A Level H1 General Paper Practice Paper 1

TuitionGoWhere Practice Paper — General Paper H1 A-Level

Answer Key — Practice Paper 2, Version 1 of 5

Section A: Comprehension & Analysis (Questions 1–15)

Question 1 (2 marks)

Answer: The phrase "the future role of human practitioners" refers to how doctors, nurses, and other medical professionals will function in a healthcare landscape increasingly dominated by AI. It questions whether human clinicians will remain central to patient care or become secondary to automated systems, and how their responsibilities, skills, and professional identity may change as AI takes over tasks traditionally performed by humans.

Marking Notes:

• 1 mark for identifying that it concerns how doctors/medical professionals will operate alongside AI.
• 1 mark for explaining the uncertainty about whether their role will diminish, change, or remain essential.
• Students must use their own words. Direct lifting of "future role of human practitioners" without explanation = 0 marks.

Question 2 (2 marks)

Answer: The word "surpasses" means to exceed or do better than. In this context, it means that AI systems have, in some cases, demonstrated greater accuracy in detecting cancers than experienced radiologists. The effect of this word choice is striking and somewhat provocative — it challenges the assumption that human expertise is superior and positions AI as a potentially better alternative, thereby strengthening the pro-AI argument and capturing the reader's attention.

Marking Notes:

• 1 mark for explaining the meaning of "surpasses" (exceeds/does better than).
• 1 mark for explaining the rhetorical effect (challenges assumptions, strengthens the pro-AI position, creates impact).
• Accept any reasonable explanation of rhetorical effect.

Question 3 (3 marks)

(a) (1 mark)

Answer: The study found that an AI system outperformed six radiologists in reading mammograms, reducing both false positives (incorrectly identifying cancer where none exists) and false negatives (failing to detect cancer that is present).

Marking Notes:

• 1 mark for identifying that the AI outperformed radiologists in mammogram reading.
• Must mention either false positives or false negatives, or the general finding of superior accuracy.

(b) (2 marks)

Answer: The author includes this reference to lend credibility and authority to the argument that AI can be highly effective in healthcare diagnostics. By citing a prestigious peer-reviewed journal (Nature), the author appeals to scientific authority, making the claim more persuasive. It also serves as concrete evidence to support the broader point that AI has immense potential benefits in healthcare.

Marking Notes:

• 1 mark for identifying the purpose of lending credibility/authority to the argument.
• 1 mark for explaining how the prestigious source (Nature) strengthens the persuasiveness of the claim.
• Accept alternative valid purposes (e.g., providing concrete evidence, supporting the proponents' case).

Question 4 (2 marks)

Answer: By "paradigm shift," Dr. Vasquez means a fundamental and dramatic change in the way healthcare is practised. It suggests that AI is not merely an incremental improvement but a transformative force that is fundamentally altering the methods, assumptions, and structures of medical practice. The shift is so significant that the old model of healthcare is being replaced by a new one in which AI plays a central role.

Marking Notes:

• 1 mark for explaining "paradigm shift" as a fundamental/dramatic change.
• 1 mark for contextualising it within healthcare (transforming how medicine is practised).
• Must use own words. Simply restating "paraphrase shift" = 0 marks.

Question 5 (2 marks)

Answer: "Algorithmic bias" refers to systematic errors or unfairness in AI systems that arise from the data on which they are trained. If the training data is unrepresentative or reflects existing prejudices, the AI will produce biased outcomes. An example from the passage is dermatology AI tools that were trained predominantly on images of lighter skin tones, causing them to perform poorly when diagnosing conditions in patients with darker skin.

Marking Notes:

• 1 mark for defining algorithmic bias (systematic errors from unrepresentative/flawed training data).
• 1 mark for providing the dermatology example or the Science study example about racial bias in healthcare needs prediction.
• Either example from the passage is acceptable.

Question 6 (3 marks)

(a) (1 mark)

Answer: The study found that the commercial algorithm exhibited significant racial bias by systematically underestimating the severity of illness among Black patients when predicting their healthcare needs.

Marking Notes:

• 1 mark for identifying that the algorithm underestimated illness severity in Black patients.

(b) (2 marks)

Answer: The potential consequence is that such biases could worsen existing health disparities rather than reduce them. If AI systems consistently underestimate the healthcare needs of certain racial groups, those patients may receive less attention, fewer resources, or delayed treatment, leading to poorer health outcomes for already disadvantaged populations. This would deepen inequality in healthcare rather than promoting equity.

Marking Notes:

• 1 mark for identifying that health disparities would worsen/deepen.
• 1 mark for explaining the mechanism (underrepresented groups receive poorer care/delayed treatment).
• Must go beyond simply restating the passage.

Question 7 (2 marks)

Answer: The "accountability gap" refers to the situation where an AI system causes harm (e.g., a misdiagnosis) but no individual or organisation can be clearly held responsible. Unlike a human doctor who can be sued or disciplined, an AI cannot bear legal or moral responsibility. This creates a void in which errors occur but there is no clear party — whether the hospital, the developers, or the clinicians — who can be adequately held accountable.

Marking Notes:

• 1 mark for identifying that it refers to a lack of clear responsibility when AI causes harm.
• 1 mark for explaining why this gap exists (AI cannot be held liable; legal frameworks are inadequate).
• Must use own words.

Question 8 (2 marks)

Answer: By describing medicine as "an art," the author means that medical practice involves qualities that go beyond technical knowledge and scientific procedure. These include empathy (the ability to understand and share patients' feelings), intuition (the capacity to make judgements based on experience and instinct rather than purely on data), and effective communication with patients as whole persons with emotional and psychological needs. It suggests that the human, relational aspects of medicine are just as important as the scientific ones.

Marking Notes:

• 1 mark for identifying that it refers to non-technical aspects of medicine.
• 1 mark for providing specific qualities (empathy, intuition, communication, human connection).
• Any two relevant qualities are acceptable for the second mark.

Question 9 (3 marks)

(a) (1 mark)

Answer: The survey found that 68% of patients expressed concern about AI being used in their healthcare, with many stating they preferred human judgement in matters of life and death.

Marking Notes:

• 1 mark for identifying the 68% figure and the nature of the concern.

(b) (2 marks)

Answer: The author uses this survey to demonstrate that public sentiment is wary of AI in healthcare, which supports the argument that the adoption of AI faces significant resistance from patients themselves. It adds weight to the concerns raised by critics who argue that AI could erode the doctor-patient relationship. By showing that a majority of patients are uncomfortable with AI, the author suggests that successful implementation will require addressing public trust and preferences, not just technical challenges.

Marking Notes:

• 1 mark for identifying that it shows public resistance/concern about AI.
• 1 mark for explaining how this supports the critics' argument or the broader point about the need to consider patient perspectives.

Question 10 (2 marks)

Answer: By "augments human capabilities," Dr. Tan means that AI enhances and improves what doctors can do rather than replacing them. It suggests that AI acts as a supplementary tool — handling routine or data-intensive tasks — so that doctors can perform their jobs more effectively and focus on areas where human skills are most needed, such as patient interaction and complex decision-making.

Marking Notes:

• 1 mark for explaining "augments" as enhances/supplements/improves.
• 1 mark for contextualising it (AI assists doctors rather than replacing them; handles routine tasks so doctors can focus on complex care).

Question 11 (2 marks)

Answer: This suggests that the author views technology as a means to an end rather than an end in itself. The author believes AI should serve to enhance human care by freeing clinicians from routine tasks, allowing them to devote more attention and time to the interpersonal aspects of medicine. The implication is that technology is most valuable when it supports and strengthens the human elements of healthcare rather than replacing them.

Marking Notes:

• 1 mark for identifying that technology should serve/enhance human care.
• 1 mark for explaining the relationship (AI handles routine tasks → doctors focus on patients).

Question 12 (3 marks)

Answer:

Benefit 1: AI can reduce healthcare costs by improving efficiency and reducing unnecessary procedures.

Benefit 2: AI can help countries with ageing populations and strained healthcare budgets sustain quality care by generating significant savings.

Concern: The upfront costs of AI implementation (infrastructure, training, maintenance) are substantial, and the benefits may disproportionately favour wealthier nations and institutions, widening the global health divide.

Marking Notes:

• 1 mark for each correct response (3 marks total).
• Benefits and concern must be from paragraph 8.
• Students must use their own words to some extent; direct lifting may receive reduced credit.

Question 13 (2 marks)

Answer: The author mentions Singapore to provide a concrete, real-world example of a country that is actively investing in AI healthcare. This serves to illustrate that the economic and demographic arguments for AI in healthcare are not merely theoretical — they are already being acted upon by forward-thinking nations. It also grounds the discussion in a specific context (an ageing population with a Smart Nation strategy), making the argument more tangible and relevant, particularly for readers in Singapore or similar developed nations.

Marking Notes:

• 1 mark for identifying that it provides a concrete/real-world example.
• 1 mark for explaining the purpose (illustrates that AI adoption is already happening; grounds the argument in a specific context).

Question 14 (3 marks)

Answer: Dr. Vasquez's statement is significant because it encapsulates one of the central themes of the passage: that AI is not an objective or neutral tool, but rather a product of human design that carries the values, priorities, and biases of its creators. In the context of the passage, this statement ties together the concerns about algorithmic bias (paragraph 4), the accountability gap (paragraph 5), and the need for inclusive training data and transparent regulation (paragraph 9). It serves as a reminder that the ethical challenges of AI in healthcare are not merely technical problems to be solved, but reflections of deeper societal choices about justice, equity, and compassion. The statement reinforces the passage's overall message that careful governance and inclusive design are essential.

Marking Notes:

• 1 mark for explaining the meaning of the statement (AI reflects creators' biases/priorities).
• 1 mark for connecting it to specific concerns in the passage (algorithmic bias, accountability, need for regulation).
• 1 mark for explaining its broader significance (ties together the passage's themes; reinforces the need for ethical governance).
• Award partial credit for incomplete but valid responses.

Question 15 (2 marks)

Answer: The author maintains a balanced view, presenting both the benefits and concerns of AI in healthcare in roughly equal measure. The passage dedicates paragraphs 2–3 to the potential benefits (diagnostic accuracy, drug development) and paragraphs 4–6 to the concerns (bias, accountability, erosion of human care). The concluding paragraph calls for balanced governance rather than outright rejection or uncritical acceptance. Evidence of this balance can be seen in the author's use of phrases like "proponents argue" and "critics urge caution," which give equal voice to both sides.

Alternative acceptable answer: The author leans slightly towards a cautious/measured view, as the final paragraph emphasises the need for "robust regulatory frameworks" and "inclusive training data," suggesting that the author believes the concerns must be addressed before AI can be fully embraced.

Marking Notes:

• 1 mark for stating a clear position (balanced, positive, or negative).
• 1 mark for providing relevant evidence from the passage to support the position.
• Accept "balanced" or "cautious" as valid positions, provided they are supported with evidence.

Section B: Summary (Question 16)

Question 16 (8 marks)

Model Summary:

The passage raises several concerns about AI in healthcare. Firstly, algorithmic bias is a significant issue: AI systems trained on unrepresentative data may perform poorly for underrepresented groups, as seen in dermatology tools that struggled with darker skin tones and a commercial algorithm that underestimated illness severity in Black patients, potentially worsening health disparities. Secondly, there is an accountability gap — when AI causes harm, it is unclear who should be held responsible, as current legal frameworks cannot adequately address this. Thirdly, critics worry that over-reliance on AI could erode the doctor-patient relationship by reducing medicine to a data-driven exercise and stripping away empathy and human judgement, which many patients value. Finally, the high costs of AI implementation may widen the global health divide, as wealthier nations benefit disproportionately.

(Word count: 138 — accept slightly over if content is strong; penalise if significantly over 120)

Marking Scheme:

Content Points (1 mark each, max 5):

1. Algorithmic bias from unrepresentative training data.
2. Specific example: dermatology tools performing poorly on darker skin / racial bias in healthcare needs algorithm.
3. Consequence: worsening health disparities / deepening inequality.
4. Accountability gap — no clear party responsible when AI causes harm.
5. Erosion of doctor-patient relationship / loss of empathy and human judgement.
6. Patient preference for human judgement (BMA survey).
7. High implementation costs widening the global health divide.

Award marks for any 5 of the 7 content points above.

Common Mistakes:

• Lifting directly from the passage instead of paraphrasing.
• Including benefits of AI rather than focusing on concerns.
• Exceeding the 120-word limit.
• Writing in note form instead of continuous prose.

Section C: Application (Questions 17–20)

Question 17 (2 marks)

Answer: One potential benefit is that the AI triage system could significantly reduce waiting times in emergency departments by quickly and accurately assessing patients' symptoms and assigning priority levels. This aligns with the passage's argument that AI improves efficiency in healthcare by handling routine tasks — in this case, the initial assessment — faster and potentially more accurately than manual processes, allowing hospitals to manage patient flow more effectively.

Marking Notes:

• 1 mark for identifying a valid benefit (reduced waiting times, improved efficiency, faster assessment).
• 1 mark for linking it to ideas from the passage (AI improves efficiency, handles routine tasks).

Question 18 (2 marks)

Answer: One potential risk is algorithmic bias. If the AI triage system is trained on data that underrepresents certain communities, it may misclassify patients from those groups — assigning them a lower priority than they actually need. This mirrors the passage's concern about the Science study, where the algorithm underestimated illness severity in Black patients, potentially leading to delayed or inadequate treatment for vulnerable populations.

Marking Notes:

• 1 mark for identifying a valid risk (algorithmic bias, misclassification, reduced role of nurses).
• 1 mark for linking it to ideas from the passage (algorithmic bias, underrepresented data, health disparities).

Question 19 (2 marks)

Answer: The "accountability gap" would apply to the AI triage system in the following way: if the system incorrectly classifies a patient's priority level — for example, assigning a low priority to a patient who actually has a life-threatening condition — and that patient suffers harm as a result, it would be unclear who is responsible. The hospital that deployed the system, the developers who created the algorithm, and the clinicians who relied on its output could all deflect blame. Current legal frameworks do not clearly assign liability for AI-driven decisions, creating a dangerous void in which patients may suffer without recourse.

Marking Notes:

• 1 mark for explaining the accountability gap in the context of the scenario.
• 1 mark for providing a specific example of how it might apply (misclassification leading to harm with no clear responsible party).

Question 20 (1 mark)

Answer: The Ministry of Health could ensure that the AI system is trained on diverse and inclusive datasets that adequately represent all demographic groups in the population, to minimise algorithmic bias. Alternatively, the Ministry could establish clear regulatory frameworks that define accountability when AI systems make errors, ensuring that patients have recourse if harmed.

Marking Notes:

• 1 mark for suggesting a valid measure drawn from the passage's recommendations.
• Acceptable answers include: inclusive training data, transparent algorithms, regulatory frameworks, ongoing dialogue between stakeholders, maintaining human oversight.

End of Answer Key

Section A: Questions 1–15 — 35 marks Section B: Question 16 — 8 marks Section C: Questions 17–20 — 7 marks Total: 50 marks