DoAssignment study guide

A1.8 · Analyse data against predictions and evaluate error or bias

Learn to analyse data against predictions and evaluate error or bias through clear examples and targeted practice.

Ontario Grade 12 Biology

Scientific Investigation Skills and Career Exploration

Ontario Grade 12 Biology — study topic A1.8

In SBI3U, you learned to use evidence to investigate living systems. For example, an investigation might compare plant growth under different conditions. A prediction states what you expect to observe; data are the recorded observations or measurements. A prediction is not automatically confirmed just because one result agrees with it. In this lesson, you will compare results with predictions, consider how large and consistent the differences are, and judge whether error or bias could have affected the evidence. The examples use hypothetical classroom data. They are not reports of real experiments.

What you will learn

Compare observed biological data with a prediction.
Use differences and patterns in data to judge how well a prediction is supported.
Identify possible sources of error and bias.
Explain how uncertainty and limitations affect a conclusion.

1. From an SBI3U question to a testable prediction

A biological investigation begins with a question about a living system. A system is the part of nature being studied, such as a group of seedlings or a population in a habitat. The investigator identifies what will be changed or compared and what will be measured. A prediction states the expected result before the data are examined.

Suppose a class asks whether seedlings given more light will grow taller over a set period. A clear prediction identifies the comparison and expected direction: seedlings receiving more light will have a greater average increase in height than seedlings receiving less light. Average means the total of the measurements divided by the number of measurements.

The prediction helps guide the investigation, but it does not determine what the data must show. After measurements are collected, compare the observed result with the predicted direction and, where possible, the predicted size. A result that differs from the prediction is still evidence. It may mean the prediction was not supported under the conditions tested, or that the investigation had limitations.

Write down the prediction before interpreting the results.
State what is compared, what is measured, and the expected direction.
A result that does not match a prediction is not automatically a failed investigation.

2. Compare evidence with the prediction

Begin by checking that the data answer the question. Confirm that the recorded measurement is the one named in the prediction, and that the comparison groups or conditions are clear. Then describe what the data show without adding an explanation that has not been tested.

A simple comparison uses the difference between two group averages. If the prediction expects one group to have a larger average, check whether the measured averages have that order. The difference can show how far apart the averages are, but it does not by itself explain why they differ or prove that the tested condition caused the difference.

Look at the individual measurements as well as the averages. Averages can hide variation, which means that measurements within a group are not all identical. If the values overlap greatly or vary widely, the difference between averages may be less convincing than it first appears. If the pattern is similar across repeated trials, that consistency adds support, although it does not remove every possible source of error.

Keep the conclusion proportional to the evidence. Say that the results support, do not support, or are inconclusive for the prediction under the tested conditions. Inconclusive means the evidence does not clearly distinguish between the possibilities. Avoid claiming that a result proves a general rule for all organisms or conditions.

\text{difference} = \text{group 1 average} - \text{group 2 average}

Check direction, size, and consistency of the observed pattern.
An average summarizes a group but can hide individual variation.
Describe evidence first; limit the conclusion to the conditions tested.

3. Identify error, bias, and limits

Error is a difference between a recorded value and the value that would have been obtained through a perfectly accurate measurement. In biology, perfect measurement is rarely possible. Instruments may have limited precision, meaning they cannot distinguish arbitrarily small differences. A ruler marked in millimetres, for example, cannot reliably provide a measurement to a much smaller fraction of a millimetre.

Random error is unpredictable variation that can make some readings a little higher and others a little lower. Repeating measurements can help reveal this variation and make a group pattern easier to judge. Systematic error is a consistent measurement problem, such as a balance that always reads too high. Repetition alone may not correct a systematic error.

Bias is a tendency in the investigation or interpretation that favours a particular result. For example, if a person measuring plant height knows which plants received more light and unconsciously measures those plants differently, the procedure may introduce bias. A consistent measurement rule, and where practical a person who does not know the group assignment, can reduce this risk.

A limitation is a feature of the investigation that restricts what can be concluded. A small number of organisms may not represent a wider population well. Differences in water, temperature, or starting size may also make it difficult to attribute a result to the condition being tested. Name a specific limitation and explain how it could affect the comparison; do not use the word as a vague excuse to dismiss an unexpected result.

Evidence and uncertainty should be considered before claiming that a model describes a real organism. A model is a simplified representation used to organize an explanation or prediction. Data may support a model in the tested setting, but other conditions or organisms may produce different results.

Random error varies unpredictably; systematic error shifts measurements consistently.
Bias can influence how evidence is collected or interpreted.
Explain how a specific error or limitation could affect the conclusion.

4. A practical analysis sequence

Use a consistent sequence when analysing data. First, restate the prediction. Next, identify the relevant measurements and compare their pattern with the expected direction. Then consider how much variation is present and whether repeated trials show a similar pattern. Finally, identify plausible error or bias and state a conclusion that reflects both the evidence and its limits.

Do not assume that any difference is meaningful simply because two averages are not identical. Ask whether the size and consistency of the difference fit the prediction, and whether a measurement or design issue could have produced it. In this course-level analysis, the goal is a reasoned interpretation, not a claim of certainty.

When proposing an improvement, connect it to the problem identified. If readings may be inconsistent, use the same measurement method and repeat measurements. If groups differ in more than the tested condition, keep those other conditions as similar as possible. An improvement can reduce a concern, but it cannot guarantee that every source of error or bias is removed.

Prediction → compare pattern → assess variation → evaluate error or bias → conclude.
An improvement should address a named weakness.
Use cautious wording that matches the strength of the evidence.

Questions to ask when evaluating evidence

Focus	Question	Why it matters
Prediction	Does the observed pattern match the predicted direction?	Shows whether the prediction is supported by the pattern.
Variation	Are measurements similar, or do they vary widely?	Shows whether an average represents the group clearly.
Error	Could a measurement problem shift or scatter the values?	Helps judge confidence in the recorded data.
Bias	Could expectations or the method favour one result?	Warns that the process may influence the evidence.
Conclusion	What claim fits the evidence and its limits?	Prevents claims stronger than the investigation supports.

Worked example

A difference in seedling growth

Hypothetical classroom data compare the increase in height of seedlings over the same period. The low-light group has an average increase of 4.2 cm; the higher-light group has an average increase of 5.1 cm. The prediction was that the higher-light group would have a greater average increase. Analyse the result and identify one limitation.

Compare the predicted direction
The prediction expects the higher-light group to have the greater average increase. Its measured average is greater, so the direction of the result matches the prediction.
Calculate the difference
Subtract the low-light average from the higher-light average. The positive difference shows that the higher-light average exceeds the low-light average by 0.9 cm.
$5.1\,\text{cm} - 4.2\,\text{cm} = 0.9\,\text{cm}$
Limit the conclusion
The averages support the prediction in this hypothetical comparison. Without individual measurements or information about other conditions, we cannot judge how consistent the difference was or conclude that light alone caused it. Unequal water or starting seedling sizes could be limitations.

Answer: The observed average is 0.9 cm greater in the higher-light group, matching the predicted direction. The data support the prediction for this comparison, but do not by themselves establish the cause.

Check: A stronger analysis would inspect the individual measurements, repeat-trial pattern, and whether groups were treated similarly apart from light.

Worked example

A prediction not supported by the averages

Hypothetical data compare the average number of leaves on two groups of plants. The prediction was that plants receiving condition A would have more leaves than plants receiving condition B. The measured averages are 8.0 leaves for condition A and 8.6 leaves for condition B. What conclusion is justified, and what should be checked before interpreting the result?

Compare the result with the prediction
The prediction expected condition A to have the greater average. Instead, condition B has the greater average, so the observed direction does not match.
$8.0 < 8.6$
State a cautious conclusion
These averages do not support the prediction under the tested conditions. They do not show why the pattern occurred. Before suggesting an explanation, check the individual measurements, the measurement method, and whether other conditions were kept similar.
Evaluate a possible concern
If one group began with larger plants, the difference in leaf averages might not be due to condition A or B. This is a possible limitation to investigate, not a fact established by the averages.

Answer: The averages run opposite to the predicted direction, so the prediction is not supported by these results. More information is needed to assess variation and possible differences between groups.

Check: Do not change the prediction after seeing the data or claim that the result proves the opposite explanation.

Worked example

Spotting a measurement bias

In a hypothetical investigation, students measure the length of leaves in two groups. The students know which condition each plant received. They expect condition A to produce longer leaves. Give one possible source of bias, explain its effect, and suggest a relevant improvement.

Identify how bias could enter
Knowing the condition may influence how a student chooses the leaf to measure or reads the ruler. If this happens more often for condition A, its recorded lengths could be shifted upward.
Connect the concern to the conclusion
A higher average for condition A would be harder to interpret because the measurement process itself might favour the expected result. This possibility does not prove that bias occurred; it identifies a risk in the method.
Choose a matching improvement
Use a written rule for which leaf to measure and how to position the ruler. If practical, have the measurer record samples without knowing their condition. These steps reduce opportunities for the expectation to affect measurement.

Answer: Knowledge of the group could influence leaf selection or ruler reading. A fixed measurement rule and, where practical, a measurer unaware of group assignment would reduce this risk.

Check: The improvement addresses possible measurement bias; it does not prove that the original measurements were biased.

Common mistakes and how to avoid them

Treating a matching result as proof that the prediction is always true.

Correction: Say that the data support the prediction under the tested conditions. Consider variation and limits before making a broader claim.

Calling any unexpected result an error.

Correction: An unexpected result may be a genuine observation. Identify a specific measurement or design concern before suggesting error.

Assuming repeated measurements remove every problem.

Correction: Repeats can reveal variation, but a consistent systematic error or bias may remain.

Naming a limitation without explaining its effect.

Correction: State how the limitation could change the measurements, comparison, or strength of the conclusion.

Lesson summary

A prediction states an expected result before data are interpreted.
Compare observed direction, size, and consistency with the prediction.
Random error varies unpredictably; systematic error shifts readings consistently; bias may favour a result.
A conclusion should match the evidence and remain limited to the conditions tested.
Suggest improvements that directly address identified error or bias.

Check your understanding

Question 1

A prediction says group X will have a higher average than group Y. The measured averages are 6.4 and 5.9, respectively. Which statement is best?

The result matches the predicted direction, so the prediction is proven for all groups.
The result supports the predicted direction in this comparison, but variation and limitations still matter.
The result must be biased because the averages differ.
The prediction should be rewritten to match the averages.

Show answer and explanation

The result supports the predicted direction in this comparison, but variation and limitations still matter.

The average for X is higher, matching the predicted direction. The comparison alone does not prove a general rule or establish whether error or bias affected the measurements.

Question 2

A measuring device gives readings that are consistently too high. Which description best fits this concern?

Random error, because every measurement is different.
Systematic error, because readings are shifted in a consistent direction.
Bias is impossible when a device is used.
No error, because repeated readings can be averaged.

Show answer and explanation

Systematic error, because readings are shifted in a consistent direction.

A consistent shift upward is systematic error. Repeating and averaging readings may not remove a consistent shift.

Question 3

Students expect condition A to increase growth and know which plants received A. What is a suitable way to reduce possible measurement bias?

Ignore measurements that do not match the prediction.
Change the prediction after measuring the plants.
Use a consistent measurement rule and, where practical, keep the measurer unaware of group assignment.
Measure only the tallest plant in condition A.

Show answer and explanation

Use a consistent measurement rule and, where practical, keep the measurer unaware of group assignment.

A consistent rule and an unaware measurer reduce opportunities for expectations to affect measurement. The other choices selectively favour the expected result.

Key terms

Prediction: A statement of the result expected before data are interpreted.
Data: Recorded observations or measurements from an investigation.
Average: A summary calculated by adding measurements and dividing by their number.
Variation: Differences among measurements or observations in a group.
Error: A difference between a recorded measurement and the value that a perfectly accurate measurement would give.
Bias: A tendency in a method or interpretation that favours a particular result.
Limitation: A feature of an investigation that restricts what can be concluded.
Model: A simplified representation used to organize an explanation or prediction.

Continue through SBI4U

View the complete SBI4U Ontario Grade 12 Biology curriculum and lessons

About this lesson and its review

Published by DoAssignment. This reviewed lesson follows Ontario Grade 12 Biology (SBI4U), expectation A1.8. It is a study resource, not an official curriculum publication.

Before publication, content is checked for structure, mathematical or chemical notation, calculations, course boundaries, and readability. Errors can still occur, so corrections are welcomed.

Official curriculum reference

Report a correction or ask a question