Forum Discussion
Interpreting/Explaining Key Influencers Visual
In this example, Key Influencer suggests that when Country = Guatemala, Over all case assessment is 1.57x more likely to be Fail compared to average (8.97%). How do we arrive at 1.57x? I was thinking that to get the likelihood, ill divide 27.02% (which is Guatemala's %Overall case assessment fail) with 8.97% (average %Overall case assessment fail of other countries) --- but that is not 1.57. I would love to hear your thoughts and learn how to BEST interpret this kind of visual. Thank youuuuuuuuu!
Hi jianlearn ,
It is right that there's no formula that will let me replicate what Power BI did, I think you can divide 27.02% with 8.97%, but that means the chances of a case failing are 57% higher than the average chance of failure for all countries.
Best Regards
Yilong Zhou
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
3 Replies
Hi jianlearn ,
The Key Influencers visual in PowerBI is designed to help you understand the factors that influence a particular outcome. It uses statistical analysis to determine how much more likely an outcome is when a certain condition is met, compared to the average. You can read this document for help: Key influencers visualizations tutorial - Power BI | Microsoft Learn
In your example, to explain the 1.57 times figure, this does not mean that Guatemala has a 157% failure rate, but rather that the chances of a case failing are 57% higher than the average chance of failure for all countries.
Best Regards
Yilong Zhou
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
thank you! am i right to say that there's no formula that will let me replicate what Power BI did? and that i should not divide 27.02% (which is Guatemala's %Overall case assessment fail) with 8.97% (average %Overall case assessment fail of other countries) to get the answer?
Hi jianlearn ,
It is right that there's no formula that will let me replicate what Power BI did, I think you can divide 27.02% with 8.97%, but that means the chances of a case failing are 57% higher than the average chance of failure for all countries.
Best Regards
Yilong Zhou
If this post helps, then please consider Accept it as the solution to help the other members find it more quickly.
