Sunday, April 10, 2011

Phenomenology : The question of validity and reliability


According to Edmund Husserl, the founder of phenomenological research method, the method of phenomenology is as rigrous as any other method and the validity of its results , steps, and the reliability of the method itself are beyond any blemish. However, viewed from the perspective of an objectivist paradigm , these claims do not seem plausible.
What actually we mean by the term validity? Validity actually means that what we are inferring from premises , or what we are concluding from the premises , actually follows from them. That we are well within the freamework of logic when we make a conclusion from premises.
What are the premises? Premises are statements on the basis of which a conclusion is drawn. In the case of deductive logic premises are already established truths or judgments. So, if one concludes , from the premises, All A's are B, Some C's are not A, therefore some C's are not B, then it is quite logical. However , in the case of inductive logic when the premises are to be formed from the particular cases, validity is to be guaranted in a different manner. Here quantitative methods or the method of natural sciences provide us with different means on ensuring validity. However, the logic remains same , and validity means , in almost all cases two things. First it is valid to draw the derived conclusion from the data, and that it is valid to generalize the drawn conclusion over other similar cases.
In quantitative or scientific method falsification of a theory means that it failed to explain one of the instances which it was supposed to explain. So, the theory , or the conclusion no longer remains valid, and both its internal and external validities are challenged. Internal validity means that the conclusion followed from the data and external validity means that it was right to generalize it over the similar cases. Thus, in scientific method , objectivity , in the form of availability of the chance of empirically varifying the results of a theory, or the testing process of a theory, provide with an oppurtunity to validate a theory.
In phenomenology objectivity is a bias, for phenomenology suspends belief in the objective world, it studies the subject and keeps its focus on the subject alone while turning away completely from the object. So, if a researcher establishes something about the inner experience of a person, or a group of people, that conclusion cannot be validated objectively.
So, how phenomenological research can be validated or can be considered as valid? In ensuring validity of phenomenological research we have to proceed stepwise. So, each step has its measures of validity. Let us see these measures one by one.
Validity
Sampling
Sampling should be logical, and sample should be purposefully drawn.
Data Collection
During data collection, the researcher has to excercise epoche' or bracketing. A researcher has to completely suspend his/her beliefs.
Data Analysis
During data analysis themes are to be selected as they emerge from the data. Although a researcher can thematize or focus a particular dimension at his/her discretion, yet there should be a logic for doing this.

Eidetic Reduction
Eidetic reduction for determining essences, should be carried without any biases. One has to imaginatively vary the example to see its possibilities and then from these possibilities one has to separate that which remained invariable in all the possibilities.

So, phenomenological research process establishes its validity in each of the steps. In making generalization one has to prove that what is being generalized is actually generalizable.

Reliability
Reliability of this method is to be ensured through clearly outlining and defining the steps involved. Thus, whoever wants to replicate the steps on the data , can easily do so to find out how reliable the proceedure was. Moreover, if there are certain generalizations made through the research, other researcher, through following the same proceedures should be able to find the same generalized conclusions.

No comments: