I am using cokriging options for interpolating two datasets(A & B) in arcmap 10.1. I have first kept A in datasets 1# and B in datasets 2#. I got 0.8461 as the root mean square prediction error(rmspe). However, when I changed the order i.e, tried B in datasets 1# and A in datasets 2#, the rmspe increased to 3.6321. My doubt is that when we change the order , does it assign different weightage to the dataset which results in different rmspe?
I searched it on ESRI blogs but didn't get any solution.
Kindly help me with this.