I can’t seem to wrap my head around the concept of a log-linear trend model…
Why would a linear trend model not be appropriate for a time series that exhibits exponential growth ?? Why would the regression risiduals be persistant?
It’s easy to see this is if you create a graph in Excel.
In column A – cells A1 through A101 – put the values 0.0, 0.1, 0.2, . . ., 9.8, 9.9, 10.0
In cell B1, put “=2^A1”, then copy this down column B to cell B101.
Highlight all of these cells, and insert a scatter chart.
Right click on the curve in the chart, and click “Add Trendline…”
The default trendline is linear, so click the Close button.
You’ll see that the trendline (regression line) is a lousy fit to the exponential function.
Cool. Thanks s2000 magician
You’re welcome.