The other common user error is to use a "Line" chart instead of an "XY
(Scatter) chart. This will result in coefficients that are actually wrong,
instead of merely rounded beyond usefulness.

The "Line" chart occurs earlier in the chart-type menu and is misleadingly
named, inviting the obvious (but wrong) interpretation that it has something
to do with whether you want to connect the points by a line. A "Line" chart
is a chart where the x-axis is composed of categories instead of a numeric
scale (unclear how MS got "Line" out of that). If you provide x-axis
information at all, then it is interpreted as category labels, even if the
range solely contains numeric values. If you request a trendline (unclear
why MS supports trends on non-numeric x-values), then the x-values for the
trend are taken to be 1,2,3,... regardless of the actual x-data provided.

Jerry

"Bernard Liengme" wrote:

Why do you think it incorrect. My bet is you have too few decimals
displayed. Format the trendline equation to show 15 places and try again.
Then experiment with LINEST to get polynomial coefficients as shown at
http://www.stfx.ca/people/bliengme/E...Polynomial.htm
best wishes
--
Bernard V Liengme
www.stfx.ca/people/bliengme
remove caps from email

"Jimg" wrote in message
...
I am trying to determine the equation defining a curve in Excel. When I
apply a trendline using a polynomial equation, the equation returned is
incorrect. Is there a method to "fit a curve" and determine the equation
defining the curve?