Đề luyện thi tốt nghiệp THPT môn Tiếng Anh nâng cao số 23
Đề luyện thi tốt nghiệp THPT môn Tiếng Anh số 23, gồm 40 câu trắc nghiệm nâng cao về từ vựng, ngữ pháp, liên kết văn bản và đọc hiểu theo định hướng đánh giá năng lực.
Câu 1.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (1).
A. conceptually superficial
B. ethically consequential
C. administratively effortless
D. technically routine
Câu 2.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (2).
A. initiatively designed
B. worker-informed
C. initiating
D. initiative-led
Câu 3.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (3).
A. brief
B. examinesed
C. examines
D. examinesing
Câu 4.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (4).
A. with
B. to
C. for
D. on
Câu 5.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (5).
A. take deeper into
B. make an evidence
C. challenge opaque decisions
D. advance the deeply
Câu 6.
It is tempting to reduce evaluating workplace algorithms to checking whether a scheduling app runs quickly. That assumption overlooks how (1) ______ the work becomes when evidence conflicts. The new (2) ______ initiative ‘Rules Behind the Screen’ begins with a session that (3) ______ the wider consequences of each decision. Tasks are matched (4) ______ learners’ existing skills. Advanced participants may (5) ______; beginners, who are unfamiliar with (6) ______, first practise in a controlled setting.
Choose the option that best fits blank (6).
A. to documenting automated prompts
B. documenting automated prompts
C. have documenting automated prompts
D. document automated prompts
Câu 7.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (7).
A. has been reshapeing
B. reshape
C. reshaped
D. reshapeing
Câu 8.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (8).
A. oversee to
B. look over
C. look after
D. overlook
Câu 9.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (9).
A. Instead of
B. In view of
C. On behalf of
D. As far as
Câu 10.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (10).
A. number
B. quality
C. volume
D. standard
Câu 11.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (11).
A. responsibly
B. randomly
C. altogether
D. temporarily
Câu 12.
Practical Change, Not Symbolic Change
When organisations persist in assigning shifts and ratings through unreviewed models, they can (7) ______ long-term harm. At the same time, decision-makers may (8) ______ early warnings because the damage is dispersed. Select human review procedures where evidence supports it. (9) ______ adopting a fashionable stand-alone measure, compare several interventions. The aim is to cut the (10) ______ of preventable harm. Document performance scores, location data, and customer feedback (11) ______ and manage them (12) ______.
Choose the option that best fits blank (12).
A. responsibility
B. responsibly
C. responsive
D. responsible
Câu 13.
a. Linh: I thought algorithmic management was mainly about checking whether a scheduling app runs quickly.
b. Minh: The new evidence suggests the issue is far more ethically consequential.
c. Linh: Then I should read the briefing before making that assumption again.
Choose the best arrangement to make a meaningful exchange or paragraph.
A. b – c – a
B. a – b – c
C. a – c – b
D. c – b – a
Câu 14.
a. Researcher: We need to decide whether the pilot on algorithmic management should expand.
b. Analyst: What outcome matters most to the affected groups?
c. Researcher: Not just speed; they may coordinate complex operations and make some procedures more consistent when objectives and data are appropriate.
d. Analyst: Then the evaluation must include distributional effects, not only averages.
e. Researcher: Agreed. We will revise the indicators before the next phase.
Choose the best arrangement to make a meaningful exchange or paragraph.
A. a – b – c – d – e
B. a – c – d – e – b
C. b – c – d – e – a
D. e – d – c – b – a
Câu 15.
a. First, our team reviewed the assumption that checking whether a scheduling app runs quickly.
b. We then compared that assumption with observations from several settings.
c. The comparison revealed an important limitation: a model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
d. Consequently, we redesigned the study to include contextual evidence.
e. The revised design produced conclusions that were narrower but more defensible.
Choose the best arrangement to make a meaningful exchange or paragraph.
A. b – c – d – e – a
B. a – b – c – d – e
C. e – d – c – b – a
D. a – c – d – e – b
Câu 16.
a. My interest in algorithmic management began with a local question rather than a grand theory.
b. During university, I learned that the issue involved workers, employers, software vendors, unions, and labour regulators.
c. I therefore combined technical training with community-based fieldwork.
d. That combination eventually led me to coordinate the ‘Rules Behind the Screen’ programme.
e. I now value the role because it joins analysis with practical accountability.
Choose the best arrangement to make a meaningful exchange or paragraph.
A. b – c – d – e – a
B. e – d – c – b – a
C. a – c – d – e – b
D. a – b – c – d – e
Câu 17.
a. The town first treated algorithmic management as a narrow technical issue.
b. A series of disruptions then exposed links to public services and local livelihoods.
c. Officials responded by creating a cross-sector working group.
d. The group tested meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use.
e. Although results improved, the programme retained monitoring because a model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Choose the best arrangement to make a meaningful exchange or paragraph.
A. b – c – d – e – a
B. e – d – c – b – a
C. a – c – d – e – b
D. a – b – c – d – e
Câu 18.
A public discussion about algorithmic management often begins with a simplified claim. (18) ______. The project did not accept that framing without investigation. (19) ______. Once responsibilities had been agreed, the team moved from consultation to analysis. (20) ______. This prevented a single dramatic case from dominating the conclusions. (21) ______. The aim was not to promise a universal solution but to identify defensible actions. (22) ______.
Choose the sentence that best fits blank (18).
A. By contrast, the project abandoned all measurements and relied entirely on intuition.
B. That simplification is risky: a model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
C. The final recommendations followed a broader principle: meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use.
D. In response, the ‘Rules Behind the Screen’ team began by consulting workers, employers, software vendors, unions, and labour regulators.
Câu 19.
A public discussion about algorithmic management often begins with a simplified claim. (18) ______. The project did not accept that framing without investigation. (19) ______. Once responsibilities had been agreed, the team moved from consultation to analysis. (20) ______. This prevented a single dramatic case from dominating the conclusions. (21) ______. The aim was not to promise a universal solution but to identify defensible actions. (22) ______.
Choose the sentence that best fits blank (19).
A. The next stage examined the mechanism: these systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match.
B. By contrast, the project abandoned all measurements and relied entirely on intuition.
C. In response, the ‘Rules Behind the Screen’ team began by consulting workers, employers, software vendors, unions, and labour regulators.
D. Evidence was then compared across settings rather than treated as universally identical.
Câu 20.
A public discussion about algorithmic management often begins with a simplified claim. (18) ______. The project did not accept that framing without investigation. (19) ______. Once responsibilities had been agreed, the team moved from consultation to analysis. (20) ______. This prevented a single dramatic case from dominating the conclusions. (21) ______. The aim was not to promise a universal solution but to identify defensible actions. (22) ______.
Choose the sentence that best fits blank (20).
A. Evidence was then compared across settings rather than treated as universally identical.
B. By contrast, the project abandoned all measurements and relied entirely on intuition.
C. In response, the ‘Rules Behind the Screen’ team began by consulting workers, employers, software vendors, unions, and labour regulators.
D. The next stage examined the mechanism: these systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match.
Câu 21.
A public discussion about algorithmic management often begins with a simplified claim. (18) ______. The project did not accept that framing without investigation. (19) ______. Once responsibilities had been agreed, the team moved from consultation to analysis. (20) ______. This prevented a single dramatic case from dominating the conclusions. (21) ______. The aim was not to promise a universal solution but to identify defensible actions. (22) ______.
Choose the sentence that best fits blank (21).
A. Evidence was then compared across settings rather than treated as universally identical.
B. That simplification is risky: a model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
C. By contrast, the project abandoned all measurements and relied entirely on intuition.
D. The final recommendations followed a broader principle: meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use.
Câu 22.
A public discussion about algorithmic management often begins with a simplified claim. (18) ______. The project did not accept that framing without investigation. (19) ______. Once responsibilities had been agreed, the team moved from consultation to analysis. (20) ______. This prevented a single dramatic case from dominating the conclusions. (21) ______. The aim was not to promise a universal solution but to identify defensible actions. (22) ______.
Choose the sentence that best fits blank (22).
A. The next stage examined the mechanism: these systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match.
B. By contrast, the project abandoned all measurements and relied entirely on intuition.
C. The final recommendations followed a broader principle: meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use.
D. In response, the ‘Rules Behind the Screen’ team began by consulting workers, employers, software vendors, unions, and labour regulators.
Câu 23.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
Which of the following is NOT presented as a central feature of the issue?
A. Cooperation among multiple stakeholders
B. A mechanism that produces varied outcomes
C. The need to interpret evidence in context
D. A universal solution that works identically in every setting
Câu 24.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
The word ‘overlooks’ in paragraph 1 is OPPOSITE in meaning to ______.
A. misses
B. recognises
C. neglects
D. ignores
Câu 25.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
The phrase ‘These processes’ in paragraph 2 refers to ______.
A. checking whether a scheduling app runs quickly
B. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use.
C. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match.
D. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes.
Câu 26.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
The word ‘mechanical’ in paragraph 3 could best be replaced by ______.
A. carefully calibrated
B. physically engineered
C. automatic and uncritical
D. socially negotiated
Câu 27.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
Which option best paraphrases the final sentence of paragraph 3?
A. Evidence is too uncertain to support any practical response.
B. The evidence should guide action, but its limits and context must still be considered.
C. Once evidence exists, the same action should be taken everywhere.
D. Context matters only when no scientific observations are available.
Câu 28.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
Which of the following is TRUE according to the passage?
A. The passage argues that technical monitoring is unnecessary.
B. The issue can be solved by changing individual behaviour alone.
C. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes.
D. The passage treats all affected groups as experiencing identical conditions.
Câu 29.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
In which paragraph does the writer explain a causal mechanism?
A. Paragraph 2
B. Paragraph 1
C. Paragraph 4
D. Paragraph 3
Câu 30.
Paragraph 1. Algorithmic management uses data-driven systems to allocate tasks, monitor performance, set schedules, or evaluate workers. Public debate, however, often reduces the issue to checking whether a scheduling app runs quickly. That narrow view overlooks the technical, social, and institutional relationships that determine real outcomes.
Paragraph 2. These systems translate managerial priorities into rules and scores, often operating continuously and at a scale that human supervisors could not match. These processes explain why isolated observations can be misleading and why changes may differ among places or groups.
Paragraph 3. Case studies in logistics, platform work, and healthcare show that the same tool can improve coordination while also reducing worker autonomy. The evidence therefore supports careful action, but not mechanical conclusions. A model can be statistically accurate overall yet still treat particular groups unfairly or misunderstand exceptional circumstances.
Paragraph 4. Meaningful consultation, explainable criteria, data minimisation, and accessible appeal procedures are central to responsible use. Researchers increasingly study not only model performance but also how workplace power, incentives, and collective voice shape outcomes. In this sense, progress depends on both technical capacity and accountable cooperation among workers, employers, software vendors, unions, and labour regulators.
In which paragraph does the writer discuss future-oriented methods and cooperation?
A. Paragraph 1
B. Paragraph 4
C. Paragraph 2
D. Paragraph 3
Câu 31.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
Where in paragraph 1 does the following sentence best fit?
‘A broader evaluation must therefore examine distribution, governance, and long-term capacity.’
A. [II]
B. [IV]
C. [I]
D. [III]
Câu 32.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
The phrase ‘privilege visible outputs’ in paragraph 2 could best be replaced by ______.
A. remove every measurable result
B. hide outcomes from public review
C. guarantee that benefits will last
D. give priority to results that are easy to display
Câu 33.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
The word ‘These’ in paragraph 2 refers to ______.
A. short funding cycles, incomplete data, fragmented duties, and pressure for quick results
B. durable outcomes
C. complex systems
D. affected people
Câu 34.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
According to paragraph 2, which of the following is NOT a force behind narrow decisions?
A. Short funding cycles
B. Fragmented responsibilities
C. Pressure for rapid results
D. Long-term institutional learning
Câu 35.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
Which option best summarises paragraph 3?
A. Maintenance capacity is unrelated to policy effectiveness.
B. A seemingly successful measure can redistribute harm or create new weaknesses.
C. Average improvement proves that no one is disadvantaged.
D. Technical measures always benefit vulnerable groups first.
Câu 36.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
The word ‘credible’ in paragraph 4 is OPPOSITE in meaning to ______.
A. reliable
B. unconvincing
C. trustworthy
D. defensible
Câu 37.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
Which of the following is TRUE according to the passage?
A. Monitoring unintended effects helps programmes adapt when circumstances change.
B. Rapid announcements usually produce durable outcomes.
C. Consultation should occur only after a programme has ended.
D. A single indicator is sufficient in complex systems.
Câu 38.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
Which option best paraphrases the final sentence of the passage?
A. In complicated situations, the ability to learn may matter as much as confidence in the first design.
B. Learning is useful only when an initial plan completely fails.
C. Initial plans should never be revised once implementation begins.
D. Complex systems become simple when enough indicators are collected.
Câu 39.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
What can be inferred from the passage?
A. The slowest programme is always the most effective.
B. Programmes that look efficient in the short term may be fragile if they ignore governance and maintenance.
C. Affected groups necessarily agree on every policy choice.
D. Quantitative indicators should be eliminated from evaluation.
Câu 40.
[I] Efforts to address algorithmic management have expanded as its links to wider social, environmental, and institutional outcomes have become clearer. [II] Yet expansion has also exposed a recurring mistake: success is often measured through a single convenient indicator. [III] For that reason, an apparently efficient intervention may fail when it ignores who bears the costs and who can influence decisions. [IV]
Paragraph 2. Several forces encourage narrow decision-making. Agencies may face short funding cycles, incomplete data, fragmented responsibilities, and pressure to announce rapid results. These constraints can privilege visible outputs over durable outcomes.
Paragraph 3. The consequences are not merely technical. A measure may improve an average while worsening conditions for a vulnerable group; it may solve one stage of a problem while transferring costs to another; or it may create dependence on systems that local institutions cannot maintain.
Paragraph 4. A more credible approach begins by defining the problem with affected people, testing assumptions, comparing alternatives, and monitoring unintended effects. Although this process is slower, it makes adaptation possible when conditions change. In complex systems, the quality of learning can be as important as the apparent certainty of an initial plan.
Which option best summarises the passage?
A. Technical expertise alone can resolve algorithmic management without attention to distribution or institutions.
B. Because algorithmic management is complex, governments should postpone action until complete data are available.
C. Addressing algorithmic management requires broader evaluation, participation, comparison of alternatives, and continuous learning rather than reliance on one convenient indicator.
D. The best way to address algorithmic management is to select the fastest measurable intervention and avoid later changes.
