Every research is conducted in one of the paradigms of the research, different paradigms do not mean that the research will have different characteristics. Regardless of any epistemology, paradigm, methodology or research method, every research will have same characteristics. Why? Because these characteristics define research. The data collection, analysis and the presentation of knowledge should be done keeping in mind these research characteristics.
The most important research characteristic is that it is always scientific. It means that every research should be conducted in an empirical manner. The data collection and its processing should be done using scientific procedures. In physical sciences and medicines the researcher uses experiments, observations, and another primary source of data collection to make the research scientific. In social sciences, on the other hand, the researcher might not have as much control and he cannot avoid the subjectivity. The social science researchers make the research empirical by using only valid and reliable sources of data collection.
If we observe around us natural science is very objective in nature. For example, when hydrogen and oxygen in a specific molecular weight react with each other H2O is formed, if the molecular weight of both or only one element is changed the resultant product will be different; when fire is exposed to oil it readily bursts and catches everything that is closer to it; and when water is poured on the ground it always flows toward the downhill or slope. These examples show that in natural science everything has a fixed behavior, given that the conditions are also same. In social sciences, on the other hand, the same condition can have two or more than two meanings, the reason being that social sciences studies human behaviors and behaviors cannot be assessed objectively. This subjectivity can be the part of the research and it is often tolerated. Subjectivity should not be confused with bias which is unforgivable in research. As Kirk, J. in his book Qualitative Research Methods Series states that natural science is strongly linked with its commitment to subjectivity and on the other hand social science is pluralistic in nature. A variety of models may be applied to the same object for various purpose. 
Valid and reliable
The basic aim of every research is to do the research in a valid and reliable manner. It should be realized that, however the methods used to achieve validity and reliability are different in social science and physical science. In physical science, the goal of the researcher is to do the experiments in such a way that if someone else repeats those experiments the result is same. In social science, the researcher achieves validity and reliability in its own way. The research tools, procedures, and techniques should be well tested and they should be relevant to the research topic. Validity alone cannot work in research and every research tool and procedure should be valid as well as reliable.
Control is the most important characteristic of research and a well-executed research must have extreme control or rigor on the research process. In the quantitative research, a great amount of control can be easily exerted by the researcher. Quantitative researchers mostly use experiments and real life observations as a research tool, in experiments as well as in observations the researcher define the independent and the dependent variable. Any other variable that can affect the study are extraneous variables and all such variables are kept under control so they do not ruin the research. In a qualitative research, on the other hand, having control is difficult. Hammersley, 2005 says that to serve evidence-based policymaking we probably need to invent a … myth for qualitative work, that is we too have clear-cut guidelines and criteria, maybe not randomized control trials, but we have our criteria.
It seems sarcastic to say that we will develop a criteria to have a control in qualitative research, qualitative research process is in use from last 50 years, we should have already developed a framework for controlling the question of external influences and subjectivity in research.
Generalizability implies that the research conducted on the basis of a sample is representative of the whole population to which it reflects. The generalizability of the research is highly dependent on how well the sample has been taken from the population. The sample should be highly representative of the population. Taking a large sample size does not always assure that the resultant data will be highly generalizable, sometimes having large sample size makes the research difficult. Larger sample size means greater effort on the part of the researcher, rather the researcher should make an effort to improve the quality of the research by avoiding biases.
Research process should have a systematic, structured and organized plan. This plan should have to be free of any loopholes. It should start with a question or a problem that needs to be solved or answered and step-by-step it should go through each next phase to completion.
- Kirk, J. Miller, M. Qualitative Research Methods Series I, Sage Pub. New Delhi, ISBN 0-8039-2560-3. Pp-12
- Hammersley, Martyn (2005) Close Encounters of a Political Kind: The Threat from the Evidence-Based Policy-Making and practice Movement, Qualitative Researcher. 1: 2–4.