The process step by step

1) Evidence-based Assessment

Our unique assessment is as simple as it is effective and takes just 2 to 5 minutes per pupil! Pupils read two short passages of text from a monitor with an eye tracking device attached, one out loud and one silently.  Each passage is then followed by some comprehension questions.

By monitoring spontaneous eye movement we can quickly analyse the key reading components and gain a detailed insight into how reading ability varies when pupils process text across different levels.

   Current reading assessments are simply not good enough. When we look at a standard test score in isolation, it tells us a child can’t read. It doesn’t tell us why that child can’t read. Lexplore Analytics gives us much more than a simple score or percentile.  

Bernadette McLean – Former Principal of the Helen Arkell Dyslexia Centre

As our assessment takes a completely different format to most traditional tests pupils tend to thoroughly enjoy their experience; many even ask if they can sit it again! By replacing more time-consuming assessment methods ,we can help decrease the overall assessment time for pupils as well eliminating marking and administration time for teachers.

   I found it more enjoyable than other tests that we do because it was fun, short and less tiring. The stories were interesting and I wanted to read them to find out what happened.  

Pupil – KS2


2) Machine Learning Analysis

Once a pupil has sat our assessment the recordings are uploaded to a secure, cloud-based platform. Machine learning analysis is then able to provide teachers with immediate and accurate results without the need for any marking, paper work or manual evaluation.

”    200 children went through the assessment process in a week. Within five minutes we had results for reading activity and an accurate attainment level. Previously, it took teachers an entire term to do accurate assessments on their whole class.  

Adam Luxford – Freemantle Community Academy

Developed from extensive and prolonged research, these models are already performing at a very high accuracy level ( around 96% ). They also continue to learn from new data, meaning the more our technology is use the more accurate the overview of reading it is able to give.

Our machine learning platform is also able to take into account the natural progress a child would make though an academic year, so even minor percentiles progressions equate to real progress following interventions.


3) Results and Intervention

Within our secure results portal teachers can find a percentile score, a reading age, a year level equivalent and a standard score for each pupil. They can also see a unique visual representation the reading process and for the first time what reading silently looks like.

Individual measures are also displayed alongside previous results, the school average and the national average for each reading component, including reading speed and fixation time. This enables teachers to monitor the most minor progress at a granular level.

”     The tests showed one of our KS2 girls was finding reading difficult but this was something we hadn’t spotted before. She had developed effective coping strategies to manage those difficulties, we hadn’t realised that she needed support 

Jayne Mullane – Headteacher – Mersey Vale Primary

Reading speed in particular has proven to be useful for teachers building a picture of candidate needs for exam access arrangements. Therefore, results for both aloud and silently are displayed alongside a pupil’s overall reading profile.

Our machine learning platform is also able to provide a sticky words list for each pupil based on their fixation time. Teachers can then clearly see whether pupils are struggling to decode particular graphemes or phonemes.

Teachers can then work with our dedicated interventions package, Lexplore Intensive, to support all pupils, both emergent and fluent readers, inline with their attainment level. With regular testing schools can closely monitor progress and ensure interventions are having the intended impact.

As well as individual reading profiles, the our Results Portal also provides a powerful overview of reading across multiple levels from a multi-academy trust all the way down to an individual class.

”     The presentation of results is clear, visual and easily accessible both with regards to class, whole school and individual results. This enables schools to easily share results with staff, governors, parents and Ofsted. Being able to see the child’s eye movements has also been fascinating, insightful and at times mesmerising.   

Adam Luxford  – Freemantle Community Academy


The Research The Technology Why Lexplore?