Second mathematical problem for me within a fortnight! Who'd have thought it
My supervisor is in control of 2 teams in my office (of which I am one). The other team spend the day taking amendments from customers to their licences with us and then produce updated versions which are sent to me to check and produce a record which I submit monthly to my/our supervisor.
Problem: The record consists of basic documentation on whether or not their updated contracts were correct or not. Whilst the numbers are fine, it is unreflective of the person's true performance. In this case, one person got 110 out of 160 contracts correct whereas another (a very smug another too..) only did TEN and got 8 correct.
Since this morning, the rest of the other team has been most displeased about this person and their lack of work and quite franky, it's unacceptable. I appreciate it is easy to disect these figures and clearly see that she only did 10 (the soop isn't stupid). But without a definitive grade which represents correct AND output you still have the problem of umming and arring over exactly how many contracts must be completed to justify a good percentage. Also, 6 months down the line, the fact that only 10 were completed will be forgotten but 80% will remain on record.
Question: Is there a diversity index which give me this definitive grade I could suggest to my supervisor we add to the record? I've been having a look around and I've had a little play around with Simpson's and Shannon's before in regards to species diversity of place but this isn't really my field and I'm not sure how I'd translate these figures into those kind of equations. But a final accuracy score of 0 - 1 (or x100) like these would give, would be ideal. Any help/pointers would be fantastic =)
I'll give a small example:
A) Contracts 70: Correct 50, Incorrect 20
B) Contracts 10: Correct 8, Incorrect 2
C) Contracts 164: Correct 110, Incorrect 54
D) Contracts 31: Correct 17, Incorrect 14.
Thanks in advance for any help