In textbooks for social scientists, critical rationalism based on the works of Karl Popper, Imre Lakatos and others is presented as the epistemological basis for research. However, many researchers pay little attention to the epistemological aspects of their work. For example, many published studies aim at the verification of hypotheses and theories, ignoring the fact that this is not possible. Some journals even reject articles which mainly present falsifications. This paper discusses possible reasons for and consequences of the unpopularity of proper epistemology in the social sciences.
The reasons include the following:

1. Critical rationalism has been developed for the natural sciences, especially physics. Unlike physicists, social scientists have to cope with an ever changing reality (social life varies between countries, and societies, and over time). Nomothesis is difficult, most theories are of middle range (Robert Merton).
2. Even well corroborated theories on social issues are prone to falsification by irregular observations. There is almost always a surprise during an election campaign or a group of people who do not behave like they are expected to according to sociological theories. Social scientists are confronted with the unexpected to a far greater extent than their colleagues from the natural sciences.
3. An inductive approach (rather than the deductive one demanded by critical rationalism) is inherent to many research projects in the social sciences. This holds especially for non-academic research, e. g. marketing surveys.

All in all, the practice of social scientific research is governed more by epistemological anarchy (as proposed by Paul Feyerabend) than by critical rationalism.
Regarding the consequences for social science methodology, the paper discusses the following issues:

1. Nomothesis is already difficult in the social sciences, but is restrained even more by a lack of attention for the systematic development of theories as proposed by critical rationalism.
2. Induction misleads people to believe in uncertain findings.
3. Social sciences are especially prone to influence by extrinsic interests and ideology. The deductive approach of critical rationalism helps to minimize (or to shed light on) such influences.

The paper eventually discusses measures to improve sensitivity for epistemological issues among social scientists.