I am not sure how effective I was at explaining the use of trend projection. However, since linear regression is an extension of the same topic within my forecasting series, hopefully I am able to explain the concept a little more clearly.
Linear regression is the art of drawing a line to represent the relationship between points plotted between two axis. This line is the smallest sum of squares between all of the points and this line can be projected forward to help make predictions and educated guesses. In the case of trend projection, an analyst is looking at the data value, typically demand or sales over time. However, linear relationships can be drawn without using time as one of the axis. Consider a relationship drawn between sales and payroll. If pay is based on a commission, a positive linear relationship could show how a higher commission may lead to greater sales. This is just one example.
Another would be how sales might increase as advertising increases. An analyst could predict a certain level of sales based on a specified level of advertising by using this model.
Now as I mentioned in my previous post, predictions in this world are never 100% accurate. However, by using the Standard Error of the estimate, also known as the standard deviation about the line, one can predict the probability that a forecasted value will land within a particular range of the predicted value.
I hope that I haven't lost too many of my readers trying to understand this. I will not try to explain the statistics of this any further. Instead, I will direct you to a wonderful free online statistics resource http://onlinestatbook.com/.
Well, thanks for reading. The next part of this series will be in how to track and follow a forecast.
----Sincerely, Trevor Stasik.
To return to initial post about forecasting, click HERE.
To visit the next post in the forecasting series, click HERE.
Forecasting, Linear Regression, Trend Projection