People per UK Household in 1981 and 2001

All category percentages are reported correctly, and the response exceeds 150 words. However, one-person households ranked fourth in 1981, not second, and the decline comparison overlooks the four-percentage-point fall in six-person households.

Progression is logical, moving from overview to rising and then falling categories, with cohesive devices such as 'after which' and 'a similar picture' used naturally.

Vocabulary provides useful variety for reporting trends, including ranking first and underwent a mild rise. However, family classifications is less accurate than household categories, least popular implies preference rather than frequency, and changes in shares need percentage-point units.

Sentence structures are varied and generally controlled, including complex relative and comparative constructions, with occasional slips such as '2-person household made up' and 'in 1-person household'.

The overview correctly identifies opposing trends across smaller and larger households, and most category values are accurate. However, the 2001 two-person share of 34% is omitted, “most households” overstates a 31–34% category, and changes require percentage-point units.

Information is logically grouped by year and progresses coherently with good use of referencing and linking devices, though transitions like 'Regarding the first stacked bar' and 'finishing at' feel somewhat mechanical at times.

The response uses varied household vocabulary, including “occupants” and household-size descriptions, but expressions such as “enjoyed upswings”, “other demographics” and “heading the list of total demographics” are imprecise or unnecessarily elaborate.

Sentence structures are varied and mostly accurate, including participle clauses and inversion ('Remaining in the last place were...'), with only occasional awkwardness such as 'rising up by 9%'.

The overview and most category values are accurate, with useful grouping by direction of change. However, the two-person rise is 3 percentage points, and the six-person decline is larger, not smaller, than the declines in three- and four-person households.

Information is organised logically with a clear overview, grouping by upward and downward trends, and cohesion is managed smoothly with only minor mechanical issues such as misplaced spaces before commas.

Vocabulary is varied and generally effective, with expressions such as “breakdown”, “occupants” and “dipping”. Some descriptions, including “witnessed the same level of decrease”, are wordy, and “at lower rates” inaccurately characterises one of the declines.

A wide range of structures is used with generally strong control, including contrastive clauses such as “While the figure for the former had declined...”. Several spaces before commas need correction, and more direct subjects would make some long sentences easier to follow.

The response identifies the principal trends and rankings accurately and reports the 1981 distribution clearly. Changes in shares need percentage-point units, and giving the separate 2001 values for the declining categories would make the final comparisons more complete.

Progression is clear and cohesive, moving from an overview to detailed descriptions with well-managed linking devices, though the year-by-year organisation makes some comparisons less direct.

Vocabulary is flexible and precise for describing proportions and trends, with natural collocations such as 'most prevalent category' and 'less pronounced', and only minimal imprecision.

A wide range of structures, including complex sentences and varied passive and participle forms, is used with strong accuracy and only rare, non-impeding slips.