Dear Prashant, on the topic you're interested in, I find this textbook very helpful:

Dunning, T. (2012). Natural experiments in the social sciences: A design-based approach. Cambridge University Press.

Hi. Due to the need for replications in experimental research designs which increase by more elaborated designs, use of these research designs are limited in social sciences. Theoretically there is no limitation to use ERD in social sciences.

Hello Prashant Ji

Before starting your research, i feel that you need to be clear that what tool you will be using, who will be your respondents/ participants? What is your role in research? Will you be biased researcher or bias free researcher? What will be the reality for you? If if answer such questions then I think you will get a clear way out to carry on your research. Apart from these consultation with your supervisor and taking reference from books might help you.  

One has to really look for the objective and the principals of design - Randomisation, Replication and Local Control. The selection of subject will be most important consideration to use DOE. I think this can be applied. Various design such as CRD, RBD, LSD with factorial experiments has potential for the same.

Using rigorous experimental research design such as randomized control trial in social science is an operational challenge because selection of an appropriate comparison or control group is practically difficult. Under the circumstances one may have to use quasi-experimental research design. It will depend upon the specific hypothesis that one is interested to test by conducting research study in the field of social science.  

Experimental research designs,when done accurately ,minimising threats to validity is more reliable in reports and results.It can be used in case of behavioral testing expecially in observation of normal  or mentally/socially/psychologically challenged persons.There is no scenerio where you cannot experiment once you know the right stimuli to apply in treatment .It is not restricted to pure and applied sciences alone.For example to test for change in behaviour following treatment in one group and withdrawal of treatment in the other equivalent or non-equivalent group

The following book addresses your question:

Keith F. Punch (2005). Introduction to Social Research–Quantitative & Qualitative Approaches. London: Sage

I recommend you have a look at the studies of Harold Garfinkel. His crisis experiments may lead you to a way of developing a method for your question of research

Consider Pre-Post design with control group. One Group subjected to Treatment and the other without it, provided they are similar at baseline and there is no effect of other things on the outcome variable of interest.

Hi Dr Prashant,

There are many use cases for experimental research design in social science according to one's research problem & objective(s) - you can find examples from the above researchers' links or merely from Internet search. 

You might also refer to the following 2 links on uses cases that I'd prepared:

https://www.researchgate.net/publication/266201759_Impact_of_Project_Management_Skill_on_Academic_Achievement__An_Experimental_Study_on_Malaysia_Form_4_Students

https://www.researchgate.net/publication/266201770_Impact_of_Mind_Mapping_Technique_on_Academic_Achievement__An_Experimental_Study_on_Malaysia_Form_4_Students

Regards,

Fung

Article Impact of Project Management Skill on Academic Achievement –...

Article Impact of Mind Mapping Technique on Academic Achievement – A...

Clinical trials typically employ the statistical principles of experimental design in randomized controlled experiments. 

Experiments in the social sciences have a long and important history – that history is very well covered and argued for in

Oakley A (2000) Experiments in Knowing: Gender and method in the social sciences. Cambridge: Polity Press. (New York: The New Press.)

And the same author puts the argument pithily in

Oakley A (2004) Who’s afraid of the randomized controlled trial? Some dilemmas of the scientific method and ‘good’ research practice’. In: Seale C (ed.) Social Research Methods: A reader. London: Routledge, and pp. 510-518.

http://www.brown.uk.com/teaching/HEST5001/oakley3.pdf

Currently there is a great deal of interest in randomized experiments and quasi experiments across the social sciences – for example in political science

 Rebecca B. Morton, and Kenneth C. Williams (2010) Experimental Political Science and the Study of Causality From Nature to the Lab

http://www.cambridge.org/gb/academic/subjects/politics-international-relations/research-methods-politics/experimental-political-science-and-study-causality-nature-lab

 The Cambridge Handbook of Experimental Political Science

http://www.cambridge.org/gb/academic/subjects/politics-international-relations/research-methods-politics/cambridge-handbook-experimental-political-science

 and a new journal

Journal of Experimental Political Science

 http://journals.cambridge.org/action/displayJournal?jid=XPS

Experimental research is applicable in social research when determining a causal relationships among variables. You must select the experimental group( where tou can apply treatment) and the control group. Selection of subjects in this this group is simply by randomisation i.e by chance. This will enable you to control certain variables that may affect the experiment since the subject are selected simply by chance. The effects of the treatment on the experimental group can then be compared to the control group. The independent variables that causes the effects on the dependents variable can then be determined holding other extranous variables contant.

hi

here are the publications:

http://psycnet.apa.org/psycinfo/1993-97950-000

http://moodle.technion.ac.il/pluginfile.php/219643/mod_resource/content/0/Campbell_and_Stanley_1963.pdf

http://books.google.co.th/books?hl=en&lr=&id=9yurQt7T65oC&oi=fnd&pg=PA1&dq=experimental+research+design+in+social+science&ots=noRc_5wM9n&sig=A2zWY0CQtuJJYvt0JeGOXr0QLFM&redir_esc=y#v=onepage&q=experimental%20research%20design%20in%20social%20science&f=false

http://books.google.co.th/books?hl=en&lr=&id=XvMAYYvS1rEC&oi=fnd&pg=PT1&dq=experimental+research+design+in+social+science&ots=hNumGQ3iRl&sig=LYosBV-wf_ebgO6xDtoBdG-9vxk&redir_esc=y#v=onepage&q=experimental%20research%20design%20in%20social%20science&f=false

Dear Prashant,

Here are my talks on this topic:

https://www.youtube.com/watch?v=LASRBoalcfo : Qualitative and quantitative research ( 1)

https://www.youtube.com/watch?v=Nzb-6aOHxmk  Qualitative and quantitative research ( 2)

Dear Prashant,

As like sciences, experimental research design has wider applicability in social sciences too. The practice of experimental research has mainly seen in the disciplines like Social Work and Psychology. A large volume of such studies has already been conducted throughout the globe. As per the nature of circumstances and context of the study, experimental designs may vary. (Pre test and post test design, ex-post facto design etc..)

You may follow the below link for one of such study I have conducted.

https://www.researchgate.net/publication/308804131_Positive_Mental_Health_through_Life_Skill_Education_Empowering_Adolescents_having_Psychosocial_Problems

Article Positive Mental Health through Life Skill Education: Empower...

You may follow the link below:

Article The experimental method in the social sciences

Experimental research in science expects the observations to be independent of the observer and it expects that all factors would be known. In analyzing the observations and from that deriving the cause and effect relationship the experimenter is able to isolate the one parameter that has been altered and therefore be able to see the change it brings out. Translating that in to social sciences in the same deterministic manner is fraught with dangers. The subjects of the study are not invariant, the conditions are are never replicable and the observations are far less objective. Even the randomization exercises assume similarity of communities and an understanding of the entire act and an ability to maintain interventions such that only one or two factors are altered. It also assumes an understanding of the significance of the intervention and the manner it can be most effective. Often science and experiments have an hierarchical superiority in terms of the results emanating and that can lead to misleading knowledge and wrong application. We need to design the studies with a lot of ground work and effort and even then respond to the inferences from experimental methods carefully.