Suppose a clothing manufacturer has collected data on the number of defective socks it makes. There are six subsidiary companies (factor C) that make knitted socks. At each company, there are five brands (B) of knitting machines with 20 machines of each brand at each company. All machines of all brands are used on the different types of yarn (Y) from which socks are made: cotton, wool, and nylon. At each company, data have been collected from just two machines (M) of each brand for operation by each of four locally resident workers, using each of the yarns. And on each occasion the number of defective socks in each of two replicate samples of 100 socks is recorded. Which factors do you think should be treated as fixed and which as random? Give reasons for your decisions.