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 11 Biology

Scientific Investigation Skills and Career Exploration

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

In science, a prediction is an expected result based on a question or an idea. Data are recorded observations or measurements. The result may agree with a prediction, partly agree, or disagree. In SNC2D, you learned to ask questions, collect evidence, and consider whether a test is fair. In this lesson, you will use those skills to analyse biological data. The goal is not to make the results fit an idea. It is to compare the evidence with the prediction and judge how trustworthy the comparison is.

What you will learn

State a clear prediction before interpreting results.
Describe what biological data show and compare the pattern with a prediction.
Use averages or ranges when they make a comparison clearer.
Identify possible error or bias and explain how it could affect a conclusion.

From a question to a fair comparison

Start with a biological question that can be investigated. For example: Does the amount of light affect the number of new leaves produced by young plants over a set period? A prediction gives a specific expected pattern. One prediction might be that plants receiving more light will produce more new leaves.

Before interpreting results, identify what is being changed and what is being measured. The independent variable is the factor deliberately changed. The dependent variable is the outcome measured. A controlled factor is something kept the same so it does not provide another explanation for a difference. Here, light amount is changed, leaf number is measured, and plant type and observation period could be kept the same.

A prediction should be clear enough to compare with results. Decide what pattern would count as agreement. If the prediction says that more light will lead to more leaves, compare leaf counts from groups receiving different amounts of light. One observation may not show a dependable pattern. Reliability means how consistently a method gives similar results when repeated under similar conditions.

Write the prediction before interpreting the results.
Name the factor changed, the outcome measured, and important factors kept the same.
A prediction describes an expected result; it does not guarantee that result.

Describe the evidence before explaining it

A data set is a group of recorded observations or measurements. First describe its pattern in plain language. For example, say that the group receiving more light had a higher average leaf count. Then compare that pattern with the prediction. This keeps the observation separate from an explanation.

An observed pattern is what the recorded data show. An explanatory model is an idea about why the pattern occurred. If plants receiving more light had more leaves, that is an observed pattern. Saying that light caused the difference is a stronger claim. It is justified only if the investigation was designed well enough to address other reasonable explanations.

Use numbers when they help make the comparison clear. An average is the sum of recorded values divided by the number of values. A range is the difference between the largest and smallest values. A range can show that results vary, but it does not explain why they vary. State units whenever a measurement has units.

Compare like with like: use the same measurement, units, time period, and groups that the prediction refers to. A prediction can expect a number or amount, or it can expect a pattern, such as one group having more than another.

The average can be represented by adding the recorded values and dividing by the number of values. In the expression below,

x_i

represents a recorded value and

n

is the number of values.

\bar{x}=\frac{\sum x_i}{n}

Report what was observed before offering a reason for it.
Use averages or ranges when they clarify the comparison.
Agreement with a prediction does not by itself prove an explanation.

Check error, variation, and bias

Error is a difference between a recorded result and the value that a more accurate measurement or procedure would produce. Random error varies unpredictably, such as small differences in judging where a leaf begins. Repeating observations can make variation easier to notice, but it does not remove every error.

A consistent measurement problem can shift results in one direction. For example, if a measuring tool always gives readings that are slightly high, results made with it may also be too high. Identify the possible problem and explain how it could affect the comparison.

Bias is a systematic influence that makes a result or interpretation lean in a particular direction. Selection bias occurs when the way observations are chosen makes the sample unlike the group the investigator wants to understand. Observer bias can occur when a person's expectations influence how they record or interpret observations.

Ask whether the same method was used for each group, whether enough observations were made to notice variation, and whether the measurement rule was clear. Consider whether a group was treated differently or chosen in a way that favours the expected result. These questions help you judge confidence in the conclusion.

When results disagree with a prediction, do not discard them. Check the procedure and possible error or bias, then state what the data support. When results agree, do not claim certainty. A small or biased data set can match a prediction without showing that it is generally true.

Error concerns measurement or procedure; bias is a systematic influence on selection, recording, or interpretation.
Link each possible problem to its effect on the direction or dependability of results.
A sound conclusion is limited to what the data and method can support.

A quick evidence review

Question	What to check
What was predicted?	State the expected pattern before interpreting results.
What was observed?	Describe the data, using suitable numbers and units.
Do they agree?	Compare the observed pattern with the stated prediction.
How trustworthy is the comparison?	Look for error, bias, unequal methods, and limited observations.
What can be concluded?	Make a claim no broader than the evidence supports.

Worked example

Compare a pattern with a prediction

Practice data: Two groups of similar seedlings are observed for the same period. The prediction is that the group receiving more light will produce more new leaves. The lower-light group has leaf counts of 3, 4, and 5. The higher-light group has counts of 5, 6, and 7. Analyse the results and give a careful conclusion.

State the observed pattern
The higher-light group has larger counts in this small data set. Its counts range from five to seven, while the lower-light counts range from three to five. This describes the observations without claiming why they differ.
Compare group averages
Averages give a simple comparison of the groups. Add the lower-light counts and divide by three to get four leaves. Do the same for the higher-light group to get six leaves.
$\bar{x}_{\mathrm{low}}=\frac{3+4+5}{3}=4,\quad\bar{x}_{\mathrm{high}}=\frac{5+6+7}{3}=6$
Judge the prediction
The observed pattern agrees with the prediction because the higher-light group has the larger average. The data support the predicted pattern in this example. A stronger claim that light alone caused the difference would require confidence that other conditions, such as plant type and observation period, were comparable.

Answer: The higher-light group averaged six new leaves, compared with four in the lower-light group. This agrees with the prediction. The small data set supports the predicted pattern, but it does not establish that light was the only cause.

Check: A correct analysis distinguishes agreement with a prediction from proof of a cause.

Worked example

Notice when evidence does not match

Practice data: A prediction states that one classroom surface will have fewer visible mould spots after the same cleaning method than another surface. After the same observation period, the first surface has an average of 8 visible spots and the second has an average of 5. Analyse the mismatch. Do not assume a cause that was not measured.

Compare result and expectation
The prediction expected fewer spots on the first surface. Instead, its average is higher. The recorded pattern does not agree with the prediction.
$8>5$
Separate result from explanation
The counts show a difference, but they do not identify its cause. Differences in initial conditions, cleaning, or counting could be considered only if information about the method supports those possibilities.
State a limited conclusion
The results do not support the prediction under the conditions recorded. Before making a broader claim, check whether procedures and counting rules were the same and whether observations were repeated. Do not change or ignore data simply because they conflict with a prediction.

Answer: The observed average for the first surface is higher, not lower, so the results disagree with the prediction. The data alone do not show why. Check the method and possible sources of error before making a broader claim.

Check: The conclusion reports the mismatch and avoids inventing an explanation.

Worked example

Evaluate possible selection bias

Practice scenario: A student predicts that one kind of leaf is longer than another kind. They choose only the largest leaves of the first kind but choose leaves at random from the second kind. The measured first group has a larger average. Is the prediction supported fairly?

Identify the observed result
The measured average is larger for the first kind of leaf. This is the pattern in the recorded data.
Identify the unfair selection
The groups were selected differently. Choosing only the largest leaves for one kind is selection bias because it favours larger measurements in that group. Randomly chosen leaves from the other kind do not make a fair comparison with only the largest leaves.
Revise the conclusion
The result does not fairly establish that the first kind generally has longer leaves. The selection method could explain the larger average. A fairer comparison would use the same selection rule for both kinds, then compare the data.

Answer: Although the first group has a larger measured average, selection bias makes the comparison unfair. The data do not support a reliable general conclusion about which kind usually has longer leaves.

Check: A possible bias matters because it can push the observed pattern in a particular direction.

Common mistakes and how to avoid them

Saying a prediction was proven because the results agreed with it.

Correction: Say that the results support or agree with the prediction under the conditions tested. Consider the method and its limits.

Treating a result that disagrees with a prediction as a failed investigation.

Correction: A mismatch is useful evidence. Report it accurately, then examine possible error or bias.

Naming a possible source of error without explaining its effect.

Correction: Explain whether it could raise, lower, or make the results less dependable.

Claiming that one factor caused a difference when other conditions may also differ.

Correction: Describe the observed pattern and limit cause claims to what the investigation can support.

Lesson summary

A prediction states an expected pattern before results are interpreted.
Describe observations first, then compare them with the prediction.
Use numbers, averages, or ranges when they clarify the evidence.
Error and bias can affect how well data represent the result being investigated.
Explain how a possible limitation affects confidence, and keep conclusions within the evidence.

Check your understanding

Question 1

A prediction expects group A to have fewer spots than group B. The recorded averages are 4 for A and 7 for B. Which statement is best?

The data agree with the prediction, but do not prove its explanation.
The prediction is proven correct for every case.
The data disagree because 4 is smaller than 7.
The averages show that no error or bias occurred.

Show answer and explanation

The data agree with the prediction, but do not prove its explanation.

Group A has the lower average, so the pattern agrees with the prediction. Agreement does not prove a cause or rule out error and bias.

Question 2

An observer expects one group to have longer leaves and repeatedly measures its leaves from the tip to the wrong point. What concern is most relevant?

Possible measurement error that could affect the recorded lengths.
Proof that the prediction is correct.
Selection bias caused by random selection alone.
Evidence that the two groups are identical.

Show answer and explanation

Possible measurement error that could affect the recorded lengths.

Using the wrong measurement points can make recorded lengths inaccurate. The effect on the comparison depends on whether the same mistake was made for both groups.

Question 3

Why is choosing only the largest leaves from one group a problem when comparing average leaf length?

It can create selection bias and make that group appear longer.
It guarantees that the prediction is false.
It removes all variation from the data.
It proves that leaf type caused the difference.

Show answer and explanation

It can create selection bias and make that group appear longer.

The selection method favours large leaves in one group, so the groups are not selected fairly. The comparison may be biased.

Key terms

Prediction: A clear statement of an expected result or pattern.
Data: Recorded observations or measurements.
Independent variable: The factor deliberately changed in an investigation.
Dependent variable: The outcome measured in an investigation.
Controlled factor: A condition kept the same to make a comparison fair.
Error: A difference between a recorded result and a more accurate result that could be obtained.
Bias: A systematic influence that favours a particular result or interpretation.
Selection bias: A problem caused when the way observations or subjects are chosen favours one outcome.

Continue through SBI3U

View the complete SBI3U Ontario Grade 11 Biology curriculum and lessons

About this lesson and its review

Published by DoAssignment. This reviewed lesson follows Ontario Grade 11 Biology (SBI3U), 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