If 8-bit, 10-bit, 12-bit, etc. ADC results are left-justified in 16 bits, you can treat the 16-bit value, regardless of the actual ADC precision, as 'N' 65536'ths of full scale. Less precise only means more zero-fill on the right. This way, you can use a common routine and simply increasing the number of ADC bits yields more resolution in your measurements without code changes.