Qualitative research
Qualitative research is research which uses only dichotomous data – that is, data which can take only the values 0 (zero) and 1 (one). In psychological research it has come to be defined often as research whose findings are not arrived at by statistical or other quantitative procedures, but this definition is entirely negative, describing qualitative research by what it lacks rather than by what it possesses. Qualitative researchers have, though, described other characteristics of qualitative research which they believe also distinguish it from quantitative research.For example, qualitative research is often said to be naturalistic. That is, it studies naturally occurring phenomena without trying to manipulate them. However, descriptive quantitative research may do the same thing.
Qualitative research is often said to be inductive. This seems to mean that the research is non-evaluative rather than that it depends on inductive logic in the usual sense. However, a reasonable argument could be made that quantitative research is often non-evaluative in the same sense.
Qualitative research emphasizes fieldwork, and this emphasis has been offered as a distinguishing mark. However, quantitative researchers also do fieldwork.
Qualitative research is also described as holistic. That is, qualitative researchers believe in studying phenomena in their entirety rather than concentrating on narrow aspects of the phenomena defined as independent or dependent variables. However, this assertion is questionable. Karl Popper has argued that we cannot know that we are studying the totality of any phenomenon and that consequently we cannot study phenomena holistically. Certainly qualitative researchers have provided few examples of research which attempts to study a phenomenon exhaustively. For one thing, conducting such research would require an ability which no one has, specifically the knowledge of enough disciplines to conduct such research. For example, qualitative researchers do not necessarily assess the medical conditions of their human subjects, but those medical conditions may easily influence their behaviour.
Other proposed distinctions between qualitative and "quantitative" (that is, statistically-based) research are also questionable – that qualitative research is more intense, for example, or that quantitative researchers assume that researchers do not influence their data.
The only valid distinction between qualitative and quantitative research seems to be that qualitative research uses only dichotomous data. Qualitative research typically features assignment of phenomena to categories. Any phenomenon assigned to a specific category can be assigned the value 1 for that category, and any phenomenon not assigned to it can be assigned the value 0. Quantitative research also uses dichotomous data, but uses continuous data as well.
Definitions of qualitative research usually treat it as the opposite of quantitative research, a distinction which has been questioned. Qualitative analysis is at the very least still descriptive research and relevant quantitative psychometric concerns such as its reliability and validity are critical to its utility.
The philosophical bases of qualitative research are found in phenomenology, ethnomethodology, and naturalistic behaviourism. Its research methods are derived from ethnography and anthropology.
In psychology, qualitative research methods include:
- participant observation
- direct observation
- unstructured interviewing
- case studies
- content analysis
- focus groups
- the results of open-ended interviews
- notes of direct observation
- written documents (answers to questionnaires, diaries, program records, and so on)
Most psychological researchers probably use both types of method. In particular, qualitative methods are widely used as exploratory methods; the results of qualitative analysis are used to design quantitative research which tests null hypotheses derived from the qualitative observations.
Nevertheless, many other researchers reject statistically-based research in favour of qualitative research. They argue that statistically-based research is invalid because it ignores context and concentrates on tiny parts of phenomena rather than on the phenomena as wholes. They also argue that quantitative research assumes a unitary reality which does not exist, since every researcher's perception of reality is influenced by his or her unique perceptions and predispositions.
The validity of these criticisms of quantitative research may be questioned. For example, much research on intelligence is conducted without the researchers assuming that they are actually measuring a real entity called intelligence. Instead they are assessing a hypothetical construct, and the value of their assessment will depend on the utility of the construct. On the other hand, many quantitative researchers, and especially many studying intelligence, have represented and interpreted their research as if it was research about an actual entity.
Popper's argument against the possibility of holism is also relevant here. but if the qualitative researchers' assertion is taken simply as an assertion that qualitative research studies more aspects of a phenomenon than quantitative research an empirical assessment of the claim could be made. Quantitative researchers often incorporate large numbers of independent variables in multiple linear regression studies, and often ignore such context as interactions between the independent variables which reduce the stability of their findings. Such studies constitute classic examples of a deficiency which qualitative researchers ascribe to quantitative research, although the adequacy of qualitative methods in analyzing large numbers of variables has not been substantiated. And regardless of the deficiencies of many quantitative researchers, it is still possible to use multiple linear regression judiciously.
Many qualitative researchers reject the traditional psychometric idea of validity – that is, the idea that measures should reflect differences in other logically related measures. Since they deny the existence of an external reality independent of personal interpretation they consequently reject the idea of assessing differences in it. However, many people have concluded that accepting the premisses of this argument leads to the logical conclusion that research of any kind is futile, and other qualitative researchers have developed non-quantitative criteria for assessing reliability and validity.
These criteria, however, are based on debatable assumptions. For example, the traditional psychometric concept of reliability has been rejected by many qualitative researchers in part because some types of reliability require repeated observation, which the qualitative researchers consider impossible in any useful sense. They have proposed a qualitative analog called dependability, which requires researchers to explain how changes in context produced changes in observations.
However, claiming that repeated observation is impossible does not demonstrate that the traditional psychometric concept lacks utility, even if repeated observation is in fact impossible. Furthermore, requiring researchers to explain how changes in context produced changes in observations raises the psychometric issues of reliability and validity again – researchers cannot explain changes with invalid measures (including classification systems), and valid measures must be reliable by definition.
Confirmability is a qualitative concept analogous to the concept of objectivity in quantitative research. It is the degree to which research results can be confirmed by other researchers, while objectivity is the obtaining of identical results by different investigators. A quantitative test is objective, for example, if different testers assign the same scores to the same test-takers. Again, quantitative techniques of correlation would seem to be relevant to the assessment of confirmability.
Transferability has been proposed as a qualitative substitute for psychometric validity. Research findings are transferable to the extent to which they can be generalized to settings other than the one in which they were made. This definition, however, does not imply that conventional psychometric methods of assessing validity are not useful for this purpose, nor that they are less useful than qualitative methods.
The prevailing opinions in psychology are probably that both approaches offer important benefits, that rejecting one or the other means renouncing some of those benefits, and that the most useful debate is about the circumstances in which the two approaches may most profitably be used.
See also qualitative marketing research
