Skip to main content

Hello, I have a dataset that contains average score of a field audit by contractor name with different number of audits. What would be the best way to normalize the data?

for example

contractor A, average Score = 98.8%, # of audits 108

contractor B, average Score = 96.4%, # of audits 79

contractor C, average Score 100%, # of audits 6

3 answers
  1. Aug 30, 2022, 12:56 PM

    Ah right @Marina Derechinskaya​ 

    For a weighted average you would need something to weight it by. So if each audit had a different number of checks [Num checks], you might say:

    SUM([Num checks] * [Score]) / SUM([Num checks])

    But if each audit has the same number of checks, or you don't have that data then I suspect you've created the best average you can. And in that case I might just show the volume of audits next to the percentage for context (so a user can see "ah, customer C got 100% but only had 6 audits, so perhaps there isn't enough data to draw too much of a conclusion").

    Ta, Steve.