It's arbitrary. 0 = noisier, 5 = less variance charted. Think of 0 as "raw" and 5 as "normalized". I like 3.
I find STD skews high for N/A cars; I think the over generous correction is more appropriate for FI cars given that there is no good correction factor to account for the change in temp due to boost/intercooling.
For N/A cars, I prefer SAE, which tends to be a bit more conservative.
That said, all of the correction factors are equally valid; it just makes it easier to compare and contrast runs done in different ambient conditions if you use the same correction.
But... check this out
Proven Power Dyno Database Thread