I have not observed any problems with the power trendline. It is
difficult to diagnose what is happening in your case, because you give
so little information.

A common problem is to have data that look like numbers but are actually
text; =COUNT(range) should give the same result as =COUNTA(range). I
would expect text data to make the power option unselectable, but I
understand you to have said that the power option is selectable, but
still fits a straight line (slope and intercept).

15 data points is minimal, include their values in the body of your
reply (no attachments, please) along with the equation that Excel
reports for the Power trendline option.

Jerry

CR Optiker wrote:

I have a scatterplot with 2 series: one a linear x-y line, one a dataset of
15 datapoints from (15.94,16.17) to (239.06, 248.37)...ie, no zroes or
negative numbers included. I request a trendline, select POWER, but it
gives me a LINEAR trendline, verifiable as linear because the equation is
is showing on the plot is a linear equation, not a power equation. At one
point I got a message about some trendlines not able to be generated due to
zeroes or negative numbers, but I don't have any of those.

Nothing I do seems to give me a power fit.

Suggestions?

Thanks!

Optiker