PTE Describe Image Tables: Vocabulary for Highs, Lows, and Exceptions

Tables create a selection problem. Rows and columns make every cell look equally available, but a PTE Academic Describe Image response lasts only 40 seconds. Your job is to find the table's structure: a maximum, a minimum, a cluster, a contrast, or a genuine exception.

Pearson currently allows 25 seconds to study the image. Official advice says to focus on main information and supporting details. This page teaches vocabulary that turns a grid into that hierarchy: peaks at, records the lowest figure, clusters around, is broadly consistent, stands apart, and with the exception of.

Read title, units, rows, and columns

Consider this original practice table of average weekly study hours:

The unit is hours per week, and the values are averages. Rows are faculties; columns are study modes. That permits two different scans:

• Across a row: how one faculty distributes time.
• Down a column: how faculties compare on one mode.

Trying to do both completely will produce a cell-by-cell reading. Choose the direction that reveals the clearest pattern.

Maximum and minimum language

Peak at identifies the highest value in a defined series:

Independent study peaks at 11.6 hours among engineering students.

The scope matters. It is the peak for independent study across the four faculties, not necessarily the highest study time ever recorded.

For a category rather than a changing series, records the highest figure is often plainer:

Engineering records the highest independent-study figure.

For the minimum:

Business has the lowest independent-study average, at 8.7 hours.

Do not say a figure bottomed out unless the image presents movement over an ordered sequence. A categorical table has a lowest value, but it may have no trend.

Cluster around and remain within a narrow range

Tutorial hours run from 1.8 to 2.1. The interesting fact is not which faculty wins by 0.1; it is the tight grouping.

Tutorial time clusters around two hours per week.

Tutorial averages remain within a narrow range of 1.8 to 2.1 hours.

Cluster around means several observations gather near a central level. It does not mean they are equal. State the approximate center or range so the listener can evaluate the description.

Broadly consistent is also useful:

Tutorial time is broadly consistent across faculties.

It describes similarity, not perfect uniformity. Avoid identical when the cells differ.

Exception language

Business group-study time is 6.4 hours, while the other faculties fall between 3.8 and 4.1. You can frame that relationship as:

Group-study time is close to four hours in every faculty except Business, where it reaches 6.4 hours.

Or:

With the exception of Business, group-study averages are tightly grouped around four hours.

With the exception of identifies a category that does not follow a stated pattern. Always name the pattern and show how the exception differs. Merely saying Business is an exception leaves the relationship unspecified.

Is it really an outlier?

An outlier is an observation unusually distant from the wider distribution, often under a defined statistical rule. A visibly separated cell may be described cautiously as standing apart:

The Business figure for group study stands apart from the other three values.

Calling it a statistical outlier would go beyond this tiny descriptive table unless a rule, distribution, or annotation supports that label. The safest PTE language often describes the visible distance without claiming a statistical classification.

Similarly, the highest value is not automatically an outlier. Engineering's 11.6 hours is near Science's 10.8 and belongs to a gradual spread.

Respect averages and units

The table reports average hours. It does not say every engineering student studies independently for 11.6 hours. Keep the measure visible:

Engineering has the highest average.

The figure is 11.6 hours per week.

Avoid converting decimals into hours and minutes unless necessary. Eleven point six hours is 11 hours 36 minutes, but that calculation adds no useful main feature and creates another chance for error.

If the cells use different units by column, compare within a unit. It would be meaningless to call 40 percent higher than 12 kilograms merely because 40 is numerically larger.

Select rows instead of reciting cells

A weak response says:

Arts is 9.2, 3.8, 1.9; Engineering is 11.6, 4.1, 2.0; Science is 10.8, 3.9, 2.1; and Business is 8.7, 6.4, 1.8.

A selective response says:

The table compares average weekly hours spent on three forms of study across four faculties. Independent study takes the most time in every faculty and peaks at 11.6 hours in Engineering, while Business records the lowest figure, at 8.7. Tutorial time is broadly consistent at about two hours. The main exception is group study in Business, which reaches 6.4 hours compared with roughly four elsewhere.

The rewrite captures a column-wide dominance, two extrema, a cluster, and one exception. It uses only three exact figures because those numbers support the hierarchy.

A rapid table scan

Write a preparation note like:

avg hrs/week | independent highest all | max Eng 11.6 min Bus 8.7 | tutorials ~2 | exception Bus group 6.4 vs ~4

Use this scan order:

1. Read the title and unit.
2. Look for a repeated row or column pattern.
3. Circle the overall maximum and minimum only if they clarify that pattern.
4. Box one tight cluster or exception.
5. Select two to four numbers as evidence.

The order is flexible. If the table highlights totals, begin with those. If conditional formatting marks a relationship, verify it against the legend before describing it.

Common mistakes

• Reading every cell. A complete inventory can still be a poor description.
• Comparing unlike units. Scan headers before ranking values.
• Dropping average. An average does not describe every member of a group.
• Calling the maximum an outlier. Describe its distance and context first.
• Using time-trend verbs on unordered rows. Faculties do not rise from Arts to Engineering merely because they appear in that order.
• Inferring cause. The table does not explain why Business students report more group study.
• Using respectively with a long list. The listener should not have to reconstruct six pairings.

Worked scan: the same table, two possible hierarchies

Consider an original table of average waiting time in minutes:

A row-first scan reveals that technical support has the longest wait on every day and that its Monday figure stands apart from its later values. A column-first scan reveals that Monday has the widest spread, from seven to 24 minutes. Both are faithful; choose one rather than attempting two complete inventories.

Row-first response:

The table compares average waiting times for four services on three days. Technical support records the longest wait throughout, peaking at 24 minutes on Monday before remaining close to 15 or 16 minutes later in the week. Advice-desk waits are broadly stable at around 12 minutes, while equipment loans and document collection remain below ten minutes.

Here peaking is legitimate because Monday, Wednesday, and Friday are ordered time points, though the table does not show what happened on Tuesday or Thursday. “Before falling steadily” would be wrong: the figure drops to 15 and then increases to 16.

Column-first response:

Monday shows the greatest variation, with average waits ranging from seven minutes for document collection to 24 for technical support; by Wednesday and Friday, the four services form a narrower range of seven to 16 minutes.

This version reports range rather than every service. It does not say Monday was generally busier, because waiting time alone does not establish demand.

Deterministic unit and scope check

Decide whether each statement is supported.

1. “Every technical-support user waited 24 minutes on Monday.”
2. “The Monday technical-support average is nine minutes above Wednesday.”
3. “Document collection has the lowest or joint-lowest average each day.”
4. “Technical support is inefficient.”

Answers: 1 is unsupported because the cells are averages. 2 is supported: 24 minus 15 equals 9. 3 is supported by the three column comparisons. 4 is an interpretation without a benchmark, cause, or service-complexity measure. The table permits a statement about longer waits, not a judgment about efficiency.

Quick check

A table reports satisfaction scores out of 10: Site A 7.8, Site B 7.7, Site C 7.9, and Site D 5.2.

1. Safe cluster statement: A, B, and C cluster around 7.8.
2. Safe exception statement: D stands apart at 5.2.
3. Unsafe claim: D is a statistical outlier, unless the image supplies a relevant basis.
4. Unsafe explanation: D is lower because its staff are unhelpful. The table provides no cause.

Practice with a new grid containing three cities and four seasons. Choose one scan direction, give an overview, and support it with no more than four cells. Then check whether you preserved every unit and qualifier.

For other visual types, see PTE Describe Image maps, processes, and photos and PTE Describe Image trends. More practice is available through the PTE Academic Vocabulary hub and Exam English hub.

Sources

• PTE Academic Speaking & Writing test format — Pearson
• PTE Academic speaking test tips — Pearson

Lyra Practice is not affiliated with or endorsed by Pearson.