Problem:
An important aspect of optimizing the fish finger production line is developing an inventory policy for raw materials used in the production process (oil, breadcrumbs, boxes, etc). A Lowliner food is considering three new policies inventories (EOQ, S-s, and S-R) and has Eddie to modify his simulation model to evaluate them. Eddie executes ten replications of his model under each inventory scenario and records the average daily inventory cost (in $Can). Results from the three models appear below:
|
Policy
|
|
Run
|
EOQ
|
S-s
|
S-R
|
|
1
|
107.5
|
102.7
|
110.6
|
|
2
|
111.9
|
92.7
|
117.4
|
|
3
|
102.7
|
99.5
|
114.5
|
|
4
|
112.4
|
93.1
|
113.6
|
|
5
|
115.2
|
89.3
|
111.8
|
|
6
|
114.6
|
96.2
|
113.1
|
|
7
|
112.6
|
89.7
|
112.0
|
|
8
|
105.7
|
93.8
|
107.4
|
|
9
|
108.5
|
101.4
|
113.4
|
|
10
|
110.1
|
86.0
|
111.3
|
|
Mean
|
101.1
|
94.4
|
112.5
|
|
Variance
|
16
|
30.1
|
6.8
|
Assume that Eddie wishes to identify the best possible policy (where lower cost is considered better). Moreover, assume that Lowerliner wants to be 90% certain (i.e. a = 0.10) that the selected policy is the best. Using this information, can any of the three policies be identified as "best"?