General Mathematics · Unit 3 · Topic 3

Time series analysis

Plot and describe time series, smooth them with moving means and medians, find seasonal indices by the average percentage method, deseasonalise, fit a least-squares trend line and forecast with it.

  • 7 lessons
  • 1 h 53 min of lessons
  • 12 worked examples
  • 11 hours in the syllabus
0 of 7

Start here

Time series plots

Lesson 1 of 7 · 12 min

Start the topic

Lessons

7
  1. Time series plots (up next)

    Recognise time series data, construct a time series plot with time on the horizontal axis, and read values and changes from it.

  2. Trend, seasonality and irregular fluctuations

    Describe a time series plot by its trend, its seasonality and its irregular fluctuations, each in the context of the data.

  3. Smoothing with moving means and medians

    Smooth a time series with a simple moving mean or median over an odd number of values, by hand and with a spreadsheet, and choose a sensible window.

  4. Seasonal indices by the average percentage method

    Measure how much each season sits above or below an average season by calculating seasonal indices with the average percentage method, by hand and with a spreadsheet.

  5. Deseasonalising data

    Remove the seasonal effect from a time series by dividing by the seasonal indices, reverse the process, and use deseasonalised values to compare seasons fairly.

  6. Fitting a trend line

    Model the long-term trend of a time series with a least-squares line against the time period t, fitted to deseasonalised values when the data is seasonal, and interpret it.

  7. Forecasting and solving time series problems

    Forecast future values by extending the trend line and re-seasonalising with the seasonal index, judge how reliable a forecast is, and carry out a full time series analysis.

  8. Topic quiz

    Ten questions across the topic, from foundation to complex unfamiliar. Your results feed your mastery.

About this topic

A time series is a set of values of one variable recorded at regular intervals: guest nights at a Sunshine Coast holiday park each quarter, the rainfall in Brisbane each month, the number of customers through a café door each day. Plot it against time and patterns appear. There is a long-term direction, a rhythm that repeats every year (or every week), and a scatter of bumps that no calendar explains.

In this topic you'll learn to see those three patterns — trend, seasonality and irregular fluctuations — and to separate them. A moving average smooths away the bumps so the trend shows through. Seasonal indices, found by the average percentage method, measure how far each season sits above or below a typical one, and dividing by them deseasonalises the data: it tells you whether a busy summer was busier than summers usually are. Fit a least-squares line to the deseasonalised values and you have a model of the trend, which you can push forward and put back into season to forecast.

Everything here is done the way businesses and government agencies do it — with a spreadsheet — so you'll see the formulas as well as the arithmetic.

Before you start

Builds on Bivariate data analysis 2 and Univariate data analysis 1.

  • The mean and median of a small data set, with and without technology, from Univariate data analysis.
  • Expressing one quantity as a percentage of another.
  • Fitting a least-squares line y=mx+cy = mx + c with technology, interpreting its slope and intercept, and the dangers of extrapolation, from Bivariate data analysis 2.
  • Entering data and simple formulas such as =AVERAGE(B2:B4) in a spreadsheet.

What QCAA says you need to know6 dot points

Describing and interpreting patterns in time series data

3 hours
  • Construct and use time series plots.LessonTime series plots
  • Describe time series plots by identifying features, including trend (long-term direction, e.g. increasing/decreasing), seasonality (systematic, calendar-related movements) and irregular fluctuations (unsystematic, short-term fluctuations).LessonTrend, seasonality and irregular fluctuations

Analysing time series data

8 hours

Subject matter from the General Mathematics General senior syllabus 2025 v1.3 (opens in a new tab), CC BY 4.0 © State of Queensland (QCAA).

The focus areas you can practise on their own. Each one maps to dot points above.

  • Time series plots

    Construct a time series plot with time on the horizontal axis and the points joined in order, and read values and changes from one.

    Practise
  • Trend, seasonality and irregular

    Describe a time series plot by its trend, its seasonality and its irregular fluctuations, in the context of the data.

    Practise
  • Moving means

    Smooth a time series with a 3-, 5- or 7-point moving mean, place each value at the middle period, and write the spreadsheet formula.

    Practise
  • Moving medians

    Smooth a time series with an odd-point moving median, and choose between the moving mean and median when the data has an outlier.

    Practise
  • Seasonal indices

    Calculate seasonal indices by the average percentage method, find a missing index from their sum, and interpret an index in context.

    Practise
  • Deseasonalising data

    Divide an actual value by its seasonal index to deseasonalise it, multiply to reverse it, and interpret deseasonalised values.

    Practise
  • Fitting a trend line

    Fit a least-squares line to a time series (or its deseasonalised values) against the time period t, and interpret its slope and intercept.

    Practise
  • Forecasting

    Use the trend line to forecast a future value, re-seasonalise it with the seasonal index, and judge how reliable the forecast is.

    Practise
  • Solving time series problems

    Carry out a full time series analysis in an unfamiliar context and use it to answer a practical question or evaluate a claim.

    Practise

Worked examples12

All 12

Common mistakes

Joining the dots out of order, or not at all

A time series plot has time on the horizontal axis, evenly spaced, and the points joined in time order by straight line segments. A scatterplot with no joins hides the pattern.

Placing a moving average at the start of its window

The 3-point moving mean of the values for periods 4, 5 and 6 belongs to period 5, the middle one. A 5-point average of periods 1 to 5 belongs to period 3. Smoothed series always start and end with gaps.

Multiplying instead of dividing to deseasonalise

Deseasonalised value == actual ÷\div seasonal index. A busy season (index above 1) should come down when deseasonalised. If your Q1 summer peak got bigger, you multiplied.

Averaging the raw values instead of the percentages

The average percentage method averages each season's percentages of its year's mean, not the raw values for that season. Averaging raw values ignores the trend and gives indices that don't add to the number of seasons.

Indices that don't add up

Quarterly indices must add to 4 and monthly indices to 12 (allowing for rounding). Check the sum — it catches arithmetic slips and finds a missing index.

Forecasting with the trend value alone

For seasonal data, the trend line gives a deseasonalised value. Multiply it by the season's index to forecast what will actually happen. Leaving it out forecasts an "average" quarter, not the one asked for.

Getting t wrong

tt counts periods from the start of the data: if t=1t = 1 is Q1 2022, then Q1 2025 is t=13t = 13, not t=2025t = 2025 or t=4t = 4. Count carefully, or use t=4(years after the start)+quarter numbert = 4(\text{years after the start}) + \text{quarter number}.

Treating a forecast as certain

Every forecast is an extrapolation. It assumes the trend and the seasonal pattern carry on unchanged. The further ahead, the less reliable it is.

Exam tips

  1. Describe a time series plot with all three features, each in context: "There is an increasing trend; there is seasonality, with a peak every Q1 (summer) and a trough every Q2; the irregular fluctuations are small." Say "no seasonality" or "no clear trend" if that is what you see.

  2. When you smooth by hand, write the working for at least one value, e.g. 38+21+293=29.33\frac{38 + 21 + 29}{3} = 29.33, and say which period it belongs to.

  3. Write seasonal indices as decimals (1.262) unless the question asks for a percentage, and check that they add to the number of seasons.

  4. Interpret a seasonal index in a sentence: "Q1 guest nights are typically 26.2% above the quarterly average."

  5. Define tt before you fit a trend line ("let t=1t = 1 be Q1 2022") and write the line with words: deseasonalised guest nights=0.556t+28.885\text{deseasonalised guest nights} = 0.556t + 28.885.

  6. For a forecast, show both steps: the trend value from the line, then × the seasonal index. Round only at the end, and give the answer in context and units.

  7. When asked how reliable a forecast is, mention extrapolation (how far beyond the data), the strength of the trend (rr) and whether the trend and seasonal pattern are likely to continue.

Where this shows up

  • The Australian Bureau of Statistics publishes seasonally adjusted figures for unemployment, retail turnover and building approvals so that real changes aren't confused with the usual seasonal ups and downs.

  • Tourism operators on the Gold Coast and in Tropical North Queensland plan staffing and pricing around the school-holiday and wet-season pattern in their bookings.

  • Electricity networks forecast demand peaks in summer heatwaves and winter evenings to plan generation and maintenance.

  • Retailers stock up for the Christmas quarter using seasonal indices from previous years.

  • Water utilities in south-east Queensland track deseasonalised consumption to see whether water-saving campaigns are working.

  • Meteorologists smooth rainfall and temperature records with moving averages to show long-term climate trends.

Save your progress with a free account