There is little doubt that, if UK schools don’t embrace Artificial Intelligence, they will soon find it embracing them. Quite how this will pan out is impossible to say. But, if you, like so many other school leaders, are wondering what the likely short term impact might be and how to get started engaging with this new, potentially powerful, agent. Here are a few suggestions…
The first one is that engaging with Earwig would be a good place to start. Here are three reasons why…
Earwig software already provides facilities for teachers covering most of the activities listed in the previous section - and these are the school activities where AI is likely to be most useful to teachers in the short term.
It has been working with hundreds of special schools for ten years, with the express aim of providing teachers with a single package that replaces several other applications, each of which provides more limited scope.
So employing AI to make life easier and more productive for teachers is just continuing the process that we started a long time ago..
As we have already seen, AI is only as good as the data that it bases its decisions on.
Of course there is lots of data in the public domain about special needs teaching, lesson content and performance management. But Earwig has a great deal more.
Hundreds of times every day a teacher creates an Earwig pupil performance record or grades a pupil, adding images, commentary and a great deal more.
Currently we hold more than 10 million records of the small achievements of special needs pupils throughout England and Wales which are being added to, in this way, every school day.
Any school whose staff use an Earwig AI ‘assistant teacher’ will therefore immediately benefit from the combined experience of more than 300 special schools, 10,000 teachers and 30,000 pupils and 10 years of performance history, based on current British SEND pedagogy and processes.
AI is not a conventional tool. It is an ‘agent’ which can make (and change) decisions on its own. This makes it an uncomfortable ally in the battle to improve SEND teaching and learning - especially for teachers who are not especially proficient with software generally.
The key to gaining ready acceptance for these upgrades is to introduce the improvements piecemeal, adding a process improvement here and a new feature there. So that, as time goes by, teachers find that they have armed themselves with a powerful bank of AI based agents without the stress of incorporating a whole new way of working in one go.
So, whether you are already using Earwig or have yet to get involved, Earwig is probably your ideal partner in the journey towards better outcomes for pupils through AI supported teachers.
Every journey starts with the first step and now it’s time to consider which of the many software tools that teachers currently use might be first in line for an AI upgrade.
Here are fourteen different Earwig functions and new facilities which we could upgrade or develop to help your staff to do their jobs better or more easily. Which ones should we be focusing on?
We have divided the options into three sections, from small projects to the bigger ones. Please read through these and, at the ned, let us know which one you think we should be focusing on to provide optimum assistance to teachers in the minimum time.
These are features that could be developed, trained and applied within twelve months - if school SLTs felt that they would be valuable additions to the current Earwig packages. What we need to understand from clients is which should be prioritised.
a. Commentary suggestions to make recording teaching evidence easier, better
With the Earwig app, the only thing that takes any time when creating an evidence record is writing the commentary, which is the core of the record. Earwig AI will write a draft commentary in each Earwig record, based on the history, the images and other data captured. This can be edited by teachers, either manually or verbally.
b. Use facial recognition to make recording teaching evidence faster, more secure
Facial recognition can connect Earwig records automatically to the pupils recognised in any image or video clip. This can speed up record creation. It can also obscure faces of children who are flagged as ‘Internal only’ for any reason. However, no facial recognition system is watertight. So it will still require oversight from teaching staff.
c. Translating commentary and reports for EAL/SEN parents.
For parents with limited English, translate into any language with a click.
d. Writing pupil end-of-year reports
Based on the records and commentary generated for each pupil during the year and the associated assessment and tracking data, Earwig AI can produce better year-end summaries of subject and overall performance than most teachers, with no staff effort. These, like all the other commentaries, must be reviewed and edited by relevant teachers, but the time and effort saved should be considerable and the result more complete.
e. Proposing new EHCP- based objectives and targets for each child
Currently, teachers have to analyse each new EHCP Outcome and laboriously create a set of Long Term (Annual) Objectives (IEPs) from these. Then, a further set of Short Term (Termly) targets for each objective.
Earwig AI will draw from that child's detailed history and also on similar experience with thousands of other pupils with similar disabilities and interests and targets - to draft a set of proposed targets to be selected and edited by the teacher.
f. Setting Expectations
Pupil performance is not only measured against ECHP targets. Most pupils are taught against a whole raft of other conventional curricula and frameworks. Because the achievement profiles of special needs pupils are usually spiky, it is considered good practice for teachers to set expectations for their SEN pupils by estimating the level that each one might achieve for each subject during the coming school year.
Given sufficient historical data, Earwig AI will propose each subject expectation, get approval from the teacher and then measure performance against this individual target score for the rest of the year.
g. Next Steps 2.0
Using the existing pupil performance information, select the most suitable Next Steps for each pupil from the target statements in each relevant curriculum framework. Update daily and provide these to subject leaders to review and teachers to pin on the wall, daily or weekly.
These are material upgrades to the existing functionality which would require more time and more development resource - but may be considered sufficiently useful and important to be worth prioritising.
h. Proposing each year's EHCP Required Outcomes for the Annual Review meeting
Take one step further up the target setting process to read inputs from therapists, social workers, parents, teachers etc. Evaluate these in the light of thousands of similar processes for children with similar needs to propose the Outcomes to be considered by the Local Authority with the school and other professionals at the Annual Review.
i. Lesson Plans
A major upgrade to the existing Earwig 'Next Steps' facility through which the system conceives and lays out a sequence of more individualised, interesting and useful lessons for each pupil for consideration by the teacher and evolves this on the basis of each performance record.
j. Performance Tracking - Anomalies Report
Much more detailed analytics reports for school management.
Provide a weekly report for school and trust leadership, highlighting achievements and performance gaps during the previous period.
k. Next Level Parent Engagement
Alert parents to recent activity at school, suggest activities to do at home to support learning and provide answers to questions on their child's school performance.
The features laid out below would require a substantial amount of development work, and therefore time. Currently, other edtech software houses provide packages which cover these areas. But it may make sense, in the long term, to integrate some or all of these into one application because of the considerable overlap with regard to the data that they use to in order to be effective.
l. Providing Lesson Content
This is a big one. There are already many lesson content providers, some of which are already using AI. But these are usually siloed in some way which means that they are only used for certain subjects or situations. Is there scope for a system to provide more individualised SEN lesson content, across curriculum areas, based on pupil progress and individual inclinations, strengths and weaknesses?
m. Tracking Behaviour
We have been asked many times whether or not we will be introducing a Behaviour Tracking Module. This is because of the high degree of overlap between the data required for behaviour tracking and performance tracking. So it would appear to be a logical extension to the existing Earwig application and avoid having to use two different software packages.
n. Provision Mapping and Tracking
Lining up pupil needs with provisions available – and tracking the effectiveness of interventions is the sort of application that AI can handle very efficiently. The effect of the interventions themselves are often already being tracked using Earwig. So, once again, this would appear to be a logical extension and save school SLTs a lot of time and effort.
o. Safeguarding Management
We have been asked many times whether or not we will be introducing a Safeguarding Module. The same logic applies here as to the Behaviour Tracking option. The high degree of overlap between the data required for safeguarding and the data already input into Earwig would appear to make it a logical extension to the existing functionality.
If you are a special school leader or SENCO and you have a view about how AI could be harnessed in the short term to make special needs teaching, easier, more efficient or better, we would love to hear from you.
Just click on the button and take ten minutes to give us your opinion.
We will collate the responses and let you know what the consensus is.