Question

Formula for calculating scores

  • 27 October 2021
  • 4 replies
  • 4810 views

Badge

Hi,

I have 10 people in my group, 8 of them responded to survey. I ended up with a score of 91% on one question. I can also see in the exec summary that “% favorable” for that question is 100. How do I end up with 91%? Unless there is different weight for 4 and 5 answers, it seems impossible to get 91%

Thanks

 


4 replies

Not a Glint employee, but came across this in prepping for our change mgmt switching from %fav to average… so to answer @soren.nielsen’s question, yes, though 4s on a 5 points scale constitute a weighted score of 75.

From Primary Metric of Analysis - People Science Explained | Glint Community (glintinc.com)...

Scaled Average Score = (Raw Average - Scale Minimum ) * 25

So if everyone scores a 4 on a 1-5 scale… 4 minus 1 = 3 times 25 = 75.

If you have 8 out of 10 respondents say 5, 5, 5, 5, 5, 4, 4, 2, the raw average would be 35/8 = 4.375.  Subtract scale min from that = 3.375, multiply by 25, you get 84.375, or rounded to 84.  Not much different in this case from % fav, which would be 87.5 ~ 88 but different.  

Hope this helps!

Hi @soren.nielsen - check out the Scores Calculation overview, especially the PDF attachment deep dive and let me know if it helps.  If not, I will need to pull in someone much more statistically inclined than myself to help :). 

 

Thanks you @bcolver for the article, but I have the same question that @soren.nielsen. Another example: If I have a group of 10 employees but only 8 respondend, ¿the score calculated as average is from 8 or 10 employees? because I have categories with score 91 and % favorability 100%. Thanks you in advance!

Badge

Thanks for the info, but it doesn’t exactly cover my question. I think I’m getting the gist of it. E.g. a question can have 100% favorable responses, but if ALL the favorable responses were 4’s then the score would “only” be 80%? (given that you get 80% for a 4??)

Userlevel 7
Badge

Hi @soren.nielsen - check out the Scores Calculation overview, especially the PDF attachment deep dive and let me know if it helps.  If not, I will need to pull in someone much more statistically inclined than myself to help :). 

 

Reply