How to decide between qualitative and quantitative research methods
The nature of 'truth': research paradigms and frameworks
Research should be about discovering 'truth' - but what exactly is 'truth'? It often depends on how one looks at things. It is therefore important as a researcher to understand how you are looking at your research and to be able to explain this to others who needs to know about your research.
Quite generally a way of looking at the world is known as a 'paradigm'. A 'research paradigm' is a 'school of thought' or 'a framework for thinking' about how research ought to be conducted to ascertain truth. Different writers tend use different terminologies when discussing research paradigms, because of where they are coming from. For practical purposes, though, various research paradigms can normally be simplified into just two:
- The 'traditional' research paradigm which is essentially quantitative.
- The 'interpretivist' research paradigm which is essentially qualitative. (The term is due to Denzin and Lincoln (1994), Handbook of Qualitative Research. Beverly Hills, CA: Sage. p. 536.).
This distinction will serve for starters but be aware that there are any number of different research paradigms in the literature and that there is no agreement among academics on how many there are or the finer distinctions between them.
The advantages of quantitative research - but where it cannot work
The traditional research paradigm relies on numerical (i.e. quantitative) data and mathematical or statistical treatment of that data. The 'truth' that is uncovered is thus grounded in mathematical logic.
The traditional research paradigm lends itself to highly valid and highly reliable research. So why do researchers ever use anything else? The reason is that the traditional research paradigm can only be used where the variables that affect the work can be identified, isolated and relatively precisely measured – and possibly, but not necessarily, also manipulated. This is how research in the natural sciences normally operates. Researchers who can work in this paradigm are fortunate because high reliability and validity are held in great esteem. The proponents of the paradigm tend to take its advantages for granted, and theses grounded in it generally take the high reliability and validity as self-evident.
Living beings, however, are affected by numerous interacting variables, such as tiredness, hunger, stress, etc and these variables cannot normally be isolated from one another or measured, and it is certainly impossible, let alone normally unethical, to hold some constant while manipulating others.
Nevertheless the traditional research paradigm can still lend itself to research touched by human and other animate behaviour if the data is numerical and if the sample is sufficiently large for the effects of individual vagaries effectively to cancel one another out. One example could be the performance of school leavers in national examinations across a country over a period of years. Another could be an investigation into the yields of a hybrid crop using large fields of control and experimental plants.
Research set in this traditional research paradigm can answer questions about what is happening and the statistical chances of something happening in the future, but - and this is a big 'but' - it cannot directly answer questions about why something is happening or may happen, nor about the existence of anything else that may be relevant, although answers to such questions may be provided by an established theory within which the research fits.
So the traditional research paradigm is generally not appropriate for research involving small samples of living beings. Then, the variables which stem from individual vagaries and subjectivity do not cancel one another out; neither can variables be readily identified or measured, let alone isolated and held constant while others are varied. Even with a large sample there are sometimes ethical or pragmatic reasons why variables cannot be held constant or manipulated experimentally.
So a different approach is needed and the research has to be set in the interpretivist research paradigm.
What this involves is more like in-depth investigations to establish a verdict in a court of law than experiments in a laboratory. The evidence can be circumstantial and even where there are eye-witness accounts, doubt can always be cast on the veracity or reliability of the observers. A verdict must be reached on what is reasonable, i.e. the weight of evidence one way or the other and on the power of the argument. Data gathered within the interpretivist research paradigm is primarily descriptive, although it may be quantitative, as for example in sizes of living areas, coded questionnaires or documentary analysis. The emphasis is on exploration and insight rather than experiment and the mathematical treatment of data.
Research set in the interpretivist research paradigm can address questions about how and why something is happening. It can also address questions about what is happening in a wider context and what is likely to happen in the future - but it can seldom do so with statistical confidence, because the 'truth' is not grounded in mathematical logic. The 'truth' has to be a conclusion in the mind of a reader (or listener), based on the researcher's power of argument. So different recipients of the research may come to understand different 'truths', just as jurists in a court of law may disagree. It is therefore important for those who use the interpretivist research paradigm to present their work as convincingly as possible. If you are working in this paradigm, your supervisor(s)* will advise you further.
Research students who use the interpretivist research paradigm normally have to do a considerable amount of justification. In contrast, those who use the traditional research paradigm often never even mention it.
Different parts of a substantial research project may well sit comfortably in different research paradigms, but it is important to be able to explain where and how this is so.
Alternative terms for research paradigms which are broadly similar to the traditional research paradigm are: quantitative, scientific, experimental, hard, reductionist, prescriptive, psychometric – and there are inevitably others.
Alternative terms for research paradigms which are broadly similar to the interpretivist one are: qualitative, soft, non-traditional, holistic, descriptive, phenomenological, anthropological, naturalistic, illuminative – and again there are others.
It must be emphasised that the similarities are in broad terms only. Many academics would argue fiercely about the significances of the differences.
You may feel that this page on the nature of truth and research paradigms leaves you, as a student, with a sense of frustration because it does not say more. However the 'more' that individuals seem to want always turns out to be the research methods and research methodology requirements of their own particular programme of research and field of study. That is where the help of supervisors or advisors is invaluable. Fortunately there is no shortage of books on research design, research methods and research techniques appropriate for particular fields of study, and you can readily find out what they are and study a selection. Then, under the guidance of others in your field of study and, in particular, your supervisors or advisors, you should be able to choose meaningfully how to progress your own research and argue for your conclusions.
However it is worth commenting here on the difference between research methods and research methodology. Although some supervisors in some fields of study would argue that they are the same, most academics regard a research methodology as an argument or a case for the methods that a researcher decides to use. In other words, a research methodology states why certain research methods are used. This is just one more example of where imagining that one is a barrister making a case in a court of law orientates to what is needed - see the page on the other roles that research students need to take on during their programme of work.
Research methods and research paradigms
It is also worth pointing out that no research methods or techniques necessarily sit in only one research paradigm. What decides the research paradigm is not a research method, but how the resulting data are to be used.
To take one example, questionnaires can be used to collect quantitative data for statistical number crunching in the traditional research paradigm, just as well as qualitative descriptions of experiences and opinions for use in making a case in the interpretivist research paradigm. So the research method of questionnaires can be use in either research paradigm.
To take another example, precise measurements of, say, the areas of rooms can be used in an argument for more or less space, in which case the data which is quantitative is used in the interpretivist research paradigm, rather than in the traditional one.
It is however true that some research methods are more likely than others to generate data suitable for one or other of the research paradigms.