OCR Level 3 IT: Data Analytics Past Papers

The Fundamentals of Data Analytics unit rewards precise use of data terminology. 'Structured data' has a defined schema (relational databases, spreadsheets); 'unstructured data' lacks predefined format (emails, images, social media posts). Data quality dimensions matter for exam answers: when evaluating a dataset, address accuracy (are the values correct?), completeness (are any values missing?), consistency (are formats standardised?), and timeliness (is the data current?). Questions often present a dataset with deliberate quality issues and ask you to identify and address them. For relational database content, normalisation is a key concept. First normal form (1NF) eliminates repeating groups; second normal form (2NF) removes partial dependencies on a composite primary key; third normal form (3NF) removes transitive dependencies. Be able to identify anomalies in an unnormalised table and explain which normal form rule is violated. SQL SELECT statements with WHERE, ORDER BY, GROUP BY, and JOIN clauses appear in many assessments — practise writing queries from scratch given a described database schema. Ethics in data analytics is examined in the context of algorithmic bias and GDPR compliance. Algorithmic bias occurs when training data reflects historical inequalities, causing machine learning models to perpetuate discrimination — know a concrete example (facial recognition accuracy disparities, hiring algorithm bias). GDPR principles apply to personal data in analytics pipelines: data minimisation means only collecting necessary attributes; purpose limitation means not reusing personal data for new analyses without consent.