Interpreting graphs of recorded data

A dated table row gives 6 January and a daily count of 366. What does this pair represent?

Use a recorded dataset rather than an invented rule

This topic uses published Australian marriage counts for selected dates in January 2023

The counts are grouped by the date the marriage occurred, not the date a form was registered

The Australian Bureau of Statistics published these observations in its 2023 release, issued in 2024. They include registrations received in 2024, and later information may revise a count

We use a fixed copy of the published numbers, accessed 25 October 2026. The graphs are redrawn here, not generated from a formula

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

Data © Commonwealth of Australia, licensed under Creative Commons Attribution 4.0 International

https://creativecommons.org/licenses/by/4.0/

Read the title, variables and units first

The horizontal variable is the January date in 2023. The vertical variable is the published daily marriage count for Australia

The horizontal spacing between consecutive dates represents one day. The vertical numbered grid spacing represents two hundred marriages

Each point pairs a date with one count. Larger counts appear higher, not farther to the right

The same date has one published count in this display, but there is no supplied algebraic rule for predicting a count from a date

The horizontal axis shows dates in January 2023. What does moving horizontally from date 3 to date 4 represent?

A vertical scale has evenly spaced labelled lines at 0, 200, 400, 600, 800 and 1000. What value is halfway between 400 and 600?

Locate the date and read its point

Start at the requested horizontal date, move vertically to its plotted point, then across to read the count scale

A graph gives position and shape. A scale does not necessarily let you read every exact unit by eye

Use the accompanying table when an exact count is asked for. Do not round a count merely to match the nearest drawn gridline

The points and table must describe the same observations

Read the point for 6 January

The graph shows published Australian marriage counts by date of occurrence for 1 January to 7 January 2023

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023, published in 2024. Accessed 25 October 2026. Includes registrations received in 2024; later information may revise these counts

Data © Commonwealth of Australia, CC BY 4.0. Graph redrawn from the published counts

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

https://creativecommons.org/licenses/by/4.0/

Use the plotted point for 6 January and the accompanying exact table. How many marriages occurred on that date in this published data?

Read the horizontal date and the vertical count scale

Locate the requested date or height on the graph, then use the table for its exact count

At 6 January, the table gives 366 marriages

Read another date without changing the scale

Move P from 1 to 7 January. The recorded points stay fixed: the control selects an observation rather than changing what happened. Watch both the date and its count

What happens to the recorded count over these seven dates?

Read the actual date labels in a new interval

Find extremes only within the displayed interval

The maximum is the largest recorded output in the displayed set; the minimum is the smallest

Look for the highest or lowest point, then match it to its date and exact count

Do not confuse the date at the highest point with the count itself

A maximum for seven displayed dates is not automatically a maximum for the month or year

Locate the highest point and report its count

The graph shows published Australian marriage counts by date of occurrence for 1 January to 7 January 2023

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023, published in 2024. Accessed 25 October 2026. Includes registrations received in 2024; later information may revise these counts

Data © Commonwealth of Australia, CC BY 4.0. Graph redrawn from the published counts

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

https://creativecommons.org/licenses/by/4.0/

Use the graph to locate the highest recorded point and the table for its exact count. What is the maximum daily count within this displayed week?

Read the horizontal date and the vertical count scale

Locate the requested date or height on the graph, then use the table for its exact count

The recorded values in time order are 70, 61, 117, 146, 202, 366, 818

The smallest value is 61 and the largest value is 818

The requested value is 818 marriages

Find the largest count in the second displayed week

Find the smallest count, not the earliest date

Separate a value from a change in value

A point gives a daily count. A change compares two such counts

A positive result means the later daily count is higher. A negative result means it is lower

This difference is not the total number of marriages across the dates. Daily totals would have to be added for that different question

A negative change does not mean the later count itself is negative

Calculate a positive change from the actual counts

The graph shows published Australian marriage counts by date of occurrence for 1 January to 7 January 2023

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023, published in 2024. Accessed 25 October 2026. Includes registrations received in 2024; later information may revise these counts

Data © Commonwealth of Australia, CC BY 4.0. Graph redrawn from the published counts

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

https://creativecommons.org/licenses/by/4.0/

Calculate the signed change in the daily count from 5 January to 6 January. Subtract the earlier count from the later count; use a minus sign for a decrease

Read the horizontal date and the vertical count scale

Locate the requested date or height on the graph, then use the table for its exact count

The value at 5 January is 202 marriages

The value at 6 January is 366 marriages

Subtract the earlier value from the later value

The change is 164 marriages

A positive change means the later daily count is higher; a negative change means it is lower

Calculate the next increase

Keep the sign of a decrease

The graph shows published Australian marriage counts by date of occurrence for 8 January to 14 January 2023

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023, published in 2024. Accessed 25 October 2026. Includes registrations received in 2024; later information may revise these counts

Data © Commonwealth of Australia, CC BY 4.0. Graph redrawn from the published counts

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

https://creativecommons.org/licenses/by/4.0/

Calculate the signed change in the daily count from 8 January to 9 January. Subtract the earlier count from the later count; use a minus sign for a decrease

Read the horizontal date and the vertical count scale

Locate the requested date or height on the graph, then use the table for its exact count

The value at 8 January is 219 marriages

The value at 9 January is 136 marriages

Subtract the earlier value from the later value

The change is -83 marriages

A positive change means the later daily count is higher; a negative change means it is lower

Subtract in chronological order

Compare the endpoints without ignoring the middle

Describe all changes, not just the final segment

Follow the graph from its first displayed date to its last. State where the count rises or falls

A higher final count does not mean every intermediate change was an increase

For equal one-day intervals, a steeper joining segment represents a larger difference between adjacent daily totals

These joins do not show how marriages were distributed within a day

Do not call a short displayed pattern a long-term trend or assume that it repeats without further evidence

Describe the first week in date order

The graph shows published Australian marriage counts by date of occurrence for 1 January to 7 January 2023

Source: Australian Bureau of Statistics, Marriages and Divorces, Australia, 2023, published in 2024. Accessed 25 October 2026. Includes registrations received in 2024; later information may revise these counts

Data © Commonwealth of Australia, CC BY 4.0. Graph redrawn from the published counts

https://www.abs.gov.au/statistics/people/people-and-communities/marriages-and-divorces-australia/2023

https://creativecommons.org/licenses/by/4.0/

Describe the direction of change across this entire displayed week

Start with 1 January and follow the dates from left to right

The lowest point is at 2 January

The earlier consecutive counts decrease towards that point

After 2 January each consecutive recorded count increases

This description concerns only the displayed week, not a long-term trend

Describe the complete second week

A joining segment is not another measurement

The records are daily totals, one output for each listed date

Joining consecutive points helps the eye follow time order. The dataset does not supply a count for a fractional date or a count at noon

A straight segment drawn between two records does not prove that the underlying activity occurred at a constant rate

To estimate between measurements in another context, first justify that intermediate inputs have meaning and state any interpolation assumption

Distinguish a drawn join from an observation

Keep predictions separate from observations

An observation is supplied by the recorded data. A prediction goes beyond those records using assumptions or a model

Extending the last segment is one possible mathematical construction, but it does not guarantee the next actual count

A suitable model needs evidence, and a prediction should be identified as a prediction rather than a measured fact

Do not turn the last change into a guaranteed rule

Describe what changed before explaining why

A graph can show that a count changed. It does not by itself identify why that change happened

An explanation involving weather, preferences or another factor needs additional evidence about that factor

State the dates, direction and size of a change before considering explanations. Keep an unsupported explanation separate from the observed relationship

Reject a cause that was not investigated

Limit the conclusion to the displayed evidence

Read, calculate, describe and qualify

Identify the source, quantity, units and dates before reading the graph

Match the date to its plotted point and use the table when an exact count is required

Locate maxima and minima within the displayed interval, and calculate signed changes as later minus earlier

Describe the whole sequence, including changes of direction rather than just endpoints

Separate a recorded value from an estimated intermediate value, a future prediction or a causal explanation

These activities reuse a finite recorded dataset. A shuffled display is not new observational evidence