There are several scientific difficulties in successfully applying social science to solve social problems. There are three main scientific issues (Randolph and Haynes, 1968):
1. the so-called “Hawthorne effects” or changes in behavior that result from the fact that individuals are subjects in an experimental study;
2. the inadequacy of existing data on social problems and individual behavior and the flaws of indirect data; and
3. the manipulability of social factors variable in social scientific analyses of problems.