The utility of the area of study referred to as the social sciences rests on several assumptions.
The first assumption is that behavior of a large and diverse demographic can be predicted using results from a relatively minute sample set. A statistical significance test can usually falsify a social sciences result prima facie.
The second assumption is that in a social sciences study, it is possible to control for all confounding factors. Given the richness and depth of the individual human experience, this seems unlikely. Even if diversity is superficially maintained, a closer examination shows that it is impossible to account for all sets of inherited and learned behavior traits, nullifying any conclusions. Because of this inherent weakness, controls in social science studies are easily attacked. Critics of social science are brought out as being “pedantic” or “antisocial” or “not willing to compromise” or “unconcerned with the fate of the children”; or accused themselves of whatever allegedly immoral behavior is under attack, or of supporting some other clearly immoral behavior that happens to have some spurious association with the behavior under attack.
The last assumption is that predicting the behavior of autonomous, sentient individuals is within the scope of science itself. Science can only be used to divine facts and make predictions regarding deterministic systems. Behavior is not deterministic, because behavior can change whimsically, or it can change because the subject realizes they are being observed. The behavior may change to “good” behavior, to avoid engaging in a criticized activity, thus tainting the result; or it may change to “bad” behavior, to thumb one's nose at the establishment, to defy authority and dare to do that which is under criticism – also tainting the result. Social science assumes that behavior is ordered and caused. This places self-determination in a very low regard.
I believe (but cannot prove) that all are false. Repeated observation has shown that whenever social science is used to determine public policy, the outcome rarely if ever meets expectations, while increased government police power and economic meddling is justified as necessary to implement the policy and prevent it from failing. Failure, however, is usually rewarded with more funding or increased powers, in an attempt to stem the tide of whatever hated behavior continues to take place.
Fortunately, social sciences are usually only used to justify nanny-state behavior and/or thought control, so the philosophical justification for such controls can be challenged simultaneously with the practical results they fail to bring.