Once data is collected and represented, the real work of analysis begins: looking for patterns (a consistent relationship, like "height increases as sunlight increases"), trends (an overall direction, like a steady rise or fall over time), and anomalies (data points that don't fit the pattern — like one plant in the low-sunlight group that grew unusually tall). Anomalies are worth investigating rather than ignoring — they might reveal a measurement error, an unaccounted-for variable, or occasionally something genuinely interesting the original hypothesis didn't predict.
Example
A graph of plant height against sunlight hours shows a clear upward trend — but one plant that received little sunlight still grew tall (an anomaly). Investigating further reveals that plant was accidentally given extra fertiliser, explaining the unexpected result without disproving the overall sunlight-height pattern.
Key terms
Trend:
An overall direction or pattern in data over time or across a variable.
Anomaly:
A data point that doesn't fit the overall pattern.
Questions
1. A pattern in data is:
A consistent relationship between variables
A completely random, meaningless result
Something that never appears in real data
Always caused by a mistake
2. A trend describes:
An overall direction in data, like a steady rise or fall
A single, isolated data point
Data with absolutely no direction
A type of measuring equipment
3. An anomaly is:
A data point that doesn't fit the overall pattern
The most common, typical data point
Something that should always be deleted immediately
A type of graph
4. When you notice an anomaly in your data, you should:
Investigate it rather than ignoring it
Always delete it immediately with no investigation
Assume it disproves the entire investigation
Ignore it completely with no further thought
5. An anomaly might be caused by:
A measurement error or an unaccounted-for variable
Nothing at all, ever
Only a computer malfunction
A change in the researcher's mood
6. A steady upward trend in a graph over time suggests:
A consistent increase in the measured variable
No relationship between the variables
A random, meaningless pattern
Data has definitely been recorded incorrectly
7. Analysing data mainly involves:
Looking for patterns, trends and anomalies
Ignoring the data entirely after collecting it
Randomly guessing a conclusion
Deleting any inconvenient results
8. Why might investigating an anomaly (rather than deleting it) sometimes lead to a genuinely interesting discovery?
The anomaly could reveal an unaccounted-for factor or a pattern the original hypothesis didn't predict
Anomalies never reveal anything useful about an investigation
Deleting anomalies is always the correct scientific approach
Investigating unusual results is a waste of scientific time
9. In the plant example, discovering the anomalous plant received extra fertiliser helps explain the unusual result by:
Identifying an unaccounted-for variable affecting that specific data point
Proving the entire sunlight-height pattern is false
Showing that fertiliser has no effect on plant growth
Confirming there was no explanation at all for the anomaly
10. Why is it important to distinguish between a genuine trend and random noise (natural variation) in data?
A real trend suggests an underlying cause, while noise could just be normal variation with no meaningful pattern
Noise and genuine trends are always identical and cannot be distinguished
Every fluctuation in data always represents a genuine, meaningful trend
Distinguishing between these has no bearing on the accuracy of a conclusion
11. Why might looking at a full data set (not just the first and last points) be important when identifying a trend?
The pattern across all the data reveals whether the change is consistent, not just a coincidence between two points
Only the first and last data points ever matter for identifying a trend
Looking at the full data set never adds any useful information
Trends can always be reliably identified from just two data points
12. Why might immediately deleting an inconvenient anomaly without investigation be considered poor scientific practice?
It could hide a genuine source of error or an interesting finding, and undermines the honesty of the analysis
Deleting inconvenient data is always the most scientifically rigorous approach
Anomalies never provide any useful information worth investigating
Investigating anomalies always wastes valuable research time with no benefit
13. A data set shows plant height increasing with sunlight for 9 out of 10 plants, with one clear exception. What is the most appropriate next step?
Investigate the exception for a possible explanation, while still recognising the overall pattern in the other 9
Ignore the overall pattern in the 9 plants because of the one exception
Assume the entire experiment is invalid because of a single anomaly
Delete the anomalous data point with no explanation and move on
14. Why might a pattern only become visible once enough data points have been collected and analysed together?
A small number of data points might not reveal a clear, reliable pattern due to natural variation
Patterns are always immediately obvious from a single data point
The amount of data collected has no bearing on whether a pattern is visible
Patterns only ever emerge from anomalies, never from consistent data
15. Why might failing to investigate an anomaly risk drawing an inaccurate overall conclusion from an investigation?
An unexplained anomaly could indicate a flaw in the method or an important unaccounted-for factor affecting the whole result
Anomalies never have any bearing on the accuracy of a conclusion
Ignoring anomalies always leads to a more accurate final conclusion
Investigating anomalies is only relevant to the write-up, not the conclusion
16. Why might a scientist look for a plausible explanation for an anomaly before deciding whether to include or exclude it from their final analysis?
Excluding data without justification could bias the results, while a plausible explanation provides valid grounds for exclusion
Data can always be excluded from an analysis with no justification needed
Anomalies should always be included in every analysis with no consideration
The reason for excluding a data point has no bearing on the integrity of the analysis
17. Why might identifying a clear trend in data still not be enough, on its own, to confidently explain WHY that trend is occurring?
A trend shows a pattern exists, but explaining its underlying cause often requires further investigation or reasoning
Identifying a trend is always sufficient to fully explain its cause
Trends and their causes are always identical and require no further investigation
A visible trend guarantees the cause is already obvious and requires no further thought
18. Why might comparing your own data's patterns and trends against previous, similar published research be a valuable step in analysis?
It helps assess whether your findings are consistent with established understanding or reveal something new or unusual
Comparing with prior research has no value in analysing your own data
Your own data should always be analysed in complete isolation from other research
Consistency with prior research is never relevant to evaluating new findings
19. Why might a repeated anomaly across multiple independent investigations be treated differently from a single, one-off anomaly in just one study?
A repeated anomaly across independent studies suggests a genuine, reproducible phenomenon rather than a one-off error
Repeated and single anomalies should always be treated in exactly the same way
A repeated anomaly is always less significant than a single one-off anomaly
Anomalies occurring across multiple studies never carry any additional scientific weight
20. A weather station's temperature data shows a clear seasonal trend, but one single day shows an extreme, unexplained spike. Why should this anomaly be checked before including it in a broader climate analysis?
It could be a sensor malfunction rather than a genuine temperature event, which would distort the analysis if left unchecked
A single anomalous reading always represents a genuine, significant climate event
Anomalies should always be included in analysis with no verification
This anomaly definitely disproves the entire seasonal trend
21. Understanding how to analyse data for patterns, trends and anomalies mainly helps you to:
Draw well-supported conclusions from data, while appropriately investigating unexpected results
Assume every data point fits a pattern perfectly with no exceptions
Ignore unusual results without any further consideration
Avoid drawing any conclusions from collected data at all
Answer key (parent copy)
1. A consistent relationship between variables
2. An overall direction in data, like a steady rise or fall
3. A data point that doesn't fit the overall pattern
4. Investigate it rather than ignoring it
5. A measurement error or an unaccounted-for variable
6. A consistent increase in the measured variable
7. Looking for patterns, trends and anomalies
8. The anomaly could reveal an unaccounted-for factor or a pattern the original hypothesis didn't predict
9. Identifying an unaccounted-for variable affecting that specific data point
10. A real trend suggests an underlying cause, while noise could just be normal variation with no meaningful pattern
11. The pattern across all the data reveals whether the change is consistent, not just a coincidence between two points
12. It could hide a genuine source of error or an interesting finding, and undermines the honesty of the analysis
13. Investigate the exception for a possible explanation, while still recognising the overall pattern in the other 9
14. A small number of data points might not reveal a clear, reliable pattern due to natural variation
15. An unexplained anomaly could indicate a flaw in the method or an important unaccounted-for factor affecting the whole result
16. Excluding data without justification could bias the results, while a plausible explanation provides valid grounds for exclusion
17. A trend shows a pattern exists, but explaining its underlying cause often requires further investigation or reasoning
18. It helps assess whether your findings are consistent with established understanding or reveal something new or unusual
19. A repeated anomaly across independent studies suggests a genuine, reproducible phenomenon rather than a one-off error
20. It could be a sensor malfunction rather than a genuine temperature event, which would distort the analysis if left unchecked
21. Draw well-supported conclusions from data, while appropriately investigating unexpected results