The Design of Experiments

A good theory is a good beginning, but it is only that.  Martin Heidegger once wrote, “It is not surprising that we can question Nature.  The amazing thing is that she answers!”  In this unit, we will examine how to design our questions to give Nature the best chance of responding. 


The ideal way to do this is through an experiment.  But in the social sciences (and even in the natural sciences) we cannot always achieve control over all the variables and randomly assign subjects to experimental conditions (these terms will make sense—indeed, they will become your friends—as you work through this unit).  So we will also study quasi-experimental research design.  These are designs which don’t quite stack up to the rigorous demands of the classic experiment, but at least the specific weaknesses of the design are known beforehand and one can hedge the interpretation of the results accordingly.  Besides, as the old joke has it:  “When do you play in a crooked crap game?  When it’s the only game you’ve got.”  And when do you use a less than ideal research design?  When it’s the only one you’ve got.


Finally, we will consider the ethical values which should also be designed into research. The rights of human subjects must always be protected.  One must consider the principles of beneficence, equity, and respect and the practice of informed consent and confidentiality.


For Further Information:





© 1996 A.J.Filipovitch
Revised 11 March 2005