An important aspect of the scientific approach consists of relating questions to evidence. When presented with a controversial issue, investigators, whether in the natural or the social sciences, will look for relevant evidence. In some fields scientists are able to generate observations that provide evidence for use in testing their hypotheses. Experimental sciences such as chemistry and some branches of psychology have an advantage because it is possible for them to produce relevant evidence through controlled laboratory experiments. Other sciences, such as astronomy, cannot do this. They must wait for natural events to produce observations that can be used as evidence in tasting their theories. The evidence that then arises does not come from laboratory conditions under which everything is held constant except the forces being studied. Instead, it arises from situations in which many things are changing at the same time, and great care is therefore needed in drawing conclusions from what is observed. Not long ago economics would have been put wholly in the group of nonexperimental sciences. It is still true that the majority of evidence that economists use is generated by observing what happens in the economy from day to day. However, a significant and growing amount of evidence is now being generated under controlled laboratory conditions.
Social scientists seek to understand and to predict human behavior. Scientific predictions are based on the discovery of stable response patterns, but are such patterns possible with anything as complex as human beings? Does human behavior show sufficiently stable responses to factors influencing it to be predictable within some stated margin of error? This positive question can be settled only by an appeal to evidence and not by armchair speculation. The question itself might concern either the behavior of groups or that of isolated individuals.
