Teach data handling as a complete evidence-based workflow rather than a list of isolated commands. This Unit 03 pack transforms Chapter 3 of AI & Machine Learning: Foundations & Data – Volume 1 into four structured, classroom-ready lessons covering data quality, array structure, shape, dtype, vectorisation, broadcasting, DataFrame inspection, missing-data decisions and clean-data pipelines.

Students learn to inspect before trusting data, identify problems using evidence, choose appropriate checks and justify cleaning decisions before visualisation or modelling. They use NumPy to reason about arrays and compatible operations, and Pandas to connect head(), shape, dtypes, isna().sum(), describe() and value_counts() to specific questions and limitations. Missing values, suspicious ranges, duplicates, inconsistent categories and mixed data types are explored through practical scenarios.

Designed as the classroom teaching companion to AI & Machine Learning: Foundations & Data – Volume 1, the package includes a Teacher Guide, four-lesson plan, detailed Teacher Handbook, Student Summary Notes, printable Worksheet, full Answer Key, Activity Cards, Exit Tickets, a 34-slide presentation deck and a Read Me First guide. Resources support upper-secondary computing, introductory data science, AI and machine learning, vocational education, enrichment and homeschool settings.

The no-prep structure supports direct teaching, group work, assessment and independent revision. The central message is simple: good machine learning starts with good data handling.

By the end of this unit, students will be able to:

• Explain why data quality matters before modelling
• Identify missing values, duplicates, suspicious ranges and inconsistent categories
• Explain NumPy arrays using shape, dtype, axes and vectorisation
• Predict when broadcasting is compatible or likely to fail
• Explain what common Pandas inspection commands reveal
• Distinguish what a check shows from what it does not prove
• Compare possible responses to missing data
• Justify cleaning decisions using evidence, risk and context
• Build an inspect → decide → clean → document → re-check workflow
• Communicate remaining limitations before visualisation or modelling

Lesson 1 – Data Quality First
Inspect messy data, identify issues, cite evidence and explain risk before cleaning.

Lesson 2 – NumPy Array Thinking
Explore shape, dtype, vectorisation, broadcasting and meaningful array operations.

Lesson 3 – Pandas Inspection
Connect inspection commands to evidence, decisions and limitations.

Lesson 4 – Missing Data and Clean Pipelines
Compare cleaning strategies, document decisions and re-check results before visualisation.