First of all you could have used buckets. Secondly, it doesn't seem like you have enough data judging from your charts.
Dice did a study and found that
>when you control for education, level of experience and parallel job titles, says Dice, men and women earn the same amounts.
http://www.forbes.com/sites/susanadams/2013/03/20/women-are-...
> At that point I made a decision to change this by enforcing a new hiring rule. Next couple of new employees are going to be women up until we balance our team
I hope you get sued