COMP 3270 / Module 4 / Unit 11 / Lesson 2

Lesson 11.2 · Weeks 7–8 · Reading: Chapter 8

Claim-Evidence Alignment, Corroboration, and Triangulation

Align evidence with atomic claims. Classify alignment as direct support, partial support, qualified support, contradiction, or non-addressing. Assess source independence, distinguish corroboration from repetition, and communicate conflicting evidence responsibly.

Estimated study time: 4–5 hours Prerequisite: Lesson 11.1 Module 4 · Weeks 7–8

Table of Contents

1. Lesson Overview

In Lesson 11.1, you learned to distinguish information from evidence, sources from citations, and to evaluate authority, relevance, currency, and provenance. In this lesson, you will go further: you will learn to align evidence with specific claims, assess whether the evidence directly supports, partially supports, qualifies, contradicts, or does not address the claim. You will also learn to evaluate source independence and distinguish corroboration from mere repetition.

You will explore triangulation—comparing different sources, methods, or datasets—and learn to preserve disagreement and uncertainty rather than forcing a false consensus. These skills are essential for responsible verification, especially when claims are complex or consequential.

This lesson prepares you for Lesson 11.3 (lateral reading and citation tracing) and for the cumulative Unit 11 project and Assignment 5.

⚖️ Evidence and verification reminder: In this lesson, you will learn to evaluate whether a source actually supports a claim, and whether multiple sources provide independent corroboration or merely repeat the same information. You will also learn to preserve disagreement and communicate uncertainty proportionately.

2. Learning Outcomes

By the end of this lesson, you should be able to:

  1. Align evidence with atomic claims and classify evidence as directly supporting, partially supporting, supporting with qualification, contradicting, or not addressing the claim.
  2. Identify when the wording of a claim exceeds the evidence (overstatement).
  3. Assess source independence and determine whether multiple sources are independent or share a common origin.
  4. Distinguish corroboration (independent support) from repetition (echoing the same source).
  5. Develop a triangulation plan using different sources, methods, or perspectives.
  6. Communicate conflicting evidence responsibly, preserving disagreement and uncertainty.
  7. Identify evidence gaps and determine when one source is insufficient for a consequential claim.
  8. Apply the author-approved CLAIM framework (alignment step) to evaluate claim-evidence relationships.
  9. Integrate TRACE error analysis with CLAIM verification to prioritise claims and correct wording.

3. Before You Begin

4. Unit 10 Retrieval Practice

The following questions review TRACE and claim analysis from Unit 10. Answer each before revealing the feedback.

Question 1: Decompose the following compound claim into at least two atomic claims: “The new AI tutor has been proven to increase student test scores and reduce dropout rates, according to a 2023 study.”

Show feedback

Atomic claims:

  • Claim 1: “The new AI tutor increases student test scores” (based on a 2023 study).
  • Claim 2: “The new AI tutor reduces dropout rates” (based on the same study).
  • Claim 3: “The study ‘proved’ these effects” (the word “proven” is a strong claim that may not be supported).

Each atomic claim must be verified independently.

Question 2: Using TRACE, what type of error is most likely if a claim states “the study proved a causal relationship” but the original study was a correlational survey?

Show answer

Answer: Invalid inference — the claim infers causation from correlation. Also possible: unsupported claim (if the claim goes beyond what the evidence can support).

Question 3: Why is prioritising claims for verification important? Give one criterion for prioritisation.

Show answer

Answer: Prioritisation ensures that the most consequential claims (those with high impact on decisions, safety, or well-being) are verified first. A criterion could be the potential harm if the claim is false, or the centrality of the claim to the overall argument.

5. Opening Scenario: An AI-Generated Claim with Misalignment

This scenario runs throughout the lesson. Read it carefully and return to it as you learn each new concept.

📌 Scenario — A Public Health Communicator’s Query

You are a public health communicator preparing a briefing on the effectiveness of a new digital mental health app. You ask an AI assistant for a summary of the evidence. The AI produces this response:

“A 2022 randomized controlled trial (RCT) conducted by researchers at the University of Toronto found that the app, MindWell, significantly reduced symptoms of depression and anxiety in young adults. The trial, published in the Journal of Digital Health, reported a 25% reduction in depression scores and a 30% reduction in anxiety scores compared to a control group. These findings conclusively prove that MindWell is an effective treatment for clinical depression and anxiety disorders. Moreover, a 2023 review by the Canadian Mental Health Association (CMHA) supports the app’s efficacy and recommends its use as a first-line intervention.”

The AI cites:

  • A 2022 RCT from the University of Toronto.
  • A 2023 review from the Canadian Mental Health Association.

Your task: Before you dive into verification, consider:

  1. Does the RCT evidence support the claim that the app is effective for clinical depression and anxiety? What qualifications might be needed?
  2. Is the CMHA review independent of the RCT? Does it provide corroboration or does it simply cite the same trial?
  3. What wording in the AI response overstates the evidence? Identify specific phrases.
Show guided analysis of the scenario

1. Alignment of RCT evidence: The RCT shows reductions in scores for young adults, but the claim says “clinical depression and anxiety disorders.” The study likely measured symptom scales, not clinical diagnoses. Also, it was conducted in young adults; the claim generalises to all ages. The word “conclusively prove” is an overstatement—one RCT does not prove a treatment effective for a disorder; multiple trials and consensus are needed.

2. Independence of CMHA review: The CMHA review may be independent, but it may have relied heavily on the same RCT. To determine independence, we must check whether the review cites other studies or simply repeats the RCT’s findings. If the review is based on the same RCT, it does not provide corroboration—it’s repetition.

3. Overstated wording: “Conclusively prove,” “effective treatment for clinical depression and anxiety disorders,” “recommends use as a first-line intervention” — these are strong claims that exceed what a single RCT can support. A responsible statement would qualify: “A 2022 RCT found reductions in symptom scores in a sample of young adults; further research is needed to establish effectiveness for clinical populations.”

6. Core Content

6.1 Claim-Evidence Alignment

Definition — Claim-Evidence Alignment

Claim-evidence alignment is the relationship between a specific claim and the evidence used to support it. Evidence may:

  • Directly support — the evidence confirms the claim as stated.
  • Partially support — the evidence supports part of the claim but not all.
  • Support with qualification — the evidence supports the claim only under certain conditions, which must be stated.
  • Contradict — the evidence directly opposes the claim.
  • Not address — the evidence is irrelevant to the claim.

To evaluate alignment, you must first decompose the claim into atomic propositions. Then, for each atomic claim, determine whether the evidence speaks to it, and if so, how.

Example: Claim: “The AI tutor improves student performance and satisfaction.” Evidence: A study shows a 10% increase in test scores but no significant change in satisfaction. Alignment: the evidence partially supports (supports performance, not satisfaction). If the claim said “improves both,” that part is unsupported.

Common mistake: Treating a source that supports a similar but different claim as if it supports the exact claim. For example, a study on adult learners may not support a claim about K-12 students.

6.2 Wording Strength and Overstatement

The strength of your wording must match the evidence. Words like “proves,” “causes,” and “demonstrates” are strong causal claims, while “suggests,” “may,” “could,” and “is associated with” are more tentative. If the evidence is correlational, causal wording is overstatement.

Example: Evidence: A correlational study shows that people who use the app have lower anxiety. Wording: “The app reduces anxiety” — overstatement (correlation ≠ causation). Corrected wording: “App use is associated with lower anxiety in this sample.”

Qualification: When evidence supports a claim but with limitations, you must state those limitations: “In a sample of university students, the app was associated with a 20% reduction in anxiety scores; however, the study was short-term and used self-report measures.”

6.3 Source Independence

Definition — Source Independence

Two sources are not independent if they share the same origin, rely on the same data, copy each other, or belong to the same reporting chain. Independence means the sources were produced separately, without relying on each other for the key evidence.

Independence is crucial for corroboration. If multiple sources all trace back to the same original study, they do not provide independent verification—they are repetitions.

Common dependencies:

  • One source cites the other.
  • Both cite the same primary study.
  • Both rely on the same press release.
  • Both use the same dataset generated by a single organisation.
  • One is a summary of the other (e.g., a news article about a research paper).
  • AI-generated content that repeats the same pattern without new evidence.

To assess independence, examine the reference lists, methods sections, and provenance of each source. If they share a common primary source, they are not independent.

6.4 Corroboration vs. Repetition

Definition — Corroboration vs. Repetition

Corroboration is the process of obtaining sufficiently independent evidence that supports the same material claim. Repetition is when the same evidence is reported multiple times without adding new support.

Finding the same statistic in five different websites does not corroborate it if they all cite the same original study. The strength of corroboration comes from independent investigations, different methods, or different datasets that converge on the same conclusion.

Example: Claim: “Coffee reduces the risk of type 2 diabetes.” Corroboration: Three separate cohort studies, from different countries, with different methods, all find a similar association. Repetition: Ten news articles all reporting the same single study.

Corroboration increases confidence, but it does not eliminate uncertainty. Disagreement among independent sources is also valuable and must be preserved.

6.5 Triangulation

Definition — Triangulation

Triangulation involves comparing different sources, methods, perspectives, datasets, or forms of evidence to gain a more complete understanding. It can help identify strengths and weaknesses of each source and reduce bias.

Triangulation does not automatically resolve conflicting evidence; it may reveal that evidence is robust, or it may highlight unresolved contradictions. The goal is to use multiple lenses to understand a claim more fully, not to force agreement.

Example: To verify a claim about the effectiveness of a teaching intervention, you might compare:

  • Quantitative test scores (RCT).
  • Qualitative interviews with students.
  • Teacher observations.
  • Student retention data.

If all point in the same direction, confidence increases. If they conflict, you must document the disagreement and seek explanations.

Common mistake: Assuming that triangulation eliminates uncertainty. Even with multiple sources, there may be systematic biases or gaps. Always acknowledge remaining limitations.

6.6 Conflicting Evidence

Definition — Conflicting Evidence

Conflicting evidence occurs when credible sources disagree on a material claim. Responsible verification requires preserving disagreement, identifying differences in method or scope, and avoiding false certainty.

When sources conflict:

  • Document the disagreement explicitly.
  • Analyse why they differ (e.g., different populations, methods, time periods, definitions).
  • Avoid cherry-picking evidence to support a preconceived conclusion.
  • If the conflict is material and unresolved, escalate to a subject-matter expert.
  • Communicate the uncertainty to your audience.

Example: Two systematic reviews on the same topic reach different conclusions. One finds a positive effect; the other finds no effect. Instead of choosing the one you like, document both, explain the differences (e.g., one included older studies, the other used stricter inclusion criteria), and state that the evidence is mixed.

6.7 Evidence Gaps

Definition — Evidence Gap

An evidence gap exists when the available evidence is insufficient to support or refute a claim. This can happen when no source exists, sources are irrelevant, outdated, inaccessible, or when the claim exceeds the scope of the evidence.

Evidence gaps are not failures; they are honest assessments. You must document gaps rather than conceal them. When a claim is consequential and evidence is insufficient, you should escalate, withhold judgment, or qualify the claim.

Common gaps:

  • No source addresses the claim.
  • Available sources are too old or use different populations.
  • The original source is paywalled or inaccessible.
  • Only one dependent source exists.
  • Metadata are incomplete, making verification impossible.
  • The claim is speculative and goes beyond the evidence.

When you identify a gap, state: “The claim is unverifiable with the evidence currently available; further research is required.”

7. CLAIM Walkthrough: Alignment and Independence

Note: The author-approved CLAIM framework is introduced fully in Lesson 11.4. In this lesson, we focus on the operational elements concerning claim-evidence alignment and independence of sources. If the exact author-approved expansion has been supplied to your instructor, replace this box with the approved terminology.
Operational CLAIM element: Assess claim-evidence alignment
  • Definition: For each atomic claim, determine how the evidence relates to it: directly supports, partially supports, supports with qualification, contradicts, or does not address.
  • Guiding questions:
    • Does the source actually say what the claim states?
    • Are there important qualifications or limitations in the source that are omitted?
    • Does the claim generalise beyond the source’s scope?
  • Example: Claim: “The intervention reduced anxiety.” Source: “The intervention was associated with a reduction in anxiety scores in a sample of 50 college students.” Alignment: Partial support (sample size and population are limited).
  • Common mistake: Assuming that because a source is authoritative, it automatically supports the claim. Always check the actual content.
Operational CLAIM element: Assess source independence
  • Definition: Determine whether sources are independent or whether they share a common origin, data, or reporting chain.
  • Guiding questions:
    • Do these sources cite each other or the same primary source?
    • Were they produced by separate investigators or organisations?
    • Do they rely on the same dataset or measurement?
  • Example: Two news articles both cite the same press release. They are not independent for the claim about the study findings.
  • Common mistake: Counting multiple webpages as corroboration when they all repeat the same original source.

8. Verification-Process Walkthrough

This walkthrough demonstrates how to move from an atomic claim to evaluating alignment, independence, corroboration, and triangulation. We use the opening scenario claim: “MindWell significantly reduced symptoms of depression and anxiety in young adults.”

  1. Stage 1: Decompose and align the claim. Atomic claims: Claim A “MindWell reduced depression symptoms (in young adults)”; Claim B “MindWell reduced anxiety symptoms (in young adults)”. Evidence: The RCT reported a 25% reduction in depression scores and a 30% reduction in anxiety scores. The study population was young adults (ages 18–25). Alignment: Claim A and B are directly supported for the study population, but the claim uses the word “significantly” — we need to check if the reduction was statistically significant.
  2. Stage 2: Identify overstatement. The AI response goes beyond: “conclusively prove” and “effective treatment for clinical depression and anxiety disorders.” The RCT did not claim to treat clinical disorders; it measured symptom scores. The population was young adults, not all adults. The wording should be qualified: “In a sample of young adults, the app was associated with reductions in depression and anxiety scores; however, this was a single RCT and does not establish clinical efficacy.”
  3. Stage 3: Assess independence and corroboration. The CMHA review is cited as support. To determine independence, we must check whether the CMHA review included only the RCT or also other studies. If it is a narrative review that summarises the RCT, it is not independent. If it is a systematic review that included the RCT along with other studies, it may provide corroboration if its conclusions are consistent.
  4. Stage 4: Triangulation and conflicting evidence. Are there other studies with conflicting findings? If the CMHA review ignored contradictory evidence, that’s a problem. We should search for systematic reviews or meta-analyses that might have contradictory results. If conflicting evidence exists, we must preserve that disagreement.
  5. Stage 5: Document evidence gaps and final assessment. Gaps: no evidence on long-term effects; no evidence on clinical populations; the study was short-term and self-reported. Final assessment: The claim “MindWell significantly reduced symptoms” is supported (for the study population, with qualifications). The claim “proves effective treatment for clinical disorders” is unsupported and should be corrected. The CMHA review may not provide independent corroboration if it relies on the same study.

9. Worked Example: Evaluating a Claim About Remote Work

Claim: “A 2023 study by the University of California found that remote work increases productivity by 15% and reduces employee turnover.”

Step 1: Decompose the claim

Step 2: Locate and examine the evidence

The actual study (found by searching) is a survey of 500 tech workers in California, self-reported productivity, and turnover intentions (not actual turnover). It found a 15% increase in self-reported productivity and a 10% decrease in turnover intentions.

Step 3: Assess independence and corroboration

We find two other studies: one from a European university (2022) that found a similar productivity increase, and a meta-analysis (2024) that found mixed results. The European study is independent (different sample, country, method). The meta-analysis includes conflicting evidence.

Step 4: Conflicting evidence and evidence gaps

Step 5: Corrected wording and final assessment

Final disposition: Claim A is partially supported with qualifications; Claim B is unsupported (misstated). The overall claim is overgeneralised.

10. Guided Learning Activities

🧩 Activity 11.2.A: Claim-Evidence Alignment Table
  • Purpose: Practice classifying evidence as supporting, partially supporting, supporting with qualification, contradicting, or not addressing.
  • Learning outcome: Align evidence with atomic claims.
  • Estimated time: 20 minutes.

For the following claim and evidence, complete the alignment table.

Claim: “Online learning is less effective than face-to-face instruction for student engagement.”

Evidence A: A meta-analysis of 50 studies found that on average, online students had slightly lower engagement scores, but the difference was small and varied by discipline.

Evidence B: A survey of students in a hybrid course found that 70% reported being more engaged in the online portion than in the face-to-face portion.

Evidence C: A blog post by an educator argues that online learning is inherently disengaging.

Classify each evidence item.

Show activity feedback
  • Evidence A: Partially supports with qualification — it supports the claim that engagement is lower, but the difference is small and varies by discipline.
  • Evidence B: Contradicts — it shows that in a hybrid course, students felt more engaged online.
  • Evidence C: Does not address (or unsupported opinion) — it is an argument without data, and the author’s authority is not established.

Overall assessment: The claim is not supported by the evidence; evidence is mixed.

🧩 Activity 11.2.B: Wording Strength Audit
  • Purpose: Identify overstatement and select proportionate wording.
  • Learning outcome: Identify overstatement.
  • Estimated time: 15 minutes.

For the following evidence, select the most appropriate wording from the options below:

Evidence: A correlational study found that people who exercise regularly have lower blood pressure.

Options:

  1. “Exercise causes lower blood pressure.”
  2. “Exercise is associated with lower blood pressure.”
  3. “Exercise lowers blood pressure.”
  4. “Exercise may be associated with lower blood pressure.”
Show activity feedback

Answer: The most appropriate wording is d — “Exercise may be associated with lower blood pressure” (or b, if the association is well-established). Option a and c imply causation, which is not supported by a correlational study. Option b is acceptable if the study is robust, but “may” adds appropriate caution.

Explanation: The evidence does not establish causation; wording should reflect that.

🧩 Activity 11.2.C: Independent or Repeated?
  • Purpose: Assess source independence.
  • Learning outcome: Assess source independence.
  • Estimated time: 15 minutes.

You have four sources that all report the same finding: “A new drug reduces cholesterol by 20%.”

  • Source 1: The original clinical trial published in a medical journal.
  • Source 2: A news article from a major newspaper that summarises the trial.
  • Source 3: A blog post that cites the news article.
  • Source 4: A press release from the drug company that sponsored the trial.

Which sources are independent of each other? Which are repeated? Explain.

Show activity feedback

Independent: Only Source 1 (the original trial) is independent of the others. Source 4 (press release) is from the sponsor; it may be independent of the trial report but has a conflict of interest. Sources 2 and 3 are dependent — they derive from Source 1 and/or Source 4.

Repeated: Sources 2 and 3 repeat the original trial findings without adding new evidence. Source 4 may also repeat the trial findings but is a primary source for the company’s own statement.

Conclusion: To corroborate the finding, you would need independent trials, not just media repeats.

🧩 Activity 11.2.D: Corroboration and Triangulation Plan
  • Purpose: Develop a triangulation plan for a complex claim.
  • Learning outcome: Develop a triangulation plan.
  • Estimated time: 20 minutes.

Claim: “The new policy of four-day work weeks increases employee well-being and productivity.”

Design a triangulation plan: identify at least three different types of evidence or sources you would use to verify this claim, and explain what each could contribute. Also, identify potential conflicts you might encounter.

Show activity feedback

Possible triangulation plan:

  • Quantitative study: A controlled trial or longitudinal survey measuring well-being (e.g., job satisfaction, burnout) and productivity (e.g., output metrics).
  • Qualitative interviews: Employees’ subjective experiences, perceived stress, work-life balance.
  • Organisational data: Absenteeism, turnover, customer satisfaction, or financial performance before and after the policy.
  • Industry comparisons: Studies from other companies or countries that have implemented similar policies.

Potential conflicts: Quantitative data may show no change while qualitative data show positive perceptions; different industries may have different outcomes; short-term vs. long-term effects may differ.

11. Exercises

Complete the following exercises. For each, show your reasoning and final answer, then check against the feedback.

📝 Exercise 11.2.1: Align Evidence with Atomic Claims
  • Difficulty: Introductory
  • Estimated time: ~10 min
  • Learning outcome: Align evidence with atomic claims.

Claim: “The new software update reduces loading time and improves user satisfaction.”

Evidence: A test of the update showed a 20% reduction in loading time, but a user survey found no significant change in satisfaction scores.

How does the evidence align with each atomic claim? Classify each.

Show answer
  • Atomic Claim 1 (loading time): Directly supports (20% reduction).
  • Atomic Claim 2 (user satisfaction): Contradicts (no significant change).

Overall: The claim is partially supported; the satisfaction part is unsupported.

📝 Exercise 11.2.2: Identify Partial Support
  • Difficulty: Introductory
  • Estimated time: ~10 min
  • Learning outcome: Identify partial support.

Claim: “The intervention is effective for all age groups.”

Evidence: The intervention was effective in adults aged 18–40, but not in those over 60.

How does the evidence align with the claim?

Show answer

Answer: The evidence partially supports the claim — it supports effectiveness for one age group, but contradicts effectiveness for older adults. The claim should be qualified: “effective for adults aged 18–40; not effective for older adults.”

📝 Exercise 11.2.3: Identify Overstatement
  • Difficulty: Intermediate
  • Estimated time: ~10 min
  • Learning outcome: Identify overstatement.

Evidence: An observational study found that people who eat breakfast have lower BMI.

Claim: “Eating breakfast causes weight loss.”

What is the overstatement? Suggest corrected wording.

Show answer

Overstatement: The claim implies causation, but the evidence is correlational.

Corrected wording: “Observational studies suggest an association between eating breakfast and lower BMI; causation has not been established.”

📝 Exercise 11.2.4: Assess Source Independence
  • Difficulty: Intermediate
  • Estimated time: ~10 min
  • Learning outcome: Assess source independence.

You have three sources:

  • Source A: A 2022 study in a peer-reviewed journal.
  • Source B: A 2023 news article that cites Source A.
  • Source C: A 2023 blog post that quotes Source B.

Are these sources independent? Explain.

Show answer

Answer: No, they are not independent. Source B depends on Source A, and Source C depends on Source B (and indirectly on A). They all trace back to the same original study. They do not provide corroboration.

📝 Exercise 11.2.5: Distinguish Corroboration from Repetition
  • Difficulty: Intermediate
  • Estimated time: ~10 min
  • Learning outcome: Distinguish corroboration from repetition.

You are verifying the claim that “a new teaching method improves math scores.” You find:

  • Study 1: A randomised trial in one school.
  • Study 2: A randomised trial in a different country with similar results.
  • Study 3: A review article that summarises both Study 1 and Study 2.

Which provide corroboration? Which are repetition? Explain.

Show answer

Corroboration: Study 1 and Study 2 are independent trials (if they are truly independent). Together they provide corroboration.

Repetition: Study 3 (the review) repeats the findings of Studies 1 and 2; it does not add new evidence, though it may synthesise them.

📝 Exercise 11.2.6: Develop a Triangulation Plan
  • Difficulty: Advanced
  • Estimated time: ~15 min
  • Learning outcome: Develop a triangulation plan.

Claim: “The new software increases developer productivity.”

Describe three different types of evidence or methods you would use to triangulate this claim. Explain what each method would contribute and how they might conflict.

Show answer

Possible triangulation:

  • Quantitative metrics: Lines of code, commits, or story points completed before/after.
  • Surveys: Developer self-reported productivity and satisfaction.
  • Qualitative interviews: Developer perceptions of workflow and efficiency.
  • Project timelines: Time to deliver features before/after.

Potential conflicts: Metrics may show improvement while surveys show frustration; qualitative interviews may reveal that the software is hard to use, leading to reduced productivity in the long run.

📝 Exercise 11.2.7: Communicate Conflicting Evidence
  • Difficulty: Advanced
  • Estimated time: ~12 min
  • Learning outcome: Communicate conflicting evidence.

Claim: “Coffee consumption reduces the risk of heart disease.”

You find:

  • Study A: A meta-analysis of 10 cohort studies found a 15% reduction in risk (statistically significant).
  • Study B: A recent large randomised trial found no significant effect.

How would you communicate this conflicting evidence in a responsible way? Draft a brief statement.

Show answer

Draft statement:

“Evidence on coffee consumption and heart disease risk is conflicting. A 2022 meta-analysis of observational studies suggested a 15% reduction in risk, but a recent large randomised controlled trial found no significant effect. Given the limitations of observational studies (confounding) and the strength of the RCT, the association remains uncertain. More rigorous studies are needed before any definitive conclusions can be drawn.”

12. Homework Problems

Formative · Ungraded · Recommended preparation for Assignment 5 and the Unit 11 cumulative project.

🧠 Problem 11.2.1: Claim-Evidence Alignment

Claim: “The new policy reduced traffic accidents.”

Evidence: After the policy, the number of traffic accidents decreased by 5%, but the city also implemented other measures, such as speed bumps and traffic cameras.

How does the evidence align with the claim? Identify any confounding factors.

Show solution design

Alignment: The evidence supports the claim, but other measures confound the effect. The claim should be qualified: “After the policy was implemented, accidents decreased by 5%; however, other interventions were also introduced, so the policy’s specific effect is uncertain.”

🧠 Problem 11.2.2: Wording Strength

Evidence: A single small pilot study found that a new therapy reduced depressive symptoms in 10 participants.

Draft a sentence about this finding that accurately reflects the evidence.

Show answer

Possible sentence: “A small pilot study (n=10) found that the therapy was associated with a reduction in depressive symptoms; however, due to the small sample size and lack of a control group, these findings are preliminary and require confirmation in larger studies.”

🧠 Problem 11.2.3: Source Independence

You have two sources: Source A is a government report on unemployment rates. Source B is a news article that cites Source A. Are they independent for the claim “unemployment is 6%”? Explain.

Show answer

Answer: No, they are not independent. Source B depends on Source A. They do not corroborate each other; they repeat the same information. The authoritative source is Source A (the government report).

🧠 Problem 11.2.4: Corroboration vs. Repetition

You are verifying the claim “the new drug reduces pain.” You find:

  • Study 1: A randomised trial by the drug company.
  • Study 2: An independent randomised trial by a university.
  • Study 3: A blog post that summarises Study 1.

Which provide corroboration? Which are repetition? Explain.

Show answer

Corroboration: Study 2 (if truly independent) provides corroboration. Study 1 and Study 2 are independent trials.

Repetition: Study 3 repeats Study 1. Also, Study 1 may have conflict of interest; Study 2 is stronger corroboration if it confirms the findings.

🧠 Problem 11.2.5: Triangulation

Claim: “A new teaching method improves student critical thinking.”

Describe a triangulation strategy using at least three different types of evidence. What would each contribute? How might they conflict?

Show solution design

Possible plan:

  • Standardised test: Measure critical thinking scores before and after (quantitative).
  • Student portfolios: Analyse samples of student work for evidence of critical thinking (qualitative).
  • Teacher observations: Teachers’ assessments of student discussion and reasoning (observational).
  • Student self-reports: Surveys on perceived improvement in thinking skills.

Potential conflicts: Test scores may improve while portfolio analysis shows superficial gains; teacher observations may differ from student self-reports.

🧠 Problem 11.2.6: Conflicting Evidence

Claim: “Social media use is linked to depression in teenagers.”

You find:

  • Study A: A longitudinal study found a positive correlation.
  • Study B: Another longitudinal study found no correlation after controlling for family support.

Draft a balanced statement that communicates the state of evidence.

Show answer

Draft statement: “Research on social media and teen depression is mixed. Some longitudinal studies have reported a positive association, while others have found no significant effect after accounting for factors like family support. The relationship may be complex and context-dependent. Further investigation is needed to clarify the role of social media, and current evidence does not support a definitive causal claim.”

🧠 Problem 11.2.7: Evidence Gap

Claim: “The new policy reduced crime in the city.”

Available evidence: One internal report from the police department, not peer-reviewed, showing a 10% decrease in reported crime.

What are the evidence gaps? What additional evidence would be needed?

Show answer
  • Gaps: No independent verification; report may be biased; no control for other factors (e.g., economic changes); crime may be underreported; no long-term data.
  • Additional evidence: Independent evaluation, control areas, victimisation surveys, analysis of reporting practices.
🧠 Problem 11.2.8: Transfer Problem — Apply to Your Field

Identify a claim from your field or profession that you have recently encountered. Find two sources that address it. Evaluate alignment, independence, and whether they provide corroboration or repetition. Write a brief assessment.

Show solution design
  1. State the claim clearly.
  2. Identify two sources (real or synthetic).
  3. For each source, evaluate alignment: Does it directly support, partially support, etc.?
  4. Assess independence: Do they share a common origin?
  5. Determine if they provide corroboration or repetition.
  6. Conclude whether the claim is supported or needs qualification.

13. Mini-Project: Claim-Evidence Alignment and Corroboration Review

📁 Project: Claim-Evidence Alignment and Corroboration Review
  • Purpose: Apply alignment, independence, corroboration, and triangulation skills to a set of claims and sources.
  • Learning outcomes: Align evidence with claims, assess independence, identify corroboration vs. repetition, develop triangulation, communicate conflicting evidence.
  • Estimated effort: 90–150 minutes.

Supplied claim set

Claim Set A:

  • Claim 1: “The new AI-based tutoring system improves math scores by 20%.”
  • Claim 2: “The system reduces student anxiety.”
  • Claim 3: “The system is equally effective for all students regardless of prior knowledge.”

Supplied sources

  • Source 1: A 2023 randomised controlled trial (RCT) conducted by the system’s developers, published in an open-access journal. The trial found a 20% improvement in math scores for students using the system, but only for those with average prior knowledge. It found no significant effect on anxiety.
  • Source 2: A 2024 independent evaluation by a university research group that replicated the RCT with a larger sample. They found a 15% improvement (not statistically different from 20%) and a small reduction in anxiety (p=0.05).
  • Source 3: A 2024 news article that cites Source 1 and says “AI tutor boosts math scores by 20% and reduces anxiety.”
  • Source 4: A 2023 blog post by the system’s CEO, describing the system’s features and citing Source 1.

Required deliverable

Complete the following sections for the chosen claim set (or your own). Use the supplied sources or find your own.

  1. Human purpose: Why is this verification being done? (e.g., to decide whether to adopt the system)
  2. Intended audience: Who will use the verified information? (e.g., school administrators)
  3. Atomic claim list: Break down each claim into atomic propositions.
  4. Evidence required for each claim: What kind of evidence would be needed?
  5. Source assigned to each claim: Which source(s) address each atomic claim?
  6. Source authority: Evaluate each source’s authority.
  7. Source relevance: Evaluate relevance.
  8. Source currency: Evaluate currency.
  9. Claim-evidence alignment: For each atomic claim, classify alignment (direct support, partial, qualification, contradiction, not address).
  10. Source independence: Map relationships: which sources are independent? Which are dependent?
  11. Shared-source relationships: Identify any common origins.
  12. Corroboration: Which claims have independent corroboration?
  13. Triangulation: What different methods or perspectives are represented?
  14. Contradictions: Are there any contradictions among sources?
  15. Missing qualifications: What important qualifications are missing from the claims?
  16. Evidence gaps: What is missing?
  17. Revised wording: Rewrite each claim to accurately reflect the evidence.
  18. Remaining uncertainty: State what remains uncertain.
  19. Specialist-review need: Is specialist review required?
  20. Preliminary recommendation: Based on the evidence, what is your recommendation?

Assessment criteria

  • Accurate atomic claim decomposition
  • Correct alignment classification
  • Sound independence assessment
  • Clear identification of corroboration vs repetition
  • Appropriate revised wording
  • Comprehensive documentation of uncertainty and gaps
  • Professional, evidence-based recommendation
Show staged hints

Hint 1: Start with atomic claims. Break each claim into the smallest meaningful propositions. For example, “improves math scores by 20%” is one atomic claim; “reduces anxiety” is another; “equally effective” is a third.

Hint 2: Assess independence carefully. Source 1 and Source 2 are independent (different research groups). Source 3 and Source 4 depend on Source 1 (and possibly Source 2). So for corroboration, look at Sources 1 and 2.

Hint 3: Revise wording to match evidence. For Claim 1: “The system improves math scores by 20%” — Source 1 found 20% for average prior knowledge; Source 2 found 15%. So a revised wording could be: “Two studies, one by the developers and one independent, found improvements of 15–20% in math scores, but effects may vary by prior knowledge.”

For Claim 2: Anxiety reduction is not supported by Source 1, but Source 2 found a small effect (p=0.05). So the claim is partially supported with weak evidence.

For Claim 3: No evidence supports equal effectiveness; in fact, Source 1 suggests it only works for average prior knowledge. Claim 3 is unsupported.

Show illustrative partial example (different claim)

Claim: “A new drug reduces blood pressure by 10%.”

Source 1: Manufacturer’s trial: 10% reduction, but no control group.

Source 2: Independent trial: 8% reduction, placebo-controlled.

Alignment: Source 1 supports the claim (but weak design). Source 2 partially supports (8% vs 10%).

Independence: Sources are independent.

Corroboration: Both show reduction, so corroboration (though magnitude differs).

Revised wording: “Two trials, one independent, suggest a reduction of 8–10% in blood pressure; however, one trial lacked a control group. Further confirmatory studies are needed.”

14. Self-Check Quiz

9 questions · Formative · Ungraded. Expand each item to see feedback.

Question 1: If evidence supports only part of a claim, the alignment is best described as:

Select one:
Show answer and explanation

Answer: b.

When evidence supports some aspects of a claim but not others, it provides partial support.

Question 2: Which of the following is an example of overstatement?

Select one:
Show answer and explanation

Answer: b.

“Proved” and “causes” are strong claims that often exceed the evidence, especially from a single study.

Question 3: Two news articles that both cite the same scientific study are considered:

Select one:
Show answer and explanation

Answer: c.

They are not independent; they repeat the same source. They do not provide corroboration.

Question 4: What is triangulation?

Select one:
Show answer and explanation

Answer: b.

Triangulation uses diverse evidence to reduce bias and increase confidence, but it does not automatically resolve conflicts.

Question 5: When credible sources disagree on a claim, the responsible approach is to:

Select one:
Show answer and explanation

Answer: c.

Responsible verification acknowledges conflicting evidence and explains why it exists.

Question 6: An evidence gap exists when:

Select one:
Show answer and explanation

Answer: d.

Any of these situations creates an evidence gap that must be documented.

Question 7: Which of the following is NOT a sign of source dependence?

Select one:
Show answer and explanation

Answer: d.

This is a sign of independence, not dependence. Different authors and methods suggest they are independent.

Question 8: In the CLAIM framework, assessing claim-evidence alignment helps you to:

Select one:
Show answer and explanation

Answer: b.

Alignment ensures that the evidence is relevant and matches the claim’s scope and strength.

Question 9: When triangulating evidence, if different methods yield conflicting results, you should:

Select one:
Show answer and explanation

Answer: b.

Triangulation is not about forcing agreement but about understanding the evidence landscape, including disagreements.

15. Assignment Connections

📝 Assignment 5: Verification and Fact-Checking Assignment

In Assignment 5, you will need to align each claim in an AI-generated passage with the evidence you find. You will classify each claim’s alignment (support, partial, qualification, contradiction, not addressed) and evaluate whether multiple sources provide independent corroboration or mere repetition.

As you complete this lesson, retain:

  • Your ability to decompose claims and assess alignment.
  • Your skill in identifying overstatement and selecting proportionate wording.
  • Your understanding of source independence and how to distinguish corroboration from repetition.
  • Your approach to triangulation and handling conflicting evidence.
  • Your documentation of evidence gaps and uncertainty.

In Assignment 5, you will produce a claim-evidence alignment table that documents each claim’s alignment and a correction record. The exercises and mini-project in this lesson directly prepare you for that task.

📓 Assignment 1 journal prompts

Reflect on the following questions in your Assignment 1 journal.

  1. How does claim-evidence alignment change your evaluation of a source? Describe a specific example where you found that a source did not actually support the claim as stated.
  2. Why are repeated sources not independent? Provide an example from your own experience or from the lesson.
  3. What did you learn about triangulation? How can you use different types of evidence to strengthen (or qualify) a claim?
  4. How should conflicting evidence be communicated? Why is it important to preserve disagreement rather than hide it?
  5. How has your approach to verifying claims changed after this lesson? What will you do differently when you encounter a claim with multiple sources?

Write in your own voice; there are no model responses.

Show reflection planning checklist
  • I have discussed a specific example of claim-evidence alignment.
  • I have explained why independence matters.
  • I have described a triangulation plan or example.
  • I have reflected on how to handle conflicting evidence.
  • I have connected my reflection to my own verification practice.

16. Key Takeaways

Remember these points
  • Claim-evidence alignment is the core of verification: evidence may support, partially support, qualify, contradict, or not address a claim.
  • Wording must match the evidence — avoid overstatement; use cautious language when evidence is weak.
  • Source independence is essential for corroboration; repeated sources are not independent.
  • Corroboration comes from independent sources with different origins, not from multiple citations of the same study.
  • Triangulation uses diverse methods or sources to build a fuller picture, but it does not eliminate uncertainty.
  • Conflicting evidence must be preserved and explained, not hidden or averaged away.
  • Evidence gaps are honest assessments — document them rather than pretend they don’t exist.
  • Always ask: Does the evidence actually support the claim as stated?

17. Key Terms

Claim-evidence alignment
The relationship between a claim and the evidence, classified as direct support, partial support, support with qualification, contradiction, or not addressing.
Overstatement
Using wording that is stronger than the evidence supports (e.g., “proves” when the evidence is correlational).
Source independence
The condition that two sources were produced separately and do not share the same origin, data, or reporting chain.
Corroboration
Independent evidence that supports the same claim, increasing confidence.
Repetition
Multiple sources that rely on the same original evidence; they do not provide independent support.
Triangulation
The practice of using different sources, methods, or datasets to cross-check and deepen understanding.
Conflicting evidence
Evidence from credible sources that supports different conclusions.
Evidence gap
A situation where the available evidence is insufficient to support or refute a claim.
Qualification
A statement of conditions or limitations that must accompany a claim to accurately reflect the evidence.
Wording strength
The degree of certainty expressed by language (e.g., “proves” vs. “suggests”).

18. Looking Ahead: Lesson 11.3

You have now learned to align evidence with claims, assess independence, and handle corroboration and triangulation. In Lesson 11.3, you will build on these skills to:

These skills will help you detect fabricated citations, misattributions, and chains of repetition that obscure the original evidence.

After Lesson 11.3, you will complete Lesson 11.4, which integrates TRACE and CLAIM to produce corrected wording and final dispositions. The cumulative Unit 11 project will synthesise all of these skills.

Continue to Lesson 11.3: Source Verification and Citation Tracing Return to Unit 11 Home

19. Sources & Further Reading

[VERIFICATION REQUIRED: External links and institutional resources should be reviewed for currency and accessibility.]

⚠️ Note

Additional authoritative sources may be provided by your instructor. If you are unable to locate a source, use the supplied-source alternatives in the activities.

👤