Although the coefficient of correlation is well defined for a few regression models, you do not calculate it directly because your calculator does that for you.
I could just tell you that it is displayed along with the statistical results (equation of best fitting curve, ) but if you do not know how to do that you will just shoot my post. If you know how to do a linear regression, stop right here and ignore what follows.
For those interested: How to do a regression on the calculatorI am sorry but I will have to assume that you know how to enter the data through the [MENU][STAT].
- After you finish entering all the data
- Press [F1:Graph], select Gph1,
Gph2, or Gph3.
- Once the graph of the experimental data is drawn, you
will have access to the diverse regression models.
- Your choices are
- [X:Linear]
- [Med-Med],
- [Quadratic :X^2]
- [Cubic;X^3]
- [Quartic: X^4].
- Press [F6] to scroll right to access the others
- [Logarithmic:LOG],
- [Exponential :EXP],
- [Power :Pwr] and
- [Sinusoidal: Sin]
After you select a model and press [EXE], the screen displays the
equation and its various coefficients, a correlation coefficient if it has meaning, etc.
On that same screen, you have a
menu option [F6:Draw]. If you select it, the regression curve will be
drawn on the same screen as your statistical graph. This allows you to
judge if the fit is good. If it is not, you have the option to choose
another model.
The other option [F5:Copy] allows you to save the regression equation in the Y1, Y2, etc. variables.