# Weighted average (files opened per person)

1. ## Weighted average (files opened per person)

Good day all you awesome people!

Have one that has my mind cracking:

Lets say we have John and Pete, and they both opened 300 and 400 files respectively for the month, giving them a nice healthy 350 files per person. Woohoo!

Now along comes Paul, opens a meager 1 file for the month, which makes it 3 people for 701 files. The average of this works out to be 233.67, which brings the KPIs for the other two guys down tremendously!

What I would like to achieve is some sort of weighted average (if possible), which gives a much more accurate answer (closer to 350 files per person).

The amount of people and files per month can vary, so this calculation needs to be super flexible.

Now lets all put the thinking caps on!

Thank you!!

Kind regards,
J

2. ## Re: Weighted average (files opened per person)

I dont know if theres a proper term for it but I always knew it as "Trimmed Mean" which involves removing the smallest and largest x% of the values and averaging the rest. Not sure however, how that works if theres only 3 numbers (Im guessing 300 would be the answer after you've removed the 1 and 400 values).

In bigger data sets however you would get something like the following:

Data
1,20,20,30,30,30,50,70,100,270

Averaging that would give you 62.1

however, removing the top and bottom 10% would give a list of
20,20,30,30,30,50,70,100

which when averaged gives 43.75 which, is more indicative of the actual scores given ie 75% of the 8 numbers are in a range of 20 to 70

3. ## Re: Weighted average (files opened per person)

Try weighted %age

That is file opened by the person divided by total files multiplied by 100
In your example it it will be
42.79
57.06
0.14

You may replace 100 by another digit say 10 to get points out of 10
Regards

There are currently 1 users browsing this thread. (0 members and 1 guests)