Data Detective
Find missing values, repeated records and unusual numbers. Calculate means, compare cleaning decisions and keep an evidence trail.
Investigate a dataset →Four connected projects take you from raw data to responsible AI evaluation. Suitable for KS3 extension, GCSE preparation and independent post-16 exploration.
Start with messy data, build a prediction, explore a trained classifier and explain what the results justify. The activities use fictional or clearly credited educational datasets, run locally in the browser and collect no student details.
Find missing values, repeated records and unusual numbers. Calculate means, compare cleaning decisions and keep an evidence trail.
Investigate a dataset →Fit a straight-line regression model using the fictional bike-hire dataset from Chapter 18, then compare it to a baseline on eight unseen days.
Train and evaluate →Use k-nearest neighbours to classify fictional sensor readings. Change k, inspect false alarms and missed positives, then evaluate limitations.
Explore a trained classifier →Complete original self-check questions and a printable project brief. Explain your cleaning, train/test separation, evaluation and ethics.
Test and evidence your learning →AI & Data Science Foundations — chapters 5–9, 13–14, 16–18 and 20 support the learning route.
All three investigations come with matching commented Python programs and sample CSV files, plus a learner worksheet and teacher notes.
Download v26 project pack (ZIP) →Choose an appropriate research question, record provenance, compare baselines and document limitations. These tasks are EPQ preparation, not a qualification.
Open EPQ research starter →See where data analysis, model development and transparent evaluation are used professionally.
Data scientist profile →These are original TutorSyntax enrichment resources informed by emerging computing education themes. They are not official Ada, Raspberry Pi Foundation, examination-board or EPQ assessment materials. Current GCSE specifications remain available elsewhere on TutorSyntax.