Unusual Weighted Average Calculation

Question

Unfortunately I don't know the term for what exactly I'm looking for. I hope you guys can direct me.

Problem at hand: Have a composite score for an employee performance, for example a score may comprise of four discreet values.

score = {
   writtenSkills: 10,
   verbalSkills: 16,
   punctuality: 20,
   timeoffRequests: 9
}

In the example above, writtenSkills, verbalSkills and punctuality are all positive skills, however timeoffRequests are negative.

I'm looking to compare and measure these individual composite scores against an arbitrary predefined benchmark score of say:

benchmark = {
   writtenSkills: 15,
   verbalSkills: 15,
   punctuality: 25,
   timeoffRequests: 5
}

Where my benchmark is the baseline (say 100%) and each individual composite score is some fraction of that benchmark (76%, 164% etc.).

If the individual metrics of the composite score were all positives, I could have done a straight up weighted average, but how can I do this for the metrics that some are positive and some negative?

Answer 1

I think you could do a weighted average as planned, but use (benchmark.timeoffRequests - score.timeoffRequests) in place of score.timeoffRequests in the average.