From Real Exams Exam Paper

A Level H1 General Paper Practice Paper 3

Free Exam-Derived Owl Alpha A Level H1 General Paper Practice Paper 3 practice paper with questions and answers for Singapore students. This page is rendered as a direct URL so the questions and answers can be discovered without pressing in-page buttons.

These static practice materials are generated from the site's syllabus and paper-generation workflow, with source and model context shown so students and parents can evaluate the material before use.

A Level H1 General Paper From Real Exams Generated by Owl Alpha Updated 2026-06-07

Back to Subject Browse Free Exam Papers

Questions

TuitionGoWhere Practice Paper — General Paper H1 A-Level

TuitionGoWhere Secondary School (AI)


Subject:	General Paper
Level:	A-Level H1
Paper:	Practice Paper 2 — Comprehension
Version:	3 of 5
Duration:	1 hour 30 minutes
Total Marks:	50
Name:	________________________
Class:	________________________
Date:	________________________

Instructions

Answer ALL questions.
Write your answers in the spaces provided.
Where questions ask you to use your own words, you must paraphrase — direct lifting from the passage will not be awarded marks.
The summary question (Section C) must be written in continuous prose within the word limit stated.
The application question (Section D) requires you to apply ideas from the passage to a new context.

Section A: Short-Answer Questions (25 marks)

Read the passage below and answer Questions 1–10.

The Rise of Artificial Intelligence in Healthcare

The integration of artificial intelligence (AI) into healthcare has accelerated at a pace that few could have anticipated a decade ago. From diagnostic imaging to drug discovery, AI systems are now capable of performing tasks that were once the exclusive domain of trained medical professionals. Proponents argue that AI has the potential to revolutionise medicine by improving accuracy, reducing costs, and expanding access to quality care in underserved regions. Yet, as with any transformative technology, the rapid adoption of AI in healthcare raises profound ethical, practical, and societal questions that demand careful scrutiny.

One of the most promising applications of AI lies in diagnostic radiology. Machine learning algorithms, trained on millions of medical images, can now detect anomalies such as tumours, fractures, and lesions with a level of precision that rivals — and in some studies surpasses — that of experienced radiologists. A landmark study published in The Lancet Digital Health found that an AI system correctly identified breast cancer in mammograms with 94.5% accuracy, compared to 88% for human specialists. Advocates of AI-assisted diagnosis contend that such tools do not replace doctors but rather augment their capabilities, allowing clinicians to focus on complex cases while routine screenings are handled efficiently by machines.

However, the enthusiasm surrounding AI diagnostics must be tempered by an awareness of its limitations. AI systems are only as reliable as the data on which they are trained. If training datasets are skewed — for instance, if they underrepresent certain ethnic groups or age brackets — the resulting algorithms may produce biased outcomes. Research from MIT and Stanford has demonstrated that some widely used diagnostic algorithms perform significantly less accurately for patients of African and Asian descent, raising serious concerns about equity in AI-driven healthcare. The phrase "garbage in, garbage out" has never been more apt.

Beyond diagnostics, AI is reshaping the pharmaceutical industry. Traditional drug development is notoriously slow and expensive, often taking over a decade and costing upwards of $2.6 billion to bring a single drug to market. AI-driven platforms can analyse vast molecular databases, predict how compounds will interact with biological targets, and identify promising drug candidates in a fraction of the time. During the COVID-19 pandemic, AI played a pivotal role in accelerating vaccine development, with companies like Moderna leveraging machine learning to optimise mRNA sequences in record time. Even so, critics caution that AI cannot fully replicate the serendipitous nature of scientific discovery, and over-reliance on algorithmic predictions may narrow the scope of research.

The ethical dimensions of AI in healthcare are equally complex. Questions of data privacy loom large: AI systems require enormous volumes of patient data to function effectively, yet patients may not fully understand how their information is being used or shared. The concept of informed consent becomes murky when data is repurposed for algorithms that patients never explicitly agreed to support. Furthermore, the issue of accountability remains unresolved. If an AI system misdiagnoses a condition and a patient suffers harm, who bears responsibility — the developer of the algorithm, the hospital that deployed it, or the clinician who relied on its output? Current legal frameworks are ill-equipped to address such questions.

There is also the matter of the doctor-patient relationship. Medicine has long been regarded as a fundamentally human endeavour, rooted in empathy, trust, and the nuanced art of clinical judgement. Critics warn that an over-dependence on AI could erode these qualities, reducing patients to data points and clinicians to mere operators of software. The author Atul Gawande has observed that "the greatest clinicians are those who combine scientific rigour with deep human connection" — a balance that no algorithm can replicate. Nevertheless, even Gawande acknowledges that AI, when used judiciously, can free physicians from administrative burdens and allow them to devote more time to meaningful patient interaction.

Ultimately, the question is not whether AI will transform healthcare — it already has — but whether society can harness its benefits while mitigating its risks. This requires robust regulatory frameworks, transparent algorithms, diverse training data, and an ongoing dialogue between technologists, clinicians, ethicists, and the public. The stakes could not be higher: getting it right could save millions of lives; getting it wrong could deepen existing inequalities and undermine trust in medicine itself.

Question 1 (2 marks)

According to the passage, what are two advantages of using AI in diagnostic radiology?

Answer: _______________________________________________________________

Question 2 (2 marks)

Explain the author's use of the phrase "garbage in, garbage out" in line 24. Use your own words as far as possible.

Answer: _______________________________________________________________

Question 3 (2 marks)

What does the word "tempered" (line 19) suggest about the author's attitude towards AI diagnostics?

Answer: _______________________________________________________________

Question 4 (3 marks)

According to the passage, what concerns have been raised about the reliability of AI diagnostic systems? Identify two distinct concerns and explain each in your own words.

(a) _______________________________________________________________

(b) _______________________________________________________________

Question 5 (2 marks)

In paragraph 3, the author states that AI-driven drug development is faster and cheaper than traditional methods. What evidence does the author provide to support this claim?

Answer: _______________________________________________________________

Question 6 (2 marks)

Explain the author's use of the word "serendipitous" (line 30) in the context of scientific discovery.

Answer: _______________________________________________________________

Question 7 (2 marks)

What does the author mean by the phrase "the concept of informed consent becomes murky" (lines 35–36)?

Answer: _______________________________________________________________

Question 8 (3 marks)

According to the passage, what are the three main ethical concerns associated with the use of AI in healthcare?

(a) _______________________________________________________________

(b) _______________________________________________________________

Question 9 (3 lines — 3 marks)

The author quotes Atul Gawande: "the greatest clinicians are those who combine scientific rigour with deep human connection." Why does the author include this quotation? What point is the author reinforcing?

Answer: _______________________________________________________________

Question 10 (4 marks)

In the final paragraph, the author states: "The stakes could not be higher: getting it right could save millions of lives; getting it wrong could deepen existing inequalities and undermine trust in medicine itself."

Explain how the structure of this sentence reinforces the author's argument. Consider the use of contrast and language choices.

Answer: _______________________________________________________________

Section B: Vocabulary in Context (5 marks)

Questions 11–15: For each question, select the option (A, B, C, or D) that best captures the meaning of the word or phrase as used in the passage. Write your answer in the space provided.

Question 11 (1 mark)

The word "augment" (line 14) is closest in meaning to:

A) replace B) enhance C) diminish D) evaluate

Answer: ________

Question 12 (1 mark)

The word "contend" (line 12) is closest in meaning to:

A) admit B) argue C) concede D) doubt

Answer: ________

Question 13 (1 mark)

The word "judiciously" (line 44) is closest in meaning to:

A) reluctantly B) sparingly C) wisely D) experimentally

Answer: ________

Question 14 (1 mark)

The phrase "loom large" (line 33) suggests that questions of data privacy are:

A) easily resolved B) becoming irrelevant C) significant and concerning D) hypothetical

Answer: ________

Question 15 (1 mark)

The word "erode" (line 41) is closest in meaning to:

A) strengthen B) gradually weaken C) completely destroy D) carefully preserve

Answer: ________

Section C: Summary (10 marks)

Question 16

Using your own words as far as possible, summarise the concerns and limitations associated with the use of AI in healthcare, as discussed in the passage.

Your summary must be in continuous prose and must not exceed 120 words.

Use only information from paragraphs 2 to 5 of the passage.

Section D: Application Question (10 marks)

Read the following scenario and answer Questions 17–20.

Scenario: AI in Singapore's Public Healthcare System

Singapore's Ministry of Health (MOH) has announced a nationwide initiative to integrate AI-powered diagnostic tools across all public hospitals and polyclinics by 2030. Under this plan, AI systems will be used to assist doctors in reading X-rays, CT scans, and blood test results. The government has allocated $500 million to fund the development and deployment of these systems.

Health Minister Ong Ye Kung stated that the initiative aims to reduce waiting times for diagnosis by 40% and address the growing shortage of healthcare professionals as Singapore's population ages. "AI will not replace our doctors," the Minister assured. "It will empower them to deliver faster, more accurate care."

However, the announcement has drawn mixed reactions. The Singapore Medical Association (SMA) has expressed concern that doctors may become over-reliant on AI recommendations, potentially weakening their independent clinical judgement. Patient advocacy groups have raised questions about data security, noting that the initiative will require the centralisation of patient medical records in a national digital database. Some Members of Parliament have called for stronger legislation to clarify liability in cases where AI-assisted diagnoses lead to medical errors.

Meanwhile, a recent survey by the Lee Kuan Yew School of Public Policy found that 62% of Singaporeans are comfortable with AI being used in their healthcare, provided that a human doctor retains final decision-making authority. However, only 34% said they fully understood how AI diagnostic tools work, and 48% expressed concern about the privacy of their medical data.

Question 17 (3 marks)

Referring to ideas from the passage, explain two potential benefits of Singapore's AI healthcare initiative.

(a) _______________________________________________________________

(b) _______________________________________________________________

Question 18 (3 marks)

The passage discusses concerns about AI and the doctor-patient relationship. How might Singapore's initiative affect the doctor-patient relationship? Use evidence from both the passage and the scenario to support your answer.

Answer: _______________________________________________________________

Question 19 (2 marks)

Based on the passage and the scenario, what two measures should Singapore implement to address public concerns about AI in healthcare?

(a) _______________________________________________________________

(b) _______________________________________________________________

Question 20 (2 marks)

Do you think the benefits of AI in healthcare outweigh the risks? Using ideas from the passage and the scenario, state your position and provide one reason to support it.

Answer: _______________________________________________________________

END OF PAPER

Section Marks Summary:

Section	Marks
A: Short-Answer Questions (Q1–10)	25
B: Vocabulary in Context (Q11–15)	5
C: Summary (Q16)	10
D: Application (Q17–20)	10
Total	50

Answers

TuitionGoWhere Practice Paper — General Paper H1 A-Level

Practice Paper 2 — Comprehension: Answer Key (Version 3 of 5)

Section A: Short-Answer Questions (25 marks)

Question 1 (2 marks)

Answer: Two advantages of using AI in diagnostic radiology are:

AI can detect anomalies (such as tumours, fractures, and lesions) with a level of precision that rivals or surpasses that of experienced radiologists. (1 mark)
AI augments clinicians' capabilities, allowing them to focus on complex cases while routine screenings are handled efficiently by machines. (1 mark)

Marking Notes:

Award 1 mark for each valid advantage, up to 2 marks.
Accept paraphrased versions of the above points.
Do not award marks for direct lifting of phrases such as "improving accuracy" without context.

Question 2 (2 marks)

Answer: The phrase "garbage in, garbage out" means that AI systems are only as reliable as the data they are trained on. (1 mark) If the training data is flawed, biased, or unrepresentative, the outputs produced by the AI will also be flawed or unreliable. (1 mark)

Marking Notes:

Award 1 mark for explaining the general principle (quality of input determines quality of output).
Award 1 mark for applying it to the context of AI and training data.
Students must use their own words. Direct lifting of "AI systems are only as reliable as the data on which they are trained" should not be awarded the second mark.

Question 3 (2 marks)

Answer: The word "tempered" suggests that the author's attitude towards AI diagnostics is cautious and measured. (1 mark) While the author acknowledges the promise of AI, they believe enthusiasm should be balanced with an awareness of its limitations. (1 mark)

Marking Notes:

Award 1 mark for identifying the cautious/qualified nature of the attitude.
Award 1 mark for explaining that the author does not dismiss AI but urges balance.
Accept synonyms such as "moderated," "balanced," or "qualified."

Question 4 (3 marks)

Answer: Two concerns about the reliability of AI diagnostic systems are: (a) Bias in training data: If training datasets underrepresent certain ethnic groups or age brackets, the AI may produce biased outcomes that are less accurate for those populations. (1.5 marks) (b) Inaccuracy across demographics: Research has shown that some diagnostic algorithms perform significantly less accurately for patients of African and Asian descent, raising concerns about equity in AI-driven healthcare. (1.5 marks)

Marking Notes:

Award 1.5 marks for each concern: 1 mark for identifying the concern, 0.5 marks for explaining it in the student's own words.
Both concerns are closely related but distinct: the first addresses the cause (skewed data), the second addresses the consequence (differential accuracy). Award marks for both if the student distinguishes them clearly.
If a student gives the same concern twice in different words, award marks only once.

Question 5 (2 marks)

Answer: The author provides two pieces of evidence:

Traditional drug development takes over a decade and costs upwards of $2.6 billion per drug, whereas AI can identify promising drug candidates in a fraction of the time. (1 mark)
During the COVID-19 pandemic, companies like Moderna used AI to optimise mRNA sequences in record time, accelerating vaccine development. (1 mark)

Marking Notes:

Award 1 mark for each piece of evidence, up to 2 marks.
Students may paraphrase the evidence.

Question 6 (2 marks)

Answer: In this context, "serendipitous" refers to the element of chance or unexpected discovery in scientific research. (1 mark) The author suggests that AI, which relies on algorithmic predictions, may not be able to replicate the accidental or unplanned breakthroughs that have historically driven scientific progress. (1 mark)

Marking Notes:

Award 1 mark for the definition in context.
Award 1 mark for explaining the implication (AI may narrow the scope of research by relying on predictions rather than allowing for chance discoveries).

Question 7 (2 marks)

Answer: The author means that the idea of informed consent becomes unclear and complicated when patient data is used to train AI systems. (1 mark) Patients may not fully understand or agree to how their data is being repurposed, making it difficult to say they have truly consented. (1 mark)

Marking Notes:

Award 1 mark for explaining "murky" as unclear/complicated.
Award 1 mark for applying it to the context of informed consent and data use.
Students must paraphrase; direct lifting of "patients may not fully understand how their information is being used" should not receive full credit.

Question 8 (3 marks)

Answer: The three main ethical concerns are: (a) Data privacy: AI systems require enormous volumes of patient data, and patients may not fully understand how their information is used or shared. (1 mark) (b) Accountability: It is unclear who is responsible when an AI system causes harm — the developer, the hospital, or the clinician. (1 mark) (c) Erosion of the doctor-patient relationship: Over-reliance on AI could reduce patients to data points and weaken the empathy and trust that define good medical practice. (1 mark)

Marking Notes:

Award 1 mark for each valid concern.
Accept alternative phrasings that capture the same ideas.

Question 9 (3 marks)

Answer: The author includes this quotation to reinforce the idea that medicine is fundamentally a human endeavour that requires both scientific knowledge and empathy. (1.5 marks) By quoting Gawande, the author lends authority to the argument that AI, however sophisticated, cannot replace the human qualities — empathy, trust, and clinical judgement — that define great doctors. (1.5 marks)

Marking Notes:

Award up to 1.5 marks for identifying the purpose of the quotation (to support the argument about the human dimension of medicine).
Award up to 1.5 marks for explaining how the quotation functions (appeal to authority, reinforcing the limits of AI).
A well-developed answer may also note that the author uses Gawande's credibility to acknowledge AI's utility while still emphasising its limitations.

Question 10 (4 marks)

Answer: The sentence uses a contrast structure ("getting it right… getting it wrong") to juxtapose the best-case and worst-case outcomes of AI in healthcare, emphasising the magnitude of the stakes. (1.5 marks) The phrase "The stakes could not be higher" is a hyperbole that underscores the urgency and importance of the issue. (1 mark) The language choices — "save millions of lives" versus "deepen existing inequalities" and "undermine trust in medicine" — create a stark contrast between a hopeful and a cautionary vision, compelling the reader to recognise that the consequences of failure are not merely technical but deeply human and societal. (1.5 marks)

Marking Notes:

Award 1.5 marks for identifying and explaining the contrast structure.
Award 1 mark for identifying the hyperbole and its effect.
Award 1.5 marks for analysing specific language choices and their impact.
Award a maximum of 3 marks if the student identifies structural and language features but does not explain their effect on the reader.

Section B: Vocabulary in Context (5 marks)

Question 11 (1 mark) Answer: B (enhance)

"Augment" means to increase or enhance, not to replace or diminish.

Question 12 (1 mark) Answer: B (argue)

"Contend" in this context means to assert or argue a point, not to admit or concede.

Question 13 (1 mark) Answer: C (wisely)

"Judiciously" means with good judgement or wisdom, i.e., using something sensibly and thoughtfully.

Question 14 (1 mark) Answer: C (significant and concerning)

"Loom large" means something is prominent, imposing, or a source of concern.

Question 15 (1 mark) Answer: B (gradually weaken)

"Erode" means to gradually wear away or diminish, not to destroy completely or strengthen.

Section C: Summary (10 marks)

Question 16

Model Summary (for reference — within 120 words):

AI in healthcare faces several limitations and concerns. Diagnostic AI systems are limited by the quality of their training data; if datasets are unrepresentative, biased outcomes may result, particularly for underrepresented ethnic groups. In drug development, while AI accelerates the process, critics warn it cannot replicate the chance discoveries that drive scientific breakthroughs. Ethically, the use of vast patient data raises privacy concerns, as patients may not fully consent to how their information is used. Accountability is also unclear when AI errors cause harm. Furthermore, over-reliance on AI risks eroding the empathy and human connection central to the doctor-patient relationship, qualities that no algorithm can replicate.

(Approximately 110 words)

Marking Scheme:

Criterion	Marks
Content: Identification of relevant concerns from paragraphs 2–5 (up to 7 content points)	7
Language: Use of own words, fluency, and conciseness	3
Total	10

Content Points (award 1 mark each, up to 7 marks):

AI systems are only as reliable as their training data.
Skewed/unrepresentative training data leads to biased outcomes.
Algorithms may be less accurate for certain ethnic groups (e.g., African and Asian descent).
AI cannot replicate serendipitous/chance scientific discoveries.
Data privacy concerns arise from the use of large volumes of patient data.
Informed consent is unclear when data is repurposed for AI.
Accountability is unresolved when AI errors cause harm.
Over-reliance on AI may erode the doctor-patient relationship/empathy.

Language Marks (award as follows):

3 marks: Excellent paraphrasing, fluent continuous prose, well within word limit.
2 marks: Good paraphrasing with minor lifting, generally fluent.
1 mark: Significant lifting from the passage, limited paraphrasing.
0 marks: Wholly lifted from the passage or exceeds word limit significantly.

Common Mistakes:

Including information from paragraph 1 (introduction) or paragraph 6 (conclusion) — these are outside the specified paragraphs.
Exceeding the 120-word limit.
Failing to use own words (direct lifting).

Section D: Application Question (10 marks)

Question 17 (3 marks)

Answer: Two potential benefits of Singapore's AI healthcare initiative are: (a) Reduced waiting times: The passage notes that AI can handle routine screenings efficiently, and the scenario states the initiative aims to reduce diagnosis waiting times by 40%. (1.5 marks) (b) Addressing healthcare workforce shortages: The passage explains that AI augments clinicians' capabilities, and the scenario highlights that the initiative will help address the shortage of healthcare professionals as Singapore's population ages. (1.5 marks)

Marking Notes:

Award 1.5 marks for each benefit: 1 mark for identifying the benefit, 0.5 marks for linking it to ideas from the passage.
Accept other valid benefits drawn from the passage (e.g., improved diagnostic accuracy, cost savings).

Question 18 (3 marks)

Answer: Singapore's initiative could affect the doctor-patient relationship in two ways. Positively, the passage notes that AI can free physicians from administrative burdens, allowing them to devote more time to meaningful patient interaction. (1 mark) The scenario echoes this, with the Health Minister assuring that AI will "empower" doctors to deliver faster, more accurately care. (0.5 marks) However, the passage warns that over-reliance on AI could erode empathy and reduce patients to data points, (1 mark) a concern echoed by the Singapore Medical Association in the scenario, which fears doctors may become over-reliant on AI and lose independent clinical judgement. (0.5 marks)

Marking Notes:

Award up to 1.5 marks for each side of the argument (positive and negative effects).
Award 0.5 marks for each reference to the passage or scenario.
Answers must draw on both the passage and the scenario for full marks.

Question 19 (2 marks)

Answer: Two measures Singapore should implement are: (a) Stronger data protection legislation to address public concerns about the privacy and security of medical data in the national digital database. (1 mark) (b) Clear legal frameworks on liability to determine who is responsible when AI-assisted diagnoses lead to errors, as called for by MPs in the scenario and discussed in the passage. (1 mark)

Marking Notes:

Award 1 mark for each valid measure linked to concerns raised in the passage or scenario.
Accept other reasonable measures (e.g., public education on how AI works, ensuring diverse training data, mandatory human oversight of AI diagnoses).

Question 20 (2 marks)

Answer: (Accept any well-reasoned position. Sample answer below.)

Sample Position (Benefits outweigh risks): The benefits of AI in healthcare outweigh the risks because, as the passage notes, AI has the potential to save millions of lives through faster, more accurate diagnosis and accelerated drug development. (1 mark) The scenario shows that Singapore's initiative could reduce waiting times by 40% and address critical workforce shortages, which are pressing needs in an ageing society. Provided that robust regulations and human oversight are in place, the risks can be effectively managed. (1 mark)

Sample Position (Risks outweigh benefits): The risks of AI in healthcare may outweigh the benefits because, as the passage highlights, biased algorithms could deepen existing inequalities, and the erosion of the doctor-patient relationship could fundamentally undermine the quality of care. (1 mark) The scenario reveals that only 34% of Singaporeans understand how AI works, and nearly half are concerned about data privacy, suggesting that public trust has not yet been earned. Without addressing these concerns, the initiative risks doing more harm than good. (1 mark)

Marking Notes:

Award 1 mark for a clear position.
Award 1 mark for a reason supported by evidence from the passage or scenario.
Accept either position (for or against) as long as it is well-supported.

END OF ANSWER KEY

Marks Summary:

Section	Marks
A: Q1–10	25
B: Q11–15	5
C: Q16	10
D: Q17–20	10
Total	50