Both quadratic and logarithmic equation fit the data.

[Statisticum]

I have plotted my data (about ten pairs) in a scatter plot an then I added the trendline to see the equation that bests fit my data. Excel returns R-square=0.85 for the polynomial (quadratic) equation and R-square=0.87 for the logarithmic equation. Which one shall I choose? Is it possible that both equations fit the data?

[MrShorty]

I'm sure it is possible for both equations to fit the data. As for choosing -- that choice needs to come from what you know of the data and the real world process the data represent and what you need to do with the regression.

For example, quadratics have a max/min/vertex. Some real world processes also have max/min points, others do not. I might take that into consideration. If the data I am analyzing cannot or should not have a max/min, then I might prefer the logarithmic equation, because that will be continuously increasing/decreasing.