Evaluating Inferential Reasoning from Evidence
Welcome to one of the most important chapters for mastering Paper 4! In this section, we are moving beyond just looking at arguments to look at the "raw materials" they are built from: evidence. Specifically, we are learning how to judge the "leap" an author makes from a piece of data to a conclusion. This is what we call inferential reasoning.
In your exam, especially in Question 3 of Paper 4, you will be asked to evaluate how well the evidence in the documents supports the claims being made. Don't worry if this sounds a bit technical—it’s essentially about being a smart "fact-checker."
Note: For related skills on how arguments are built or how to spot logical flaws, see our chapters on "Analyse the structure of the reasoning" and "Evaluate the reasoning in a document."
1. What is Inferential Reasoning?
An inference is a conclusion you draw based on evidence. Inferential reasoning is the process of connecting the dots.
Example: If you see a student's desk is covered in (a) empty coffee cups and (b) open textbooks, you might infer that they stayed up all night studying.
The evidence is the cups and books; the inference is the "all-nighter." Your job is to decide if that inference is strong or if there is a better explanation (maybe they just haven't cleaned their desk in a week!).
2. Assessing the Evidence (The "Inputs")
Before you can judge the inference, you must judge the quality of the evidence itself. The syllabus breaks this down into Credibility and Representativeness.
A. Credibility (The RAVEN/RAVON Method)
To decide if a source is reliable, we look at five main criteria:
- Reputation: Does the source (the person or organization) have a history of being truthful and accurate?
- Ability to See: Was the source actually there? Do they have direct access to the information?
- Vested Interest: Does the source have something to gain by lying or twisting the truth? (e.g., a company saying its own product is the best).
- Expertise: Does the source have specialized knowledge or professional qualifications in this specific area?
- Neutrality/Bias: Is the source side-lined by a particular prejudice, or are they balanced and objective?
B. Representativeness of a Sample
When evidence is based on a study or survey, you must check if the sample represents the whole group. Look for:
- Number: Is the sample size too small? (e.g., asking \(5\) people does not represent a city of \(1,000,000\)).
- Characteristics: Does the group reflect the diversity of the population? (e.g., surveying only teenagers about retirement plans is not representative).
- Selectivity: Was the sample "hand-picked" to get a certain result? (e.g., only surveying people who already shop at a specific store).
Key Takeaway: If the evidence is biased, unrepresentative, or from an unreliable source, any inference drawn from it will be weak.
3. Assessing the Inference (The "Leap")
Once you've looked at the evidence, you need to see if the conclusion drawn from it actually follows. Here is what to look for:
Correlation vs. Causation
This is a classic "Thinking Skills" trap. Just because two things happen at the same time (correlation), it doesn't mean one caused the other (causation).
Example: "Ice cream sales increase in June. Sunburns also increase in June. Therefore, eating ice cream causes sunburn."
The flaw: There is a third factor (the hot weather) causing both. This is an invalid inference.
The Significance of Evidence
Is the evidence actually significant to the claim?
If a document claims a new law is "massively popular" because \(51\%\) of people support it, the inference is weak because \(51\%\) is only a marginal majority. The "significance" is low.
Generalisations
Watch out for rash generalisations. This happens when an author takes a very specific piece of evidence and applies it to everything.
Example: "My dog bit me; therefore, all dogs are dangerous." The evidence (one dog) does not support the broad inference (all dogs).
4. Suggesting and Assessing Explanations
In Paper 4, you might be asked to provide an alternative explanation for the evidence. This is a great way to challenge an inference.
Scenario: A study shows that students who sit at the front of the class get higher grades.
Document Inference: Sitting at the front causes higher grades because you can hear the teacher better.
Your Alternative Explanation: Perhaps students who are already highly motivated choose to sit at the front. The motivation causes the grades, not the seat itself.
Quick Tip: If you can find a plausible alternative explanation, the original inference is challenged.
5. Working Across Documents
Since this is Paper 4, you are often looking at multiple documents. Use them together!
- Corroboration: If Document A and Document B say the same thing using different evidence, the inference becomes stronger.
- Inconsistency: If Document C provides data that contradicts the claim in Document D, the inference in Document D is weakened.
6. Common Mistakes to Avoid
- Confusing "Fact" with "Inference": A fact is what the data says (e.g., "\(20\%\) of people said Yes"). An inference is what the author thinks it means (e.g., "This proves the policy is failing"). Focus your evaluation on the second part!
- Being one-sided: If the question asks you to "evaluate," look for both strengths and weaknesses in the reasoning.
- Ignoring the "Source": Always look at who wrote the document. A scientist's inference about climate change is generally stronger than a blogger's inference due to expertise.
Summary Checklist for the Exam:
1. Is the source of this evidence credible? (RAVON)
2. Is the sample size and type representative?
3. Is there a "leap" from the data to the conclusion that is too big?
4. Is there an alternative explanation for this data?
5. Does other evidence in other documents support or contradict this?
Don't worry if this seems tricky at first! Like any skill, evaluating reasoning takes practice. Start by looking at news articles and asking: "Where did they get that number, and does it really prove what they say it proves?"