Data Analysis Quick Reference

April 24, 2026

A compact reference for data quality, transformation, statistics, visualization, reporting, and governance.

On this page

Data quality and transformation

Issue	Appropriate response
Missing values	Remove, impute, flag, or investigate according to context and impact
Duplicates	Define the uniqueness rule before deduplicating
Inconsistent formats	Standardize units, dates, categories, currency, and casing
Outliers	Investigate before excluding; they can be errors or meaningful observations
Join mismatch	Check keys, cardinality, nulls, duplicate keys, and join type
Reproducibility	Record source, transformation steps, assumptions, and version

Concept	Useful distinction
Mean	Average; sensitive to outliers
Median	Middle value; more robust for skewed data
Mode	Most frequent value
Standard deviation	Dispersion around the mean
Correlation	Statistical relationship, not proof of causation
Confidence interval	Plausible range for an estimate under stated assumptions
p-value	Evidence against a null hypothesis, not a measure of business importance

Need	Suitable visual
Compare categories	Bar chart
Show a time trend	Line chart
Show a distribution	Histogram or box plot
Show a relationship	Scatter plot
Show ranking	Sorted bar chart
Show composition	Stacked bar; use pie charts only for simple, limited categories
Show geographic pattern	Map when location is central to the decision

Identify audience, decision, metric definition, time period, denominator, refresh cadence, and owner.
Label units, axes, filters, assumptions, and material limitations.
Apply least privilege, classification, retention, and approved sharing rules before publication.
Keep lineage from source through transformation to report so a result can be reproduced and challenged.
Check whether groups or edge cases are obscured by aggregation or biased source data.

Revised on Friday, September 11, 2026

Resources