| Page |  |
 |
|
 |
 |
Moving Averages
A time-series graph can be used to look for the trend, such as an increase in
sales over time or a decrease in smoking. Cyclical or other variations can hide
the general trend, as in the example of household electrical use.
The Trend
The trend is the behaviour once the variations are 'smoothed out'. For example,
ice-cream sales are likely to fluctuate throughout the year. Also, there may be an
increase or a decrease (a trend) in sales from one year to the next.
A ‘rough and ready’ way to obtain a trend on a time-series graph is to use a ruler
to draw the line of 'best fit'. Judging by eye where to position the ruler is a
problem if the variations are large. A better way is to remove the variations by
averaging the raw data. You can then examine the trend more carefully. The
process is called moving averages. The result of the averaging is then plotted
instead of the raw data. The line of best fit (called the trend line) can then be
drawn more accurately.
The moving averages graph and the ‘best fit’ are shown below. You can see
that the electrical use remained almost constant at about 675 units. There is a
slight upward trend. The gradient of the line measures the average increase in
electricity use per quarter. You can extend the line to estimate future use.
Moving Averages for Household Electrical Use
Line of Best Fit
0
100
200
300
400
500
600
700
800
900
1000
1
2
3
4
5
6
7
8
9
10
11
12
Quarters
|
 |
|