To paraphrase Tolstoy "Happy answers are all alike; every unhappy answer is unhappy in its own way".
As we are collecting typed answers, incorrect (ie amended) answers are a potentially rich source of data for learning interventions.
We screen amended answers with AI for a number of 'interesting' errors, which are flagged if an error is made repeatedly within the timeframe (by default, 7 days). We can flag misunderstandings and even infer the likely presence of learning difficulties such as ADHD, dyslexia and dyscalculia.
© Copyright Influenca Ltd 2020 - 2025