Instead of "5 times as many" look as "a difference of 4 samples" - in an environment where the expected sampling noise is something like +/- 4.

"Statistically insignificant" means just that - this data is not sufficient to conclusively say if lower class women receive less callbacks than lower class men; maybe they do, maybe they don't.