Analyse patterns, trends, relationships and anomalies

Students distinguish pattern, trend, relationship and anomaly; support descriptions with data; use mean/range where useful; investigate unusual results before exclusion; and avoid claiming causation from correlation alone.

• Pattern: repeated arrangement or regularity.
• Trend: overall direction through ordered data.
• Relationship: how one variable changes with another.
• Anomaly: a value inconsistent with the overall pattern.

Masses 6 kg and 12 kg moving at the same speed have kinetic energies 12 J and 24 J.

Model: When speed is held constant, kinetic energy increases with mass; doubling mass from 6 to 12 kg corresponds to doubling kinetic energy from 12 to 24 J in these data.

Means can summarise repeated trials, while range shows spread. Always inspect raw repeats: an average can hide variation or an anomaly.

One high value is not itself a trend. A flat mean does not prove every repeated measurement was identical.

1. Check transcription and units.
2. Check instrument and method.
3. Repeat or compare with other trials where possible.
4. Consider whether the point could be a real rare event.
5. Retain and flag it, or justify exclusion transparently.

Never delete an unusual value merely because it makes the graph look untidy.

If two variables change together, describe an association unless the design and controls support a causal conclusion. Consider confounding variables and alternative explanations.

We do: analyse a noisy plant-growth dataset: state trend, calculate a mean, identify an anomaly and decide what to check.

You do: write a complete analysis sentence for an unfamiliar graph, including two values and one qualification.

Exit ticket: Explain why “remove every outlier” is poor scientific practice.