Identifying Data Patterns
Recognising patterns in data helps you answer numerical reasoning questions faster and avoid calculation errors. Here's how to spot common patterns in tables, charts, and graphs.
Common Patterns
Upward trend – Values increase over time. Steady growth suggests a consistent rate; accelerating growth means the rate is increasing.
Downward trend – Values decrease over time. Look for the rate of decline.
Seasonality – Values repeat in cycles (e.g. higher in Q4, lower in Q1). Useful when predicting or comparing same periods across years.
Proportional relationships – When one value doubles, another doubles (or halves). Common in ratios and percentages.
Outliers – Unusual values that don't fit the pattern. May indicate an error or a special event. Don't let one outlier skew your interpretation.
How to Use Patterns
- Estimate – If data follows a clear trend, you can estimate values between known points.
- Spot errors – If your calculated answer doesn't fit the pattern, double-check.
- Prioritise – When time is short, questions about the overall trend may be quicker than precise calculations.
- Eliminate wrong answers – If options don't match the pattern (e.g. growth when data shows decline), rule them out.
Quick Checks
- Are values generally increasing or decreasing?
- Is the change consistent or accelerating?
- Are there any obvious spikes or dips?
- Do proportions stay roughly the same across categories?
Practice with numerical reasoning questions and the numerical reasoning test.
Frequently Asked Questions
What if the pattern isn't obvious?
Focus on the specific data the question asks for. Don't spend too long looking for patterns—sometimes the data is simply irregular. Calculate what's needed.
How do I spot seasonality?
Compare the same period across years (e.g. Q4 2022 vs Q4 2023). If values are similar or follow a trend, while other quarters differ, you may have seasonality.
Should I always look for patterns first?
A quick scan can help, but don't over-invest. If the question asks for a specific value, go straight to the data. Use patterns to verify or estimate when useful.