Data PitfallsAvoiding Data Pitfalls 78 rules
PITFALL LIBRARY
Every pitfall in the taxonomy, grouped into the book’s eight domains. Pick a domain to read what each pitfall is, how to spot it and how to fix it.
6Epistemic ErrorsHow we think about data: mistaking the record for reality, and claiming more than it can show.
8Technical TrespassesJoins, nulls, types and encodings that quietly corrupt the data before analysis starts.
11Mathematical MiscuesArithmetic that goes wrong: unweighted rates, mixed units, percent versus percentage points.
- Division by Zero or Tiny Denominator
- Dropped Missing Categories
- Embedded Total Rows
- Floating-Point Accumulation Error
- Integer Division Truncation
- Misreading a Log-Scale Average
- Mismatched Units of Measure
- Partial-Period Aggregation
- Percentage Points vs Percent Change
- Uneven Aggregation Coverage
- Unweighted Average of Rates
13Statistical Slip-UpsMisreading samples, averages and significance, from survivorship bias to Simpson’s paradox.
- Mean Mistaken for Typical
- Uncorrected Multiple Comparisons
- Nonresponse Bias
- Unchecked Normality Assumption
- Omitted Uncertainty
- P-Value Misinterpretation
- Regression to the Mean
- Simpson's Paradox
- Small-Sample Extremes
- Statistical vs Practical Significance
- Survivorship Bias
- Unrepresentative Sample Comparison
- Untested Significance Claim
11Analytical AberrationsReasoning errors: correlation read as causation, extrapolating too far, gaming a metric.
- Uncontrolled Confounding Variable
- Correlation Implies Causation
- Goodhart's Law (Metric Gaming)
- Intuition/Analysis False Dichotomy
- Misleading Interpolation
- Misleading Performance Metric
- Overfit Model Extrapolation
- Spurious Correlation
- Trend Extrapolation Beyond the Data
- Unvalidated Clustering
- Wishful Baseline Forecast
18Graphical GaffesCharts that distort: truncated axes, dual axes, 3D, and the wrong chart for the message.
- 3D Chart Distortion
- Chart Omits Key Context
- Chart Type Dogmatism
- Chart Type Message Mismatch
- Cherry-Picked Chart Window
- Disproportionate Encoding
- Dual-Axis Distortion
- EDA Noise Mistaken for Signal
- Encoding Order Mismatch
- Incoherent Axis Scale
- Inconsistent Scales Across Charts
- Insufficient Precision for Comparison
- Optimize/Satisfice False Dichotomy
- Overloaded Confusing Chart
- Rainbow Color Scale
- Task-Mismatched Visualization
- Truncated Y-Axis
- Unevenly Spaced Axis
10Design DangersDesign choices that confuse or exclude: unsafe color, low contrast, unclear interactions.