Data Cleaning and Preprocessing

Active learning works for data cleaning and preprocessing because students need to experience the frustration of messy data to understand why cleaning matters. Working with real, imperfect datasets helps them see the direct impact of their decisions on analysis quality.

• Grade 9
• Computer Science
• 3 activities
• 15 min–45 min

Learning Objectives

1. Explain the necessity of data cleaning for accurate and reliable data analysis.
2. Identify common data errors such as missing values, outliers, and inconsistent formats.
3. Design a systematic approach to detect and correct errors in a given dataset.
4. Evaluate different strategies for handling missing data, considering potential biases.

Gallery Walk

⏱ 45 min·Whole Class

Gallery Walk: Data Storytelling

Groups create large-scale visualizations of a local issue (e.g., cafeteria waste or local transit times). They display their charts around the room, and other students use sticky notes to write down one 'story' or 'trend' they see in the data.

Prepare & details

Explain why data cleaning is a crucial step before data analysis.

Facilitation Tip: During the Gallery Walk, circulate and ask students to explain why they chose specific visualizations for their datasets rather than telling them if they are correct.

Setup: Wall space or tables arranged around room perimeter

Materials: Large paper/poster boards, Markers, Sticky notes for feedback

• Understand
• Apply
• Analyze
• Create
• Relationship Skills
• Social Awareness

Inquiry Circle

⏱ 30 min·Small Groups

Inquiry Circle: The Bias Hunt

Provide groups with three different graphs of the same data set, each using a different scale or chart type. Students must figure out which graph is the most 'honest' and which ones might be trying to mislead the viewer.

Prepare & details

Analyze common types of data errors and inconsistencies.

Facilitation Tip: For The Bias Hunt, provide printed survey questions so students can physically mark language that might lead respondents toward certain answers.

Setup: Groups at tables with access to source materials

Materials: Source material collection, Inquiry cycle worksheet, Question generation protocol, Findings presentation template

• Analyze
• Evaluate
• Create
• Self-Management
• Self-Awareness

Think-Pair-Share

⏱ 15 min·Pairs

Think-Pair-Share: Ethical Collection

Students are given a scenario where a new app wants to collect their location data. They discuss with a partner: What is the benefit to the user? What is the risk? Is the collection ethical?

Prepare & details

Design a strategy to address missing or erroneous data in a given dataset.

Facilitation Tip: In Think-Pair-Share, assign roles: one student explains ethical collection principles, the other identifies potential violations in a given scenario.

Setup: Standard classroom seating; students turn to a neighbor

Materials: Discussion prompt (projected or printed), Optional: recording sheet for pairs

• Understand
• Apply
• Analyze
• Self-Awareness
• Relationship Skills

Teaching This Topic

Teachers should model data cleaning with think-alouds, showing how they decide to standardize categories or handle missing values. Avoid the trap of treating data cleaning as a mechanical task. Emphasize that every decision reflects assumptions about what counts as valid data. Research shows students grasp these concepts better when they work with datasets they care about, so incorporate student-generated data when possible.

What to Expect

Successful learning looks like students confidently identifying data issues, justifying their cleaning choices, and explaining how those choices affect the stories their visualizations tell. They should connect technical steps to ethical and practical implications.

Watch Out for These Misconceptions

Common Misconception

During the Gallery Walk, watch for students assuming their visualizations are correct because they look polished.

What to Teach Instead

Have peers ask presenters to explain how each choice of chart type connects to the data's structure and purpose. Use a simple rubric during the walk to guide their feedback.

Common Misconception

During The Bias Hunt, students may think bias only comes from obvious wording like 'Do you agree that this is the best plan?'.

What to Teach Instead

Provide examples of subtle bias, such as leading scales or double-barreled questions, and have students rewrite these questions to remove bias, then discuss why their versions are better.

Assessment Ideas

Exit Ticket

After the Gallery Walk, provide students with a messy dataset and ask them to identify two data issues, explain why each matters for analysis, and propose one cleaning method for each issue.

Quick Check

During The Bias Hunt, listen for students recognizing that even minor wording changes can shift responses. Pause the activity to ask a pair to share their revised question and explain how their changes reduce bias.

Discussion Prompt

After Think-Pair-Share, facilitate a class discussion using the prompt: 'A dataset includes responses like 'yes', 'y', 'Y', and 'Yeah'. What are the implications for analysis, and what standard would you set for these responses?' Use student responses to assess their understanding of standardization.

Extensions & Scaffolding

• Challenge: Provide a dataset with deliberate outliers and ask students to research industry-standard methods for handling them (e.g., capping, imputation, or removal) and justify their approach.
• Scaffolding: Give students a checklist of common data issues (missing values, inconsistent formats) with examples to reference while cleaning their dataset.
• Deeper exploration: Have students find and analyze a real-world example of data bias in a published report or news article and present how they would redesign the data collection to reduce that bias.

Key Vocabulary

Data Cleaning	The process of detecting and correcting (or removing) corrupt or inaccurate records from a dataset. It ensures data quality for analysis.
Missing Data	Values that are not recorded or present in a dataset. Handling missing data is crucial to avoid skewed results.
Outlier	A data point that differs significantly from other observations. Outliers can be due to measurement error or represent genuine extreme values.
Data Inconsistency	When data values that should be the same are different, such as variations in spelling or formatting for the same category.
Data Validation	The process of ensuring data is accurate, complete, and conforms to defined rules or constraints before analysis.