The objective of any research is to establish the truth. According to the subject which is being studied here, the following three levels of sciences can be distinguished: (i) The physical sciences which have their objective as non- living in nature. They include Astronomy, Physics, Chemistry etc. Their application in health includes technology of drugs and tools of diagnosis and treatment. (ii) The biological sciences deal with living things including human beings. Medical science represent an application of these sciences to the protection of health and the cure from diseases. (iii) The social sciences which have human society as their objective include Sociology, History, Economics. Their application to health problems is known as public health.

All three levels of sciences are linked. Disciplines which are situated on a higher level deal with much more complicated systems than disciplines on a lower level. There is also a difference in the levels of sophistication of the different systems coming within the scope of the same group of sciences. Less sophisticated systems are included in the more sophisticated and represent their sub systems. For instance, treatment is a sub system of the control of a communicable disease; sputum examination is a sub system of Case-finding etc. Research in health is being done on the basis of data collected by observation of real conditions or in experiments. The design of the experiment depends upon the complexity of the system under observation and on the possibility of observing one or more factors of the system (variables) after isolation or equalization of the other variables or in their full complexicity. In the field of physical sciences and technology there are very good possibilities for isolation of all variables except those which are under study. This is done in the simple testing of urine for the content of INH, or in similar tests. In medicine and biology it is impossible to isolate all variables without introducing bias; therefore a control group is needed in which all other factors (which are not being studied) are equalized. This is being done in controlled trials laboratory or clinical.

Epidemiologic observations are carried out with different tools; clinical (sputum or X-ray examination, tuberculin test); sociological (interview); physical (measurements), etc. They represent a method which is closer to real life, but they consider only one aspect (sub system) of a community health problem i.e., the trend of the disease without taking care of all other factors which play a role in the determination of this trend, such as economic and cultural variables, behaviour, health services etc. The highest level of research at present represents the systems analysis (operations research). This method has as its objective the study of all variables (or "key variables" at least) which are involved in a public health problem. Such a problem is being simulated by a system of functions (model) and "optimized" (solved) by a computer. The use of mathematics is necessary for the application of these four research methods.