Our Approach
What we consider
Qwixl looks at multiple signals from routine homework to help teachers and SENCOs spot learners who may benefit from further support. We combine four types of information:
How students type
We observe patterns such as typing speed, how often students pause while writing, and how often they correct or delete text. These patterns are gathered naturally during normal assignment completion—no separate tests required.
What students write
We look at spelling, grammar, sentence structure, and phrasing in the written work they submit. This helps build a fuller picture alongside typing behaviour.
Patterns over time
Repeated patterns across multiple assignments strengthen our understanding. A single difficult task might reflect an off day; consistent patterns over weeks or terms are more meaningful.
How students use the tutor
We observe how learners interact with their personal tutor—what they ask for help with, topics they return to, and areas where they need repeated scaffolding. These observational insights complement typing and writing signals.
How we use this information
We do not react to a single piece of work. Instead, we gather data over time from multiple assignments:
Multiple submissions
Signals are collected and compared across several assignments before any indicators are raised. This reduces the impact of one-off factors like a bad day or unfamiliar topic.
Context matters
We take into account whether students are typing on a keyboard or touch screen, and whether they are in a group where English is an additional language. These factors can affect how we interpret the signals.
Supporting, not replacing
Our indicators are screening aids. They are designed to add to a teacher or SENCO's understanding—not to replace observation, conversation, or professional assessment.
Learning gaps and tutor support
We identify concepts or topics where a learner may need extra support. The personal tutor uses this to tailor its help—adapting how it communicates based on each student's learning profile, strengths, and areas for development.
Improving over time
When a SENCO confirms a diagnosis, that understanding helps us improve the accuracy of our insights for all students. We learn what works, so early identification gets better for every school using Qwixl.
We do not react to a single piece of work. Instead, we gather data over time from multiple assignments:
Multiple submissions
Signals are collected and compared across several assignments before any indicators are raised. This reduces the impact of one-off factors like a bad day or unfamiliar topic.
Context matters
We take into account whether students are typing on a keyboard or touch screen, and whether they are in a group where English is an additional language. These factors can affect how we interpret the signals.
Supporting, not replacing
Our indicators are screening aids. They are designed to add to a teacher or SENCO's understanding—not to replace observation, conversation, or professional assessment.
Learning gaps and tutor support
We identify concepts or topics where a learner may need extra support. The personal tutor uses this to tailor its help—adapting how it communicates based on each student's learning profile, strengths, and areas for development.
Improving over time
When a SENCO confirms a diagnosis, that understanding helps us improve the accuracy of our insights for all students. We learn what works, so early identification gets better for every school using Qwixl.
Student wellbeing
Every interaction a student has with Qwixl is designed with their safety in mind:
Safeguarding in the tutor
The personal tutor monitors every conversation for signs that a student may need help beyond schoolwork. If a concern is identified, the student is shown how to access support, and their teachers, parents, and SENCO are alerted.
Conversational oversight
Parents, teachers, and SENCOs can review any tutor conversation at any time. Nothing is hidden—if the tutor and a student discussed something, the adults responsible for that child can see it.
Neutral language
Students never see clinical labels or risk scores. The language used throughout Qwixl is positive, encouraging, and focused on learning—not on categorisation.
What we do not do
We do not diagnose
Only qualified professionals can diagnose conditions such as dyslexia or ADHD.
We do not label students
We highlight patterns that may be worth investigating, leaving interpretation to the school.
We do not make decisions
Teachers and SENCOs decide on next steps—whether that's observation, further screening, or referral.
Alignment with good practice
Our approach is informed by research and statutory guidance:
Early identification
As emphasised in the SEND Code of Practice, to help learners get support sooner.
The Assess–Plan–Do–Review cycle
Our signals feed into the assess and review stages, with structured support for recording interventions and review dates.
Multi-method identification
Combining different sources of information, rather than relying on a single test.
A learning-first approach
The personal tutor adapts to how each student learns best, building on their strengths while supporting areas of difficulty.