I am working on survey analysis for my boss, and he wants me to group an open ended survey question's answers into 2 groups that contain the same # of responses (the question was, how many articles have you read in the last 2 months).

I figured I'd just find the median and split it there, but the problem is, I have an overwhelming mode of 10, which is also the median. So even if I split it up as 1-9 and 10+ or 1-10 and 11+, I end up with 2 lopsided groups by about 7-11.

Is this bad? Do I NEED to have them be equal, or approximately? He made it seem that way, but I would think the rule is to have them as close as possible. But correct me if I'm wrong. Should I just pick the least lopsided of the 2 options I listed above or use another method I'm not thinking of?