R help - Sep 2014 - Chen.Deo test

I made artificial random walk signals in a spreadsheet, and ran them
through the Chen.Deo (vrtest package) test in R.

When the "k vector" contains only, say, 2-3 values, I get the correct
VR
sum = approximately 1. However, when I add one or two more k values,
suddenly the VR sum goes up to around 1.6. Eventually, if I add all k
values from 2 to 10 inclusively, I end up with a VR sum around 3.45.

I don't understand why having more k values to check affects the result of
the VR sum. Is that a normal behavior of the Chen Deo test?

Thanks

	[[alternative HTML version deleted]]