0

I have a employee dataset with the following 5 details.

projects completed
customer ratings
number of bugs reported
customer complaints
profit

I want to rank these employess based on their efficiency. We have observed the following facts of each detail as follows.

projects completed -> highly efficient employees tend to complete many projects
customer ratings -> highly efficient employees has a good customer ratings
number of bugs reported -> highly efficient employees has less number of bugs reported
customer complaints -> highly efficient employees has less customer complaints
profit -> highly efficient employees make a higher profit to the company

I am wondering if I can convert the above observations to derive a ranking function to rank the employees. Please recommend me a suitable approach to perform my task of ranking.

I am happy to provide more details if needed.

EmJ
  • 592
  • 3
  • 15
  • 2
    You can rank them in as many ways as you like. We can agree that some measures count positively and some negatively but even with that you still to have choose weights for each variable and indeed whether linear combinations make sense at all. There isn't a neutral, predefined statistical way of doing this, any more than there could be for scoring your or my friends according to say how nice they are and how intelligent they are (assuming that we have numbers on both). – Nick Cox May 22 '19 at 00:18
  • @NickCox Thanks a lot for the comment. I am curious to know if there is a way to obtain weights. I do have a labled dataset that indicates their actual `efficiency rank`. My actual target is to derive a ranking function that is quite similar to the labels given. Looking forward to hearing from you. Thank you very much. – EmJ May 22 '19 at 00:30
  • 2
    That's now a quite different question. If you have given to you an overall ranking or assessment then relating it to other variables is then a problem in regression. But you have side-stepped the main point I am making. Weights must come from outside not the data. – Nick Cox May 22 '19 at 00:38
  • @NickCox Thanks a lot. I tried regression with my data and got really bad results. That is why I am trying to use other different approaches. Are there any other methods/techniques that I can use instead of regression? Thank you very much :) – EmJ May 22 '19 at 01:04
  • Sorry, but I don't have further suggestions, not least because I am less clear than when I started about what your real goal is. – Nick Cox May 22 '19 at 01:06

0 Answers0