When data under analysis has a seasonal influence to it, further investigation cannot continue without deseasonalising the data. This involves calculating seasonal indices to tell us how a particular season (day/month/quarter) compares to the average season.
Seasonal indices have an average value of 1. This can be converted into a percentage for easier interpretation. A seasonal index of 1.3 (or 130%) would indicate that that season had 30% more than the seasonal average. An example is where Christopher works all throughout the year at a ice-cream shop and earns an average of $100,000 a season for it. If the seasonal index for summer was 1.5, then that means Christopher earns 50% more than the average $100,000. Likewise a seasonal index of 0.6 in winter would indicate that Christopher earns 40% less than the seasonal average.
A season index is defined by:
Note: the sum of the seasonal indices equals the number of seasons.
In order to remove the seasonal component of a time series, one must divide the amount by the seasonal index.
Want to suggest an edit? Have some questions? General comments? Let us know how we can make this resource more useful to you.