Duplicate records
Repeated rows or repeated identifiers
Duplicates can inflate totals and distort counts, averages, and category comparisons.
Data Analytics Case Study
Data Cleaning & Reporting
A representative practice case showing how inconsistent operational data can be profiled, cleaned, validated, prepared for analysis, and converted into a clear reporting structure. The workflow demonstrates transferable data-quality skills without presenting simulated work as a client engagement.
0
Quality Issues Covered
0
Workflow Stages
0
Validation Checks
Analytics Workflow Preview
From raw records to validated and report-ready data
Project Overview
Context, problem, objective, and ethical framing
Context
A simulated operational dataset contained records collected over time by different users. The information needed to be prepared before reliable summaries or dashboard views could be created.
Problem
Duplicates, missing values, inconsistent dates, category variations, extra spaces, mixed capitalization, and unusual values reduced the reliability of direct analysis.
Objective
Build a transparent workflow that identifies data-quality problems, applies controlled corrections, validates the cleaned structure, and prepares decision-ready outputs.
Project status
Practice project based on a simulated dataset and a representative cleaning scenario. No real client, confidential record, or unsupported business result is claimed.
Seven Data-Quality Problems
Common issues addressed by the cleaning workflow
Missing values
Blank or incomplete fields
Missingness must be measured and handled according to the meaning and importance of each field.
Inconsistent dates
Mixed formats and invalid entries
Dates require one standard format before reliable sorting, grouping, or trend analysis.
Category variations
Different labels for the same meaning
Examples include abbreviations, spelling differences, and inconsistent category names.
Spacing and capitalization
Hidden spaces and mixed text styles
Text normalization prevents visually similar values from being counted as separate groups.
Invalid data types
Numbers stored as text or mixed fields
Correct types are necessary for calculations, filters, joins, and visualizations.
Unusual values
Potential outliers or entry errors
Unusual values should be flagged and reviewed, not automatically deleted without justification.
Tools and Techniques
Resources applicable to the cleaning and reporting workflow
Excel / Google Sheets
Profiling, filters, formulas, and summaries
Spreadsheet tools support visual inspection, duplicate review, text normalization, missing-value checks, pivot summaries, and stakeholder-friendly reporting.
SQL
Queries, grouping, validation, and reproducibility
SQL can identify repeated keys, invalid values, missing fields, inconsistent categories, and aggregate results across structured tables.
Python
Repeatable transformations and quality checks
Python supports reusable cleaning rules, audit outputs, automated validation, and consistent preparation across multiple files.
Cleaning log
Transparent record of decisions and changes
A cleaning log documents the original issue, applied rule, affected field, validation result, and any decision requiring human review.
Six-Stage Analytics Workflow
Ask, prepare, process, analyze, share, and actAsk Define the reporting question and quality requirements
- Clarified the intended use of the dataset.
- Identified the fields needed for analysis and reporting.
- Defined which errors would materially affect the result.
Prepare Preserve the source and inspect the dataset structure
- Kept an unchanged copy of the raw source.
- Reviewed columns, identifiers, field types, and expected ranges.
- Created a working copy and a cleaning log.
Process Clean, standardize, and document transformations
- Reviewed duplicates and incomplete records.
- Standardized dates, categories, spaces, capitalization, and data types.
- Flagged unusual values for review rather than deleting them automatically.
Analyze Create reliable summaries from the validated structure
- Calculated descriptive totals and category distributions.
- Compared periods or operational groups where appropriate.
- Separated observed patterns from unsupported causal claims.
Share Communicate results with clear tables and visuals
- Selected only visuals that answer a reporting question.
- Added labels, units, definitions, and data-quality notes.
- Explained limitations and unresolved records transparently.
Act Convert findings into practical next steps
- Recommended improvements to future data entry.
- Proposed validation rules for high-risk fields.
- Prepared a reusable workflow for the next reporting cycle.
Before and After Cleaning
Structural improvement achieved by the workflow
Before: repeated records
Potential duplicates could inflate totals and produce misleading counts.
After: reviewed uniqueness
Duplicate rules and identifier checks separate confirmed repeats from legitimate similar records.
Before: inconsistent categories
Equivalent values appeared under different spellings, abbreviations, spaces, or capitalization.
After: standardized categories
A documented mapping creates consistent labels without losing the original meaning.
Before: mixed dates and types
Sorting, grouping, and calculation could fail when fields used incompatible formats.
After: analysis-ready structure
Dates and numeric fields use consistent types, with invalid values flagged for review.
Six Validation Checks
How cleaning quality is verified before reporting
Record-count reconciliation
Every removed or retained record must be explainable
Duplicate recheck
Confirmed repeats no longer remain under the selected rule
Missing-value review
Required fields and unresolved blanks are counted and documented
Format validation
Dates, numbers, and categories follow defined standards
Range and logic checks
Impossible or contradictory values are flagged
Summary comparison
Key totals are compared before and after transformation
Measurable and Documented Outcomes
Transparent measures based on the published case-study scope
7 issue categories
The workflow addresses duplicates, missing values, dates, category variations, spacing and capitalization, data types, and unusual values.
6 analytics stages
The project follows ask, prepare, process, analyze, share, and act as one complete workflow.
6 validation controls
Reconciliation, duplicates, missingness, formats, logical ranges, and summary comparisons are checked.
1 reusable cleaning method
The documented structure can be adapted to future spreadsheet, SQL, Python, or field-data projects.
Measurement note: these figures describe the documented scope of this representative practice case. They are not presented as results from a named client dataset or as a guaranteed business improvement.
Reporting Output
What a decision-ready deliverable should contain
Quality summary
Counts of detected issues, applied rules, unresolved cases, and validation outcomes.
Descriptive analysis
Clear totals, distributions, comparisons, and time-based summaries where the fields support them.
Visual communication
Charts or dashboard elements selected to answer defined operational questions.
Recommendations
Practical improvements for data entry, validation, documentation, and future reporting cycles.
Lessons Learned
Principles demonstrated by the case study
Preserve the raw source
Cleaning should remain reversible and auditable
A protected original copy makes it possible to verify changes and recover from mistakes.
Understand meaning before changing values
Context determines the correct cleaning rule
Similar-looking records are not always duplicates, and unusual values are not always errors.
Validate after every major transformation
Clean-looking data can still contain logical errors
Reconciliation and rule-based checks reduce the risk of silently changing valid information.
Communicate limitations
Reliable reporting includes uncertainty and unresolved issues
A professional report explains what was cleaned, what remains uncertain, and how that affects interpretation.
Skills Demonstrated
Technical and analytical capabilities shown in the workflowRelated Learning Evidence
This practice workflow is aligned with the structured analytics process developed through the Google Data Analytics Professional Certificate.
Need Cleaner, More Reliable Data?
I can help review spreadsheets, standardize operational records, document data-quality issues, prepare descriptive reports, and build clear dashboard-ready datasets.