Anomalies threaten data reliability, so spotting unexpected values, judging their causes and deciding their influence is essential for valid biological conclusions in experimental work.

Identifying anomalies

An anomaly is a measurement that does not fit the pattern or trend shown by the rest of the data. It differs markedly from other replicates that were taken under the same controlled conditions. Anomalies weaken confidence in results because they distort calculated values such as means, rates, gradients or percentages, leading to flawed interpretations.

Anomalies are first detected by visually inspecting raw data during or immediately after collection. This supports the specification focus on identifying anomalies and considering their impact on analysis and conclusions. When examined early, steps can sometimes be repeated while the apparatus remains set up, improving data quality without altering the investigation.

Recognising anomalies in quantitative data

Students are expected to distinguish clearly between normal variation and genuine anomalies. All biological data contain natural variation, even when variables are controlled effectively. However, certain features indicate that a reading may be anomalous:

• A value lies far outside the range of other repeats

• A point on a graph breaks the established pattern, relationship or trend

• A result contradicts biological expectations based on scientific understanding

• A single measurement disproportionately changes a calculated mean

Graphical presentation is a powerful tool.

When plotted, anomalous points typically fall well above or below the line or curve of best fit.

A scatter graph with a single anomalous point that breaks the linear trend, illustrating how anomalies can be visually identified when they fall well outside the expected relationship. Source.

In OCR A-Level Biology, scatter graphs and line graphs are especially useful for identifying measurements that do not align with a correlation. Candidates must then judge whether the point should be included, excluded or repeated.

Judging the impact of anomalies

Anomalies can influence experimental outcomes in several ways. Their impact should be evaluated before moving to analysis or interpretation. Key effects include:

• Distortion of averages, making central values unrepresentative

• Reduction in precision, increasing data scatter

• Altered gradients, affecting calculated rates in biological investigations

• Misleading trends, leading to incorrect biological conclusions

If an anomaly significantly changes the outcome, its effect must be acknowledged. The OCR specification requires students to consider their impact on analysis and conclusions, so discussion of anomalies should be explicit when presenting processed data and final statements.

Differentiating anomalies from outliers

Although often used interchangeably, the terms are not identical. An anomaly is identified within the biological context of an experiment, whereas an outlier is a mathematically extreme value. A reading can be an outlier without being anomalous, especially where natural variation is expected. For example, biological samples such as enzyme activity or cell counts often show wide variability. Only when a value is inconsistent with experimental conditions or biological reasoning should it be considered anomalous.

Causes of anomalies

Anomalies typically originate from experimental error, not true biological change. Common causes include:

• Measurement error, such as parallax errors when reading scales

A diagram showing correct eye-level reading of a meniscus, helping prevent parallax error that can lead to anomalous measurements in liquid volume experiments. Source.

• Instrument faults, for example an uncalibrated pH meter or balance

• Inconsistent technique, including timing errors or varied sample handling

• Contamination, such as unintended microorganisms in a culture

• Uncontrolled variables, where a factor has drifted without detection

Identifying the likely cause strengthens evaluations and suggests meaningful improvements.

Deciding how to handle anomalies

Once identified, anomalies should not be ignored. OCR expects students to justify their treatment of unusual data by referencing its effect on results. Appropriate responses include:

• Repeating the reading when time and resources allow

• Excluding the value from calculations if evidence shows it is invalid

• Keeping the value visible in a table or graph but marking it clearly

• Commenting on its presence when writing processed data and evaluations

At A-Level, values removed from a mean should still be recorded to maintain transparency and good scientific practice.

Supporting anomaly detection with replicates

Replicates are essential for spotting anomalous values because they reveal whether a result is reproducible. If one reading differs greatly from two or more repeats, it is more confidently identified as anomalous. This strengthens reliability and provides evidence-based justification for exclusion or repetition.