Week 2 Assignment: Data Analysis Exercise
Week 2 Assignment: Data Analysis Exercise#
Due: End of Week 2 | Format: Short report + visualization | Length: 400-600 words
📝 Assignment Overview#
You’ll select a real dataset and analyze it through the lens of what you learned about data quality, representation, and how training decisions affect AI systems.
📋 Instructions#
Part 1: Dataset Selection & Exploration#
- Find a dataset that interests you (suggestions: Kaggle, UCI ML Repository, Google Dataset Search, or your field’s data repository)
- Download or access the dataset
- Document:
- What does the dataset contain?
- How many records/observations?
- What features/variables?
- Who collected it and why?
Part 2: Critical Analysis (400-600 words)#
Write a report addressing:
Data Quality:
- What is the source and credibility of this data?
- Are there missing values, outliers, or anomalies?
- How representative is this data of the real world? What’s missing?
Potential Bias:
- How might this data reflect historical biases?
- What groups or categories are overrepresented or underrepresented?
- How could these imbalances affect an AI model trained on this data?
Practical Implications:
- If an AI system were trained on this data, what real-world impacts could result?
- Who would benefit? Who might be harmed?
- How could the dataset be improved?
Part 3: Data Visualization#
Create one visualization (chart, graph, infographic) that communicates an insight about this dataset.
📌 Rubric#
| Criteria | Points | Description |
|---|---|---|
| Dataset Selection | 15% | Interesting choice, well-sourced, thoroughly explored |
| Critical Analysis | 50% | Thoughtful consideration of data quality, bias, implications |
| Visualization | 20% | Clear, informative, effectively communicates insight |
| Writing Quality | 15% | Well-organized, clear explanations, proper citations |
💡 Tips for Success#
- Choose a dataset related to something you care about
- Don’t just describe—analyze critically
- Link back to concepts from the readings
- Your visualization should tell a story
📤 Submission#
Submit a PDF report + visualization via D2L by 11:59 PM on [Due Date].
Need help? Visit the data analysis workshop or contact your instructor.