Advanced reading practice test 3

Reading Passage 1

The Promise and Limits of Personalised Medicine

AMedical treatment has traditionally relied on guidelines developed from evidence about large groups of patients. These guidelines remain essential, but clinicians have long recognised that people with the same diagnosis can respond differently to the same treatment. Differences in age, other conditions, lifestyle and biology can affect both benefits and side effects. Personalised medicine aims to use information about an individual to make prevention, diagnosis or treatment more appropriate for that person.

BGenetic testing has helped make this approach more practical in some areas. Variations in particular genes can influence how quickly a person processes a medicine or whether a drug is likely to work. In certain cases, a test can help a clinician choose a dose or select an alternative treatment. However, genes rarely explain every relevant difference. Environment, diet, other medicines and the course of an illness may also influence the result, so genetic information should be interpreted alongside other evidence.

CThe amount of data involved can be substantial. Researchers may combine genetic information with medical histories, laboratory results, imaging and information collected by wearable devices. Such records can reveal patterns that would be difficult to detect from a single source. Yet combining data creates technical challenges: records may use different formats, measurements may not be comparable, and missing information may be concentrated in particular groups. A sophisticated analysis cannot automatically correct for weaknesses in the underlying data.

DPersonalised approaches also raise questions about fairness. If a test is developed mainly using data from one population, it may be less reliable for people whose backgrounds were poorly represented. Access can be unequal too: advanced testing may be available in specialist hospitals but difficult to obtain in rural areas or underfunded health systems. Without deliberate planning, a technology intended to improve individual care could widen existing differences in health outcomes.

EPrivacy is another concern because genetic and health data can reveal information about relatives as well as the person tested. Strong safeguards are needed to limit access, explain how records will be used and prevent information from being reused in unexpected ways. Consent is not simply a form to sign once; people should understand important changes in how their data are stored or shared. At the same time, overly restrictive data policies can make useful research harder, so governance must balance protection with responsible scientific use.

FEvidence for a personalised intervention must still be tested carefully. A treatment that works well for a subgroup in an early study may not produce the same benefit in routine practice. Researchers need to compare outcomes, consider side effects and determine whether the added information changes decisions in ways that matter to patients. A test may predict a biological response accurately without improving survival, comfort or quality of life.

GCosts also shape the value of these technologies. A test may be scientifically interesting but offer little practical benefit if it is expensive, difficult to interpret or unlikely to change treatment. Health systems therefore need to compare its costs with the outcomes it produces and consider whether resources might achieve greater benefit elsewhere. Economic evaluation does not replace clinical judgement; it helps decision-makers understand the trade-offs involved.

HPersonalised medicine is best viewed as an extension of evidence-based care rather than a replacement for it. Individual information can improve decisions when it is reliable, relevant and linked to effective options. Its value depends on careful evaluation, fair access, secure data practices and clear communication with patients. The aim is not to predict every detail of a person's future health, but to use the available evidence to make better decisions under uncertainty.

Reading Passage 2

When Historic Buildings Become Part of Modern Cities

AHistoric buildings often occupy valuable sites in city centres, where land is under pressure from housing, offices, transport and public services. Their preservation can be presented as a choice between protecting the past and meeting present needs. In practice, the decision is rarely so simple. Some buildings have architectural or cultural significance, while others may have been altered so extensively that their value is less clear. The condition of the structure, its possible uses and the needs of the surrounding community all matter.

BOne option is adaptive reuse: changing a building's function while retaining important parts of its structure. A former warehouse might become a library, a school might be converted into housing, or an old industrial site might accommodate offices and workshops. Reuse can preserve local character and reduce the need for some new construction materials. However, it may require expensive repairs, upgrades to electrical systems and improvements to fire safety or accessibility.

CEnergy performance presents a particular challenge. Older buildings may have solid walls, single-glazed windows or layouts that make modern heating and cooling systems difficult to install. Insulation can reduce energy use, but certain interventions may damage historic surfaces or trap moisture inside walls. The most suitable solution depends on the building's materials, climate and use. A standard package of improvements may be effective in one structure and harmful in another.

DPreservation can also affect who is able to live or work in an area. Restored buildings may attract visitors and investment, supporting local businesses. Yet improvements can raise rents and property values, placing pressure on long-term residents and small organisations. A project that succeeds architecturally may therefore have mixed social results. Planning decisions need to consider not only the building itself but also how benefits and costs are distributed across the neighbourhood.

EThe evidence used to judge reuse projects is often incomplete. Construction costs are visible at the beginning, while savings in energy, avoided demolition and community value may emerge over many years. Comparisons are difficult because buildings differ in size, condition and intended function. A fair assessment should examine the full life cycle, including maintenance and future adaptation, rather than comparing the initial cost of reuse with the initial cost of a new building alone.

FPublic participation can improve decisions, but consultation needs to happen early enough to influence the options. Residents may know how a building has been used, which spaces are important and what facilities the area lacks. Their views do not remove technical constraints, and not every preference can be accommodated. Still, early discussion can reveal practical uses that designers or developers had not considered.

GThe idea of authenticity is also contested. Some argue that a historic building should retain visible evidence of its age, while others favour restoring it to an earlier appearance. Both approaches can erase part of its history if applied too rigidly. A building may have changed repeatedly as people adapted it to new needs. Conservation decisions therefore involve choices about which periods and features to preserve, rather than simply returning a structure to a single original state.

HHistoric preservation is most successful when it connects cultural value with practical use. A building that remains empty may deteriorate, while a heavily altered structure may lose the features that made it significant. The challenge is to find a viable function, make necessary improvements with care and consider the effects on the surrounding community. Preservation is not an automatic rejection of development; it is a way of asking what should be retained, why it matters and how it can continue to serve people.

Reading Passage 3

Why Forecasts Fail—and How They Can Improve

AForecasts influence decisions in fields ranging from public health and energy planning to business and transport. They help people prepare for possible futures, but they are often judged as though they were promises. In reality, a forecast is a conditional estimate based on available information, a chosen method and assumptions about how relevant factors may change. Its usefulness depends not only on whether the final number is close to what happens, but also on whether the forecast was appropriate for the decision it was meant to support.

BOne source of error is the assumption that patterns from the past will continue. Historical data can reveal useful regularities, but systems change. A new technology may alter consumer behaviour, a regulation may change incentives, or an unexpected event may disrupt supply. A model trained on stable conditions may perform poorly when the environment shifts. This is sometimes described as a change in the relationship between the data and the process being forecast.

CForecasts can also fail because important variables are missing or poorly measured. A model of transport demand might include income, fuel prices and population but overlook changes in remote working. If a relevant factor is absent, the model may systematically overestimate or underestimate demand. Adding more variables is not always the answer: some data may be noisy, duplicated or available only after the decision must be made. The task is to identify which information genuinely improves the forecast.

DUncertainty has several forms. Some arises from imperfect measurement, some from the limits of a model, and some from events that cannot be predicted in detail. A single point estimate can conceal these differences. Ranges, scenarios and probability statements can communicate uncertainty more honestly, provided they are explained clearly. A wide range may appear less decisive, but it can be more useful for planning than a precise number that gives a false impression of certainty.

EPeople who use forecasts can introduce further problems. Decision-makers may focus on the scenario they prefer, ignore warnings that conflict with their plans or treat a central estimate as the most likely outcome in every circumstance. Forecasters may also face pressure to provide a simple answer when the evidence does not support one. Clear communication should therefore explain the assumptions behind a result, the factors most likely to change it and the consequences if the estimate proves wrong.

FEvaluation requires more than checking one forecast after the event. A model may appear accurate because conditions happened to be favourable, or inaccurate because an unusual event occurred despite a reasonable method. Forecasters should compare predictions with outcomes across many cases, examine the size and direction of errors, and consider whether the method adds value over a simpler baseline. They should also avoid changing a model after every surprising result without first understanding the cause.

GForecasts can be improved through regular review. Researchers can track whether errors grow in particular situations, update data when conditions change and test alternative assumptions. In some settings, combining several models produces more stable results because their weaknesses differ. However, an average of several poor models does not automatically become reliable. The quality of the inputs and the suitability of the methods still matter.

HThe purpose of forecasting is not to eliminate uncertainty but to support decisions despite it. A useful forecast makes its limits visible, gives users a sense of the range of plausible outcomes and is evaluated against the purpose for which it was created. When conditions change, forecasts should be revised rather than defended as permanent truths. Their greatest value may lie not in predicting the future perfectly, but in helping people prepare for more than one possible future.