Hard science and soft science are colloquial terms often used when comparing scientific fields of academic research or scholarship, with hard meaning perceived as being more scientific, rigorous, or accurate. Fields of the natural, physical, and computing sciences are often described as hard, while the social sciences and similar fields are often described as soft.[1] The hard sciences are characterized as relying on quantifiable empirical data, relying on the scientific method, and focusing on accuracy and objectivity.[2] Publications in the hard sciences such as natural sciences make heavier use of graphs than soft sciences such as sociology, according to the graphism thesis.
Within the areas of natural science and social science, some disciplines are viewed as "harder" than others. For example, physics is viewed as harder than, say, paleontology since the former but not the latter can make precise predictions about experimental data.
A related aspect of the hard versus soft distinction has to do with the ease of drawing strong conclusions. In soft sciences, there are often numerous variables that might have an influence on some variable of interest, and many of those variables either may be non-quantifiable or may be quantifiable but difficult to obtain data on; but further, even with plentiful data, it may be difficult to disentangle the effects of such a large number of variables. In contrast, typically in the hard sciences there are only a few, readily identified, causative variables, making it easier to infer specific causative effects.