A Level H1 General Paper Comprehension Quiz

A-Level General Paper H1 Quiz - Comprehension: Answer Key

Topic: Comprehension
Total Marks: 50

Section A: Comprehension — Short-Answer Questions [35 marks]

1. According to the passage, in what two main areas is AI being deployed in healthcare? (line 2) [2]

Answer: Diagnostic imaging and drug discovery. [1 mark for each, total 2 marks]

Explanation: The passage explicitly states in line 2: "From diagnostic imaging to drug discovery, AI systems are increasingly being deployed in clinical settings." Students should identify both areas directly from the text. This is a literal comprehension question testing the ability to locate specific information.

2. What does the author mean by the phrase "unparalleled advantages" in line 5? Use your own words as far as possible. [2]

Answer: Benefits or advantages that have never been matched or exceeded by anything else / benefits that are superior to all others. [2 marks for a clear paraphrase; 1 mark for a partial answer]

Explanation: "Unparalleled" means having no parallel or equal — unmatched. Students must avoid simply lifting "advantages that are unparalleled" and instead convey the meaning in fresh language. Acceptable paraphrases include "benefits that cannot be matched," "superior advantages," or "advantages beyond comparison."

3. Explain the author's use of the word "tempered" in line 10. [2]

Answer: The author uses "tempered" to mean that the enthusiasm for AI should be moderated, restrained, or balanced by a sober recognition of its limitations. [2 marks for a clear explanation; 1 mark for identifying "moderated" or "restrained" alone]

Explanation: "Tempered" in this context does not mean "weakened" entirely but rather "balanced" or "qualified." The author is arguing that while the enthusiasm is understandable, it must be accompanied by a realistic assessment of AI's drawbacks. A common student trap is to say "reduced" without the nuance of balance.

4. The author states that AI systems are "only as reliable as the data on which they are trained." (lines 12–13) What example does the author provide to support this claim? Answer in your own words. [3]

Answer: The author cites a 2019 study in Science about a commercial algorithm used in American hospitals that underestimated the health needs of Black patients. This happened because the algorithm used healthcare spending as a proxy for illness severity, which encoded existing racial disparities — since Black patients historically had less spent on their care, the algorithm incorrectly concluded they were healthier. [3 marks: 1 for identifying the study, 1 for explaining the mechanism (spending as proxy), 1 for stating the consequence (underestimation of Black patients' needs)]

Explanation: This question tests the student's ability to explain a complex example in their own words. Direct lifting from the passage should be penalised. The key is to show understanding of the causal chain: biased data → biased proxy → biased outcome.

5. In lines 15–17, the author refers to a study published in Science. What was the key finding of this study, and why does the author consider it significant? [3]

Answer: The key finding was that a widely used commercial algorithm systematically underestimated the health needs of Black patients because it used healthcare spending as a proxy for illness severity, inadvertently encoding racial disparities. [2 marks] The author considers this significant because it demonstrates that AI does not eliminate human bias but can actually amplify and institutionalise it, making the bias harder to detect and correct. [1 mark]

Explanation: Students must distinguish between what the study found and why the author uses it as evidence. The significance lies in the broader argument about AI perpetuating systemic bias, not just the specific finding about one algorithm.

6. What does the author mean by describing AI systems as "black boxes" in line 20? [2]

Answer: It means that the internal decision-making processes of AI systems are opaque or hidden — even their developers cannot fully explain how a particular decision or recommendation was reached. [2 marks: 1 for "opaque/hidden processes," 1 for noting that even developers cannot explain them]

Explanation: The "black box" metaphor is common in technology discourse. Students should avoid vague answers like "we don't know what's inside" and instead convey the specific idea of unexplainable decision-making processes.

7. According to the passage, why is the "black box" nature of AI problematic in a medical context? Explain in your own words. [3]

Answer: In medicine, patients have the right to understand the reasoning behind a diagnosis or treatment recommendation. If an AI system operates as a black box, neither the doctor nor the patient can understand how the recommendation was made. This creates problems of accountability — if the recommendation proves harmful, it is unclear who is responsible: the doctor, the hospital, or the AI developer. [3 marks: 1 for patient's right to understand reasoning, 1 for the inability to explain the recommendation, 1 for the accountability gap]

Explanation: This is an inference question. Students must connect the "black box" concept to the specific context of healthcare, where transparency and accountability are ethically and legally essential.

8. In lines 24–27, the author poses a series of questions about accountability. What is the purpose of these questions in the context of the passage? [2]

Answer: The purpose is to highlight the unresolved issue of who bears responsibility when an AI system's recommendation causes harm. The rhetorical questions emphasise the complexity and urgency of establishing clear accountability frameworks for medical AI. [2 marks: 1 for identifying the accountability issue, 1 for noting the rhetorical effect of emphasising complexity/urgency]

Explanation: These are rhetorical questions — the author does not expect the reader to answer them but uses them to provoke thought and underscore a gap in current legal and regulatory frameworks.

9. Explain the irony mentioned in lines 33–35 regarding AI in healthcare for lower-income countries. [2]

Answer: The irony is that lower-income countries, which have the greatest need for AI-assisted healthcare due to limited access to medical specialists, are the least able to afford the expensive development and implementation of these AI systems. [2 marks: 1 for identifying the great need, 1 for identifying the inability to afford it]

Explanation: Irony involves a contrast between expectation and reality. Students should clearly articulate both sides: the potential benefit and the financial barrier.

10. What does the author mean by "the relational dimension of care" in line 39? Use your own words. [2]

Answer: The aspect of healthcare that involves human connection, trust, empathy, and communication between doctor and patient — the emotional and interpersonal elements that go beyond purely technical medical treatment. [2 marks: 1 for human connection/trust/empathy, 1 for the idea that it goes beyond technical treatment]

Explanation: Students should avoid simply rephrasing "relational" as "relationship" without elaborating on what that relationship entails in a medical context.

11. In line 40, the author states: "The danger is not that AI will replace doctors, but that the increasing presence of AI could subtly alter the nature of medical practice..." What does the author mean by this distinction? [3]

Answer: The author distinguishes between two types of threat: the obvious threat of AI completely replacing human doctors (which the author dismisses as unlikely or not the primary concern), and the more subtle threat that AI could gradually change how medicine is practised — potentially reducing the humanistic, empathetic aspects of care and turning doctors into machine operators rather than compassionate caregivers. [3 marks: 1 for explaining the first part (replacement), 1 for explaining the second part (subtle alteration), 1 for identifying the humanistic core that is at risk]

Explanation: This question tests the student's ability to analyse a nuanced distinction. The author is making a sophisticated argument that the real danger is not dramatic replacement but gradual erosion of medical humanism.

12. What is the author's overall attitude towards AI in healthcare? Identify two words or phrases from the passage that support your answer. [3]

Answer: The author's attitude is balanced/cautiously optimistic — neither wholly supportive nor wholly opposed. The author acknowledges both the promise and the risks of AI in healthcare. [1 mark for identifying the attitude]

Supporting evidence (any two of the following, 1 mark each):

• "tempered by a sober recognition of its limitations" (line 10)
• "it would be a mistake to adopt a purely reactionary stance" (line 43)
• "a balanced, evidence-based approach" (line 46)
• "harnesses the strengths of AI while remaining vigilant about its weaknesses" (lines 46–47)

Explanation: Students must identify the nuanced, balanced attitude rather than labelling the author as simply "for" or "against" AI. The supporting evidence must come from the passage.

13. In lines 44–46, the author uses the phrase "as unwise as embracing it uncritically." What rhetorical technique is being employed here, and what effect does it have? [2]

Answer: The technique is a balanced comparison / parallel structure (simile/comparative construction) that equates two extremes — outright rejection and uncritical acceptance — as equally foolish. [1 mark for identifying the technique] The effect is to position the author as reasonable and moderate, rejecting both extremes and advocating for a middle path. [1 mark for the effect]

Explanation: This question tests recognition of rhetorical technique. The parallel structure ("as unwise as") is a deliberate rhetorical choice that reinforces the author's argument for balance.

14. According to the final paragraph, what two things are needed to ensure AI is used responsibly in healthcare? [2]

Answer: (i) Regulatory bodies must establish clear guidelines for the development and deployment of medical AI, ensuring rigorous testing, transparency, and ongoing monitoring for bias and error. [1 mark] (ii) Future healthcare professionals must be trained not only to use AI tools but to understand their limitations and to maintain the human qualities essential to good medical practice. [1 mark]

Explanation: This is a literal comprehension question requiring students to identify two distinct recommendations from the final paragraph.

15. In your own words, summarise the author's conclusion about the future of AI in healthcare as presented in the final paragraph. [2]

Answer: The author concludes that AI will inevitably play a role in healthcare, and the crucial issue is how society chooses to govern its use. The decisions made now about regulation, training, and ethical standards will have long-lasting effects on the future of medicine, requiring both human wisdom and technological innovation. [2 marks: 1 for the idea of governance/choice, 1 for the long-term consequences]

Explanation: Students must paraphrase the conclusion rather than lift directly. The key ideas are: inevitability of AI, importance of governance, and long-term impact of current decisions.

Section B: Summary [8 marks]

16. Summary of concerns raised by the author about the use of AI in healthcare (paragraphs 3, 5, and 5 — lines 12–40). [8 marks]

Content Points (1 mark each, up to 8 marks):

The following are the relevant content points from paragraphs 3, 4, and 5:

1. AI systems are only as reliable as their training data, so skewed or unrepresentative data leads to biased outcomes.
2. AI can amplify and institutionalise existing human biases (e.g., racial disparities in healthcare algorithms).
3. AI systems operate as "black boxes" — their decision-making processes are opaque and cannot be fully explained.
4. This lack of transparency is problematic in medicine, where patients have a right to understand the reasoning behind diagnoses and treatments.
5. There is a lack of clarity about accountability — it is unclear who is responsible when AI recommendations cause harm (doctor, hospital, or developer).
6. The high initial investment required for AI systems may widen the digital divide between wealthy and lower-income nations.
7. Lower-income countries that need AI most may be least able to afford it.
8. Over-reliance on AI could erode the doctor-patient relationship and the humanistic core of medicine.
9. AI could reduce patients to data points and clinicians to machine operators.
10. The compassionate, empathetic aspects of medical care could be diminished.

Marking Scheme:

• Content: Up to 8 marks — 1 mark for each relevant point accurately paraphrased in the student's own words. Points must be drawn only from paragraphs 3, 4, and 5.
• Language: Assessed holistically — students should use their own words, write in continuous prose, and stay within the 120-word limit.

Common Mistakes:

• Including points from outside paragraphs 3–5 (e.g., the benefits of AI from paragraph 2).
• Lifting phrases directly from the passage instead of paraphrasing.
• Writing in bullet points or note form instead of continuous prose.
• Exceeding the 120-word limit.

Section C: Application Question [7 marks]

17. Using ideas from the passage, explain why the SMA's concern about the training data of MediScan AI is valid. [2]

Answer: The passage explains that AI systems are only as reliable as their training data. If the data under-represents certain groups, the algorithm may produce biased outcomes. The example of the Science study showed that an algorithm trained on biased data underestimated the health needs of Black patients. [1 mark for linking to the passage's concept, 1 mark for applying it to MediScan AI's East Asian training data and the risk for other ethnic groups]

Explanation: Students must connect the general principle from the passage (biased data → biased outcomes) to the specific scenario (East Asian data → potential inaccuracy for other ethnicities in Singapore's multiracial population).

18. The passage discusses the issue of transparency in AI systems. How does this issue relate to the SMA's concern about MediScan AI's proprietary algorithm? Explain with reference to the passage. [2]

Answer: The passage describes AI systems as "black boxes" whose decision-making processes cannot be fully explained, even by their developers. This opacity is problematic in medicine because patients and doctors need to understand the reasoning behind a diagnosis. [1 mark] The SMA's concern about MediScan AI's proprietary algorithm being unauditable directly mirrors this issue — if the algorithm cannot be independently examined, its reliability and fairness cannot be verified, creating risks for patient safety and accountability. [1 mark for applying the passage concept to the scenario]

Explanation: Students must explicitly reference the "black box" concept from the passage and draw a clear parallel to the proprietary, unauditable nature of MediScan AI.

19. The survey finding that 62% of patients felt uncomfortable with AI-generated diagnoses echoes a concern raised in the passage. Identify this concern and explain how it might affect the success of MediScan AI's rollout. [2]

Answer: The passage warns that over-reliance on AI could erode the doctor-patient relationship and the humanistic core of medicine, reducing patients to data points. [1 mark] If a majority of patients are uncomfortable with AI-generated diagnoses, they may distrust the system, resist using polyclinics that deploy it, or lose confidence in their doctors — undermining the tool's effectiveness and the government's goals. [1 mark for explaining the practical impact on the rollout]

Explanation: This question tests the student's ability to apply a conceptual concern from the passage to a real-world scenario with practical implications.

20. The Ministry describes MediScan AI as a "decision-support tool." Does this address the accountability concerns raised in the passage? Explain your reasoning in no more than 3–4 sentences. [1]

Answer: Partially. The passage raises the question of who is responsible when an AI recommendation causes harm. By positioning MediScan AI as a "decision-support tool" with mandatory doctor review, the Ministry implies that the doctor retains final responsibility. [1 mark] However, this does not fully resolve the accountability issue — if the AI's recommendation heavily influences the doctor's decision, the question of shared responsibility between doctor and AI developer remains unclear, as the passage suggests.

Explanation: This is a higher-order question requiring evaluative reasoning. Students should acknowledge that the "decision-support" framing partially addresses accountability but does not fully resolve the complexity raised in the passage.

Total: 50 marks