The problem was in the phrase "other things being equal". Other things are never equal. Studies try to make them as equal as they can, and wind up differing as to which things are more important to get equal, and wind up with different results.