Dear gazo,
Given a line drawn through a two-dimensional Cartesian data plot, the
R-squared value is a measure of the aggregate difference between the
line and the data points. This value is calculated in such a way that
it falls between 0 and 1. To quote the following website, "R squared
is the relative predictive power of a model. R squared is a
descriptive measure between 0 and 1. The closer it is to one, the
better your model is."
Children's Mercy: Stats: Ask: r-squared
http://www.childrens-mercy.org/stats/ask/rsquared.asp
In the case of two sets of data, one resulting in an R-squared value
of .75 and the other in an R-squared value of .90, the latter is
deemed more accurate and a better fit to the reality that has been
experimentally measured. This is because .90 is closer than .75 to the
perfect value of 1, meaning that the points in this data set fall
closer to the line of best fit. The more tightly the data points are
clustered near the line, the less error there must have been in the
measurements, and so the better the experiment.
For more information on calculating R-squared and what it says about
the accuracy of your model, consult the following site.
University of Leicester: On-line Statistics: Linear Regression
http://www.le.ac.uk/biology/gat/virtualfc/Stats/regression/regr1.html
If you feel that my answer is incomplete or inaccurate in any way, please
post a clarification request so that I have a chance to meet your needs
before you assign a rating.
Regards,
leapinglizard
Search Queries:
r squared best fit
://www.google.com/search?hl=en&lr=&ie=UTF-8&c2coff=1&q=r+squared+best+fit
r squared regression
://www.google.com/search?hl=en&lr=&ie=UTF-8&c2coff=1&q=r+squared+regression |
