How does Excel 2003 calculate trendlines with respect to axes?
 

I'm using Excel 2003. Excel is giving me the wrong equation for the
trendline it produced. The curve resembles a parabola pointed downward with
a positive offset on the y-axis. The x-axis has points -80, -70, -60, ...,
60, 70, 80. I changed the y-axis so that it crosses the x-axis at point
zero. I do not believe that zero on my x-axis is zero according to the way
it comes up with the trendline equation. I wish to figure out how to set up
my x-axis with positive and negative numbers and still have the equation for
the trendline be accurate.

Steve -

Make sure you're using an XY Scatter chart, and not a line chart.

Make sure you are looking at the trendline formula with sufficient decimal places to
assess the coefficients.

If you're making a linear or poly fit, the sign of the data values will not affect
the accuracy of the coefficients calculated.

- Jon
-------
Jon Peltier, Microsoft Excel MVP
Peltier Technical Services
Tutorials and Custom Solutions
http://PeltierTech.com/
_______

BeefmanSteve wrote:

I'm using Excel 2003. Excel is giving me the wrong equation for the
trendline it produced. The curve resembles a parabola pointed downward with
a positive offset on the y-axis. The x-axis has points -80, -70, -60, ...,
60, 70, 80. I changed the y-axis so that it crosses the x-axis at point
zero. I do not believe that zero on my x-axis is zero according to the way
it comes up with the trendline equation. I wish to figure out how to set up
my x-axis with positive and negative numbers and still have the equation for
the trendline be accurate.

Jon, I was using a line chart. I tried using the XY Scatter chart, and that
fixed the problem. Thanks one million!
-Steve

"Jon Peltier" wrote:

Steve -

Make sure you're using an XY Scatter chart, and not a line chart.

Make sure you are looking at the trendline formula with sufficient decimal places to
assess the coefficients.

If you're making a linear or poly fit, the sign of the data values will not affect
the accuracy of the coefficients calculated.

- Jon
-------
Jon Peltier, Microsoft Excel MVP
Peltier Technical Services
Tutorials and Custom Solutions
http://PeltierTech.com/
_______

BeefmanSteve wrote:

I'm using Excel 2003. Excel is giving me the wrong equation for the
trendline it produced. The curve resembles a parabola pointed downward with
a positive offset on the y-axis. The x-axis has points -80, -70, -60, ...,
60, 70, 80. I changed the y-axis so that it crosses the x-axis at point
zero. I do not believe that zero on my x-axis is zero according to the way
it comes up with the trendline equation. I wish to figure out how to set up
my x-axis with positive and negative numbers and still have the equation for
the trendline be accurate.