Process Costing

I found this question in the revision kit (BPP learning media). Normally, no losses are expected from a process. Any abnormal losses are sold for scrap. Which of the following calculates the net cost to the company of one unit of abnormal loss? i) Total input cost / actual output units ii) Total input cost / expected output units iii) (Total input cost - total scrap value)/ expected output units iv) (Total input cost / expected output units) - scrap value The answer that I chose was option (ii): Total input cost / expected output units, yet when I checked the answer, it says option (iv): (Total input cost / expected output units) - scrap value. My argument here is that subtracting the scrap value after dividing the total input cost by the expected output doesn't make sense to me. I am confused why the answer is option (iv) and not option (ii). I humbly request for help. :-)