Evidence, Data & Models

Learning objectives

• Write claims supported by a dataset
• Identify what evidence does not justify
• Evaluate a scientific claim and a model

Materials

• Science notebook
• Provided seedling dataset
• One scientific claim from media or advertising

1. Patterns and anomalies

Data analysis identifies patterns without hiding variability. An anomalous result should be checked, repeated when possible, and documented rather than silently deleted.

2. Correlation and causation

When two variables change together, they are correlated. Establishing causation requires stronger evidence, a defensible mechanism, control of alternative explanations, and often repeated studies.

3. Models are selective

A model may explain one feature well while omitting others. Evaluate what is represented, what is simplified, the scale, assumptions, evidence, and the conditions under which the model may fail.

4. Essential question & claim-evidence reasoning

Essential question: How can a student make a useful scientific claim without saying more than the evidence actually supports?

A strong scientific claim is narrow enough to be tested against the available evidence. Evidence becomes more useful when the source, method, sample, variables, uncertainty, and alternative explanations are considered. Reasoning explains why the evidence supports the claim and where the evidence stops being sufficient.

This is the core of claim–evidence–reasoning: state what you think the evidence supports, identify the specific evidence, and explain the scientific connection without hiding limitations.

5. Worked data example

Use the seedling table from Section 1.

1. From 0 to 2 hours of daily light, average height increases from 2.0 cm to 2.6 cm: a change of 0.6 cm.
2. From 6 to 8 hours, average height increases from 4.1 cm to 4.3 cm: a change of only 0.2 cm.
3. A supported claim is: In this dataset, average seedling height generally increases as daily light increases from 0 to 8 hours.
4. An unsupported claim is: Eight hours of light is always the ideal amount for every plant. The table does not test every species, every light intensity, every temperature, or long-term outcomes.

6. Vocabulary & misconception checks

• Pattern: a repeatable or systematic relationship visible in observations or data.
• Anomaly: a result that differs noticeably from the broader pattern and deserves investigation.
• Correlation: two variables change together in a measurable way.
• Causation: evidence supports that changes in one factor contribute to changes in another.
• Assumption: a condition accepted for the purpose of a model or analysis.
• Limitation: a boundary on what the method, data, or model can justify.

7. Retrieval, mentor evidence conference & portfolio evidence

1. What should happen to an anomalous data point before anyone decides to exclude it?
2. What is the difference between correlation and causation?
3. What is one assumption and one limitation of a model you have used?
4. Why is an original source usually more useful than a social-media summary when evaluating a scientific claim?