Embedded Files
A time series graph is essentially a line graph. To identify it is a time series some measurement on time will be on the x-axis
A typical feature with time series is that there is often a lot of noise in the graph. What this means that the line will often go up and then down, and you may even see repeating patterns.
Long term trend
Long term trend
Long term trend is the overall picture of what is happening. It's ignoring all the noise in the data, and looking to see overall is it increasing, decreasing, or not changing.
If you're having trouble seeing the trend among the noise, you can blur your eyes to see the overall picture
Example description of long term trend:
Example description of long term trend:
In the graph the number of filled jobs increased from on average about 1.8 million jobs over the period mid-2004 until early-2008 where it reached 1.96 million jobs. From early 2008 until 2010 the number of filled jobs decreased to about 1.88 million. Following the decrease the number of filled jobs increased again.
A reason for this decrease could be due to the Global Financial Crisis that occurred in 2008-2009 that led to businesses making less or losing money, and therefore not being able to employ as many people.
Seasonal Effects (Short term trend)
Seasonal Effects (Short term trend)
Example description of short term trend:
Example description of short term trend:
In the graph above it shows the number of filled jobs in NZ. The graph shows a short term trend due to the repeating cycle of the data. We see that every December the number of jobs filled is at its highest point, while in every January it is typically its lowest point. A reason for this could be that every December there is a lot of jobs as lots of people are shopping to buy Christmas presents, while there are far fewer jobs filled in January as many businesses close over the period or go on holiday.
Short trend is a pattern that we see happening repeatedly at the same time each interval. From the graph on the left we can after a high peak each year the next month it drops down low.
This can be seen in the graph below which has summarised the average effect for each month.
Exercises
Exercises
Making Predictions
Making Predictions
Making a prediction about the future is also known as making a forecast (for example a weather forecast).
In order to make a prediction you need to do the following:
Continue the trend line
Add on the seasonal component
Make a prediction
When making a prediction give a specific number value (for example 9250), don’t be general statements such as "OVER 9000"
Discuss accuracy of prediction
To determine the accuracy of your prediction it is important to consider whether it is following a constant cycle. If this is the case you can likely have a good accuracy in your prediction. However if the overall trend has begun to change, or it has become more sporadic recently then there will be less certainty in your prediction.
Page updated
Google Sites
Report abuse