In non-mathematical sciences what I observed is usually people apply the existing material and methodology to different location, species, scenario, conditions etc (May be some other type of research method may exists as I don't have full knowledge of them). And in mathematical sciences researchers either apply the existing methods (with possible modifications in the methods) to different problems or the extended the existing general results. (may be some other). So a natural questions arises what is a good research. I guess an intuitive way of thinking will be better than above?