate data helps us to improve every lesson
With 2,400 pupils benefiting from our one-to-one Maths lessons every week, we are also able to learn a great deal from the aggregate data. By looking at the lessons teachers choose, and the level of progress in each lesson, we can find out not only which are the most popular i.e. the topics that children commonly struggle with, but also which are more challenging and children are likely to need more help with. This data is then used by our curriculum team of ex-teachers with expertise in UK and international (including Singapore) maths education to refine and improve the curriculum.
Third Space Learning Guide to Effective 1-to-1 Interventions
How to plan, manage and teach 1-to-1 and small group KS2 Maths interventions to make best use of the resources you have
Where do we go from here?
There are many exciting opportunities to apply research findings to the real life environment in order to deliver better structured lessons, to deal efficiently with pupils’ feedback etc. Since children learn in different ways, one of the possible research directions would be personalised education, which involve clustering of pupils by their learning behaviour, finding groups of tutors that achieve better results using particular teaching practices and, finally, map the child that learns more efficiently using one method with the tutor that teaches better with this method.
This is the level of personalisation that we aspire to – combining the knowledge and insight from data, with the flexibility and personal attention of a real person providing the one-to-one tuition.
We’ll let you know how we get on.