The physical sciences (also known as the “natural”, or “hard” sciences) emerged during the Age of Enlightenment and evolved into the disciplines we know today as physics, chemistry, astronomy, geology, and others. As the names imply, these sciences are aimed at explaining natural phenomena. A strict method of exploring questions about nature has developed with key elements such as posing a hypothesis, devising an experiment to test that hypothesis, drawing conclusions from the results, and perhaps most important of all, replication of the process by others who will arrive at similar results, thus confirming the conclusion. From these conclusions predictions could be made.
In the Twentieth century, investigators wondered whether these strict methods could be applied to the behavior of human beings. It is one thing to study the behavior of atoms, but could the scientific process be applied to the study of people with free wills. The satisfying answer is “yes” provided that studies are conducted on large groups of people. The social sciences (also known as the “behavioral”, or “soft” sciences) developed into separate disciplines like sociology, economics, political science, and others.
The key similarity is that social scientists must be as rigorous in applying their investigative methods as the natural scientists. There must be hypotheses, experiments, conclusions, replication, and predictability. The way to do this is to study a large enough sample of people to be satisfied that the conclusions are valid. The social scientist largely uses statistical analysis to study large enough numbers of people to be confident in their conclusions and predictions about the larger group they are studying.
America has just finished a long political process of electing a government, in which polls were quoted regularly. These polls were samples of people who were contacted and applying their responses to a larger number of people. Statistical methods were used to determine how large the sample must be in order to achieve trustworthy results and the sample had to be administered randomly to assure that the responses of a few people could comfortably be applied to a larger group of people. The statistical methods also require that a field of possible error be stated (usually stated as + or 3%).
One complaint that natural scientists charge social scientists is that the results can not predict the outcome of a single person in the large group. But this is rather spurious. The natural scientist can increase the heat on a container of air and predict that the pressure inside the container will increase. But that same scientist can not predict the behavior of a single atom, only the results of the atoms of air as a whole.
Thus, while the focus of the two streams is different -physical science studies things; social science studies people – the methods, rigor, and predictability are the same. Since the social sciences is the youngest branch, continuing attention to how to apply the scientific method will soon bring the reputation of social science to the high standards now enjoyed by the physical sciences.