CUTTING THROUGH THE CONFUSION

Much of what you will find if you search for information on the topic of research paradigms is very confusing. This paragraph explains why this is so, and what to do about it.

The concept of paradigms which we use in discussing research is derived from philosophy. Unfortunately the same esoteric approach to understanding paradigms is carried through, inappropriately, to most discussions of research paradigms, along with Greek technical terms (such as ontology and epistemology) which serve to confuse rather than illuminate.

Consider the two predominant paradigms: positivism and interpretivism. Put simplistically (and perhaps not entirely accurately) these arIse from two contrasting philosophical views of the world, as exemplifed by the following example.

The positivist view is: This is a rock. It is a rock. Even if there is nobody here to see it, it remains a rock.

The interpretivist (also known as constructivist) view is: There is no such thing as a rock per se. It is only a “rock” because we see it as a rock and ascribe the meaning “rock” to it. Nor will everyone ascribe the same meaning to it. All meaning is constructed: it depends on how you interpret your senses.

THREE COMMON RESEARCH PARADIGMS

Positivism

As researchers, we do not have to adopt a single viewpoint to explain everything in the world. We can leave such debates to the philosophers. Most of us come from a positivist viewpoint: in other words we really do believe that there is such a thing as a rock, quite independent of people seeing or feeling it and constructing a meaning for it.

This carries over into the positivist research paradigm which accounts for the large majority of biomedical science. This is science conducted on the premise that a precise answer to our research question exists (and has always existed), and it is just up to us to find out what it is. For example, let us imagine that we wish to know the exact value of the mean number of fingers per hand of the people in the city. (We know it will not be exactly five, as some people will have been unfortunate enough to have been born without fingers or to have suffered an amputation.) Our research precedes on the following assumptions:

• We believe that an answer (for example 4.99993) exists, whether it is researched or not.
• The answer can be stated as a number, with great precision, if our research technique is good.
• The answer does not change except for obvious reasons of cause and effect, such as people being born, dying or suffering amputations. If we know these numbers then we can recalculate the answer to the original question
• Another researcher, using a similarly good research technique, will arrive at essentially the same answer, since only one true answer exists.
• If two researchers arrive at slightly different values (for example 4.99997 fingers per hand versus 4.99985 fingers per hand), we will assume that the variation is due to the limits of sampling and methods, and is not because there are actually two different answers, one for the first researcher and another for the second.

These are the premises and consequences of what is known as logical positivism. Research within the positivist paradigm is frequently quantitative, requiring counting, sizing, mathematical operations, statistical treatment etc. Positivism underlies chemical and biological laboratory experiments, animal experiments, clinical trials, meta-analyses and many others. All are based on the concept that every research question has an invariable correct answer, just waiting to be determined.

Interpretivism

Now consider the following situation. In the finals of a national football championship, Kaiser Chiefs beat the Moroka Swallows (these are two popular South African football teams) 1-0 in a penalty shootout. The question is: Is this a triumph or a disaster?

Clearly, in this case there is no single “true” answer waiting to be found. If you are a Chiefs fan, you will celebrate the outcome as a triumph. Conversely, if you are a Swallows fan (and more so, the coach), you will see it as a disaster. Nor are these answers fixed. If you are not particularly attached to either team, your answer may well depend on your present company or on the person asking the question. Furthermore, even a committed fan of one team may change their mind, for example, during the course of discussion with fellow participants in a focus group interview.

This is the basis of interpretivism:

• There is no single correct answer, independent of people, waiting to be found.
• The answer is entirely dependent on how people interpret a situation.
• The answer may even change as the research proceeds.

Interpretivism is also known as constructivism, in reference to the fact that people construct meaning for what they experience.

As researchers we recognise that there are questions which lend themselves to research within a positivist paradigm, (e.g. What is the average number of fingers per hand in the population) while other questions are only answerable within the interpretivist paradigm (e.g. What is it like to live with missing fingers?). Interpretivist research uses different methodologies to positivist research. Most of these fall under the rubric of qualitative research.

So, in conclusion, as researchers we are not required to decide which philosophical school of thought is “correct” when it comes to the world in general. All that is required is that we understand that some research questions are best approached from a positivist point of view (very frequently requiring some form of quantitative, number-driven research), while others are best approached from the interpretivist point of view, best answered by some form of qualitative research. This is the basis of the pragmatic paradigm.

Pragmatism

This paradigm expressly recognises that research need not follow an either/or approach, but makes best use of each paradigm, as dictated by the research question. The research methodology is often described as “mixed-methods“.

COMPARISON OF THE PARADIGMS

The differing nature of the “reality” sought by research within each paradigm cascades down through multiple characteristics of the research carried out within each. These are summarised in the following table.

Next: Logic and scientific reasoning