matlab model compare - A theroy/concept math question

Hello
I having a problem to compare two sets of measurements each have 4 parameters
I tried interp1 and I got for each parameter a new measurement and now I can calculate the error from the interp data
What is the best way to calculate the error and how can I know if it's good enough
I have seen the graphs that I got from the interp data and compared it visually and it looks fine
I just think it will be better if I use a different way
I thought of taking the original data and from him create some kind of plane - which will be used as comparison and then check how much does the other data fit to him... get a the new data to distribute around the plane which will be my zero plane and that way I will know the absolute error as well

any ideas? thoughts will be most wellcome