It would be interesting to gather data on the subject, but it seems to be one of those commonly-understood truths. I suppose you could study it by showing men and women a pair of conversation transcripts about relationships, one of which accomplishes nothing, and one of which ends in a useful resolution ("I feel X," "That's terrible," versus "I feel X," "You should Y."). Then men and women could rate which conversation seemed better, according to a couple qualities.

It might be hard to get funding, though.

This isn't a topic that a study or statistics would help illuminate at all. Some stories or arguments are best made with anecdote and example. Attempts to apply scientific methods to arguments like this usually just lead to bullshit studies like the one debunked here http://languagelog.ldc.upenn.edu/nll/?p=2232 and are even worse.