Calculating Power

Hi Everyone,

I am doing a correlation analysis to see which variables impact the quota achievement of a set of sales reps at my company. I am trying to figure out how significant my results are, but am not sure if I am calculating the statistical power the right way.

I'm grouping quota achievmeent into 3 categories: <80% achievment, 80%-120% achievement, and >120% achievement

In one of my analysis' I'm looking at achievement by manager: the distribution of reps across the 3 achievement groups based on manager. There are 39 reps in total and 10 managers, so each manager has 3-5 reps.

When calculating power, I'm not sure if my sample size should be 39 or 3-5 since that is how many are in each subgroup. I used this tool:

Am I on the right path?