Mid-year convention adjustments for final year discount number for multiples method vs. gordon growth?

I don't understand exactly why one would add 0.5 to the final year discount number for the multiples method, but would leave the gordon growth method's final year discount number as is..