Skills practice questions.

A campaign group's infographic shows quarterly figures for the share of removed posts that were restored on appeal. The bars appear to grow enormously across the year, but the vertical axis begins at 60 per cent and ends at 70 per cent. What does this tell a reader about the underlying change?

• The change is exactly as large as the picture suggests, since the bars are to scale
• The figures cannot be compared because each quarter uses a different definition
• The change is smaller than the picture suggests, because the axis magnifies it
• The change must be an error, because a restored share cannot rise across a year

Answer: The change is smaller than the picture suggests, because the axis magnifies it

A truncated axis is a presentation choice rather than a falsification, so the numbers can be correct while the impression they leave is wrong. Reading the axis before reading the shape is a marked skill in source-based questions.

A source reports that pupils who spend more than four hours a day on their phones score lower in a reading test taken on the same day. A candidate writes that the source shows phone use causes weaker reading. Which objection is the strongest?

• A study of this kind should report its sample size before publishing any figure
• Reading tests measure comprehension, which is only one of the literacies that matter
• The source does not name the model of phone that the pupils were using
• Both were measured at once, so weaker readers may simply use phones more

Answer: Both were measured at once, so weaker readers may simply use phones more

Data gathered at a single moment fixes no order of events, so reverse causation and a common third factor both stay open. The examinable phrasing is that the source shows an association, with the causal reading named as an interpretation.

Source D, a national statistics bulletin, reports that 88 per cent of households have an internet connection. Source E, a charity's survey, reports that 71 per cent of adults are meaningfully connected. A candidate concludes that one of the two sources must be wrong. What is the better explanation?

• The two figures count different units and use different definitions
• The charity's smaller budget makes its fieldwork less accurate than official work
• Official statistics are always collected later than charity surveys are
• One of the two must have made an arithmetic error in the published total

Answer: The two figures count different units and use different definitions

One counts households holding a connection, the other counts adults meeting a quality threshold, so both can be correct at the same time. Reconciling the definitions before comparing the figures is the first move in every source-based question.

A researcher finds the same national broadband coverage figure in a government report and in a report by a consumer campaign group that opposes the government's policy. What does the agreement most reasonably support?

• That neither publisher has an interest in how the figure is read
• That the figure is more dependable, since sources with opposing aims agree
• That coverage figures are unaffected by who chooses to publish them
• That the government's policy has succeeded in extending coverage to rural districts

Answer: That the figure is more dependable, since sources with opposing aims agree

Corroboration by a source with a conflicting interest is stronger than repetition by an allied one, although it says nothing about whether the policy worked. The check worth making next is whether both figures came from the same original measurement.

Two sources report daily smartphone use for the same country. Source F asks respondents to estimate their own daily use; Source G reads the total from a screen-time counter on the participants' phones. Which difference should a candidate expect?

• The measured figure is likely to be lower, because counters miss the time spent on calls
• The two should agree, because both describe the same behaviour in the same population
• Self-reports are likely to differ, because people estimate their own use poorly
• The measured figure is unusable, because it was collected without asking the participants

Answer: Self-reports are likely to differ, because people estimate their own use poorly

Instrumented measurement and recall capture different things: one records device activity, the other records what a person believes and is willing to state. Naming the method that produced a figure is part of saying what the figure can support.

Part (a) of a Paper 1 question asks candidates to describe two ways in which employers use algorithmic scheduling software, for 6 marks. A candidate lists six ways in six short sentences, names no employer or system, and receives a low mark. The most likely reason is that

• the answer was written in continuous prose instead of numbered bullet points
• employers rather than shift workers were treated as the affected stakeholder group
• identification was offered where description with a named case was required
• algorithmic scheduling sits outside the content topics that Paper 1 may examine

Answer: identification was offered where description with a named case was required

A 6-mark part (a) expects two points each carrying up to three marks: the claim, its development and a named real deployment. Six bare identifications cannot reach the development or example marks, so breadth actively costs the candidate the marks that depth would have earned.