Cite this page as:
-
Introduction to qualitative methods
-
First published:
on:
Last updated:
If you want to share information about this web page...
This web-page gives you a birds perspective of empirical-holistic (qualitative) approaches to research. You will find lots of links from this web-page to other web-pages providing more information.
You will understand this page best if you first have read Introduction to research.
Ambiguity surrounding the concepts of quantitative and qualitative research, respectively
There are different ways of looking at what constitutes good research. The terms quantitative and qualitative research are used frequently, often as each other’s incompatible opposites. Since the meaning of the words varies, some clarifications may be appropriate. I want to emphasize that there are different ways of viewing this, and I do not claim to present “the truth” here. What follows below is one way (out of several possible ways) of trying to sort out the concepts. Quantitative and qualitative research are sometimes linked to:
- what perception of reality (which ontological assumptions) the researcher has.
- what theoretical (epistemological) knowledge approach the researcher has regarding, for example, empiricism versus rationalism or atomism versus holism. In essence which paradigm the researcher embraces.
- the researcher’s ability to know and handle ontological and epistemological aspects.
- what kind of research questions the researcher is interested in.
- how data is collected.
- what kind of data the researcher has collected.
- what methods the researcher uses to analyze collected data.
- who the researcher’s results benefit, in whose interest the research is conducted.
Can all this be described with the terms quantitative and qualitative research, respectively? Reasonably not. Let us therefore take a closer look at these and some other concepts.
Common concepts
- Ontology is a major branch of philosophy which discuss which world view we choose to adopt (there are several).
- Epistemology is a major branch of philosophy, which, among other things, is about how we go when we develop new knowledge. Epistemology states there are four major types of paradigms (these are shown in a figure below).
- Paradigm (=theoretical frame = perspective). The world map of philosophy of science (figure below) shows four major different paradigms. Paradigms are the spectacles through which we see the world. Each paradigm embraces different methods that share the basic principles of this paradigm. The definition of what constitutes good research is very different between paradigms. The empiric-holistic paradigm, containing qualitative methods, is one of these four major different paradigms in philosophy of science.
- Method (Analytical method). Multiple different methods are used within each paradigm. Within the empiric-atomistic paradigm we have lots of different statistical methods. This page provides an overview of the different methods we have within the empiric-holistic paradigm.
- Data collection techniques (= data collection methods). To avoid confusion between “methods” I prefer to talk about “data collection techniques” rather than “data collection methods”.
- Epirical-atomistisk paradigm (approach) is a more modern label for quantitative approach.
- Epirical-holistic paradigm (approach) is a more modern label for qualitative approach.
The concepts quantitative versus qualitative
As can be seen from the figure, there are four quadrants or main approaches that research can follow. The ontological and epistemological assumptions underlying these four different main approaches are fundamentally different and not directly compatible. Here, the researcher must make a choice (conscious or unconscious). Nothing prevents a researcher from believing that the different underlying, incompatible assumptions can coexist, i.e., one can be in the upper right quadrant in one sub-study, only to jump over to the upper left quadrant in the next sub-study. The choice of quadrant further guides the formulation of the aim and any research questions one is interested in. The aim and research questions then guide the choice of data collection technique and the choice of analysis method to analyze the collected data.
To some extent, this issue is about different perspectives. In each of the quadrants, one views reality differently. You could say that one looks at reality through different kinds of glasses. These glasses are sometimes called paradigms.
One type of glasses, called a holistic worldview, gives the image of a world where everything is interconnected. The whole is more than the sum of its constituent parts. According to this view, one gets misleading results if one demarcates a part of reality and only studies that. Another type of glasses, an atomistic worldview, shows a world that is merely the sum of its parts. According to that worldview, one can isolate a part and study it separately. Most researchers seem to believe that while the whole is indeed more than the sum of its parts, one can often get a decent understanding of reality by demarcating a part and studying it. This assumes that one knows the limitations of one’s methods. This means that one can accept several different paradigms as coexisting in different sub-studies, and in an individual sub-study, the researcher makes a choice as to which of the quadrants is suitable for what the researcher is interested in at that moment. One could speak of an overarching paradigm that encompasses the other four paradigms in a model similar to the figure above. The debate about which kind of glasses distorts the image of reality the least sometimes tends to become unnecessarily polarized.
Where in this reasoning do the concepts quantitative and qualitative fit in? The word quantitative (Medieval Lat. quantitati´vus) is used in technical terms to emphasize measurable properties, e.g., of something to be investigated. Qualitative (Late Lat. qualitati´vus) refers to quality and is used in technical terms to emphasize actual or essential properties, e.g., of something to be investigated. In a strict sense, the terms quantitative and qualitative only denote properties of the data one has collected in a research project. Tradition has, however, given these words a completely different meaning, at times different meanings in different contexts. For this reason, the terms quantitative and qualitative should be used only to describe the data that has been collected, not to denote a certain perception of reality (ontology), view of knowledge (epistemology), or analysis method. When we name the paradigm (the quadrants in the figure above), it is better to use empirical-atomistic approach to describe what many, by tradition, have usually called a quantitative research approach. What many call a qualitative research approach is most often found within the empirical-holistic approach.
Concept analysis, which is traditionally regarded as a qualitative method, can be said to be within the rationalistic-holistic paradigm. With this exception, one can say that all “qualitative methods” are found within the empirical-holistic paradigm.
If we have now established that the terms quantitative and qualitative, respectively, should preferably only be used to denote collected data, how are they then defined? A common definition is that everything that is numbers is denoted as quantitative data, and everything that is not numbers, for example text, images, or unstructured observations, is qualitative data. A problem that should be noted is that the concepts of quantitative-qualitative data are easily confused with the designations quantitative-qualitative variables. These are completely different things. Quantitative and qualitative variables, respectively, are designations within statistics for different kinds of measurement data within projects that are based on the paradigm in the upper left quadrant (empirical-atomistic). Qualitative variables consist of numbers and are therefore quantitative data.
The research process
The researcher becomes interested in a problem. In an ideal world, the researcher should then first decide what kind of world they consider themselves to be in, and after that, what kind of knowledge about the problem the researcher is seeking. Once this is done, one has to decide whether an empirical-atomistic or an empirical-holistic approach is most suitable. (A rationalistic-atomistic or rationalistic-holistic approach could, of course, have been chosen, but this is not discussed further on this particular webpage). After that, the researcher should describe, more concretely and with clear boundaries, which problem one is interested in studying. When the problem is described, one formulates an aim for a new study and, possibly (if one is leaning towards a quantitative approach), one tries to break down the aim into one or more detailed, concrete research questions / hypotheses.
Unfortunately, it is common for many to know how they want to do something, but not why. To begin by designing and describing the method in detail before the problem is described and the choice of paradigm is made is wrong! Another phenomenon, which is not directly wrong but should be mentioned, is that many researchers “get stuck” within one quadrant and only see problems and research questions that fit within that quadrant. Someone who has done a number of studies within the upper left quadrant and learned to handle the reasoning and the statistics tends to then, for the rest of their life, always conduct studies belonging to the upper left quadrant. Researchers who have done several studies within the paradigm in the upper right quadrant tend to then always get stuck on problems that are suitably studied within that paradigm.
Always staying within one and the same paradigm is not wrong, as long as one is simultaneously aware that other paradigms exist. There is a tendency for researchers whose studies are based on the paradigm in the upper right quadrant to be more aware of ontology and epistemology than researchers whose studies are based on the paradigm in the upper left quadrant.
Pinpointing differences between the empirical-atomistic and empirical-holistic paradigms
In a study based on an empirical-atomistic paradigm, it is common to have determined beforehand, after choosing the data collection technique and analysis method, what conceivable conclusions the study might lead to. In a study based on an empiric-holistic paradigm, it is usually the case that one has not, in the same way, predetermined what might emerge. A characteristic of the data collection techniques and analysis methods commonly used in a study based on an empirical-holistic paradigm is that it takes the whole into account in a way that is not possible in a study where one has chosen a data collection technique and analysis method based on an empirical-atomistic paradigm. Studies based on an empirical-atomistic paradigm have the characteristic that one gets an “objective” measure of the probability that the conclusions one has reached are correct (for example, an effect size or a p-value).
Characteristic of studies based on the empirical-holistic paradigm is that one does not know from the beginning exactly what results are conceivable. This requires a fidelity to what one is studying. The choice of method may have to be changed during the course of the project. What one wants to study often deals with qualities and not with numbers, distributions, or exact measurement values. The result may be a number of new aspects of a problem. To describe how patients experience home health care, a study based on an empirical-atomistic paradigm would have defined a number of response options. For example, very dissatisfied, dissatisfied, neutral, satisfied, and very satisfied. After that, one would determine the proportion of patients who provide answers in one of the five predefined categories. In a study based on an empirical-holistic paradigm, no categories would be defined in advance. The investigation would lead to one or more categories, but these would only be defined towards the latter half of the project. The latter type of study can thus describe new qualities of reality that were not known before. The table below presents an overview of some important differences between the empirical-atomistic and the empirical-holistic paradigms:
| Empirical-atomistic paradigm (“Quantitative approach”) | Empirical-holistic paradigm (“Qualitative approach”) | |
|---|---|---|
| Holistic view / Context | Sees the phenomenon being studied as isolated and demarcated / context-free (atomism). The more well-defined, the easier to study. | Looks at the whole (holism) where the context often has significance. |
| Historical connection | Seldom interested in historical connection. | Often interested in history and the relationship past – present – future. |
| Goal | Tries to describe and preferably also explain a relationship or prove a hypothesis. Looks for universal rules. (Control – measure – explain – predict) | Tries to understand the specifically human by getting close and take part in people’s thoughts or behavior. |
| Study protocol | A fixed research plan where sources of error are planned away in advance. All data is collected and thereafter the planned statistical analysis is run. | Fidelity to data. The way of questioning informants and interpreting collected data may need to be changed during the project. |
| Researcher’s role | The researcher remains objective and is interchangeable. | The researcher has a trustful relationship with the informants and is often not as interchangeable. |
| Studies | Studies what can be demarcated and measured, preferably “objectively”. | Studies the specifically human. Often based on perceived experiences and people’s way of giving meaning and significance to the experiences. |
| Selection of “sample” | Representative sample that provides information about an underlying larger population. | Informants chosen deliberately because they have knowledge of the phenomenon. One often acts to get a wide spread (e.g., old-young, sick long time-short time, with complications-without complications, etc) |
| Size of “sample” | Sample size can be estimated in advance through various calculations. | Size / number of individuals included is not as important. Within Grounded theory, one speaks of saturation as a sign that enough data has been collected. A sample size calculation can not be made and is therefore never made. |
| Collected data consists of | Well-defined variables. Their definitions are called operational definitions. | Uses diversity as data. Describes essences, themes, patterns, perceptions. |
| Conceivable results | Predetermined what conceivable alternative results can be obtained. | Openness to what the result will be. Often lead to surprises. |
| Transferability | The results from the sample can usually be generalized to an underlying larger population. | Looks for recurring patterns, common features. Some transferability exists, but not generalizability in the way one has in studies based on the empirical-atomistic paradigm. |
| Good research Quality is | a) Reliability b) Validity c) Reproducibility | a) Perspective awareness (=reported and discussed one’s pre-understanding) b) Internal logic (=used the correct analysis method) c) Good quality of data (=quotes from informants that support results, analysis, and conclusions). d) Legitimacy (=one can follow in the paper how the conclusion was reached). |
Theories underpinning empirical-holistic approaches
Research methods within the empirical-holistic (“qualitative”) paradigm has its roots more or less in philosophy, sociology and anthropology (se figure below). Phenomenology and hermeneutics are example of approaches having a very strong link to philosophy of science and more specifically to the life world perspective first described by Edmund Husserl.
(The figure above is adapted from Evelyn Hermansson at Gothenburg University, Sweden)
Some empirical-holistic methods have a strong link to underpinning theories while others have a less pronounced link and furthermore some, like content analysis, has almost no link. Hence, content analysis is often referred to as a method without underpinning theory.
Empirical-holistic approaches does not use figures and numbers. Hence, they don’t use any kind of statistics, are based on observations (empiricism) and use a holistic view. Hence, it is better to use the term empirical-holistic paradigm (=empirical-holistic approach) rather than the ambiguous “qualitative methods” or “qualitative approach” (see figure).
Research methods used within the empirical-holistic paradigm (“qualitative methods”) is more or less rooted in the philosophy of science. Hence, awareness of philosophy of science is considered more important when using empirical-holistic methods compared to using statistics. Phenomenology and hermeneutics are example of methods with a strong link to philosophy of science. It might be worth mentioning that phenomenology and hermeneutics can refer both to an empiric-holistic method and to a specific philosophical orientation.
Empirical-holistic (qualitative) study design
The most common methods
Empirical-holistic methods can be placed within the empirical-holistic paradigm (see figure above). The exception is concept analysis that falls within the rationalistic-holistic paradigm (see figure above). These methods (study designs) has traditionally been grouped as:
| Approach / method | Focus |
|---|---|
| Case studies | Provide an in depth description of a case and what we can learn from that case. A case can be one person, family, group, community, company, country, etc. |
| Ethnography | Aim to describe the characteristics of a culture (or subculture) and what we can learn from them. |
| Grounded Theory | Aim to develop a new theory around social phenomenons (interactions between people) based on (grounded in) actual observations. What happens and how do people manage? |
| Historical studies | To describe past events to understand the present and anticipate what might happen in the future |
| Narrative studies | Analyse the stories people tell and offer possible interpretations. |
| Life world studies | Describe lived experiences and condensate this to an essence of how this lived experience can be described. Examples are phenomenology and hermenutics. |
| Biographical Study | An in depth description and analysis of the life experience of one person. |
| Action Research Studies | These studies implement changes (actions) and describes what happens without aiming for generalizable truths. Hence, they do not usie randomisation or calculating statistics. |
| Content analysis | Focus on finding (as in inductive content analysis) or verifying (as in deductive content analysis) a pattern among any data consisting of verbal or visual observations. |
| Theoretical Domains Framework (TDF) | TDF aim to identify barriers and enablers to change of behaviour. It is mainly used in implementation research. |
Differences and similarities between various empirical-holistic methods
Most methods within the empirical-holistic paradigm have more in common than what separates them. Most methods in this paradigm are characterized by:
- The researcher should be aware of his preconceptions (= pre-understanding = preconceived opinions = prejudices) and how this may affect data collection and analysis.
- Open data collection.
- Collected data is encoded / transformed into units of meaning.
- Units of meaning are arranged and combined to form one or more categories.
- Often, several categories are combined into one (or more) overall themes / essences / theories / main interpretations. Examples of this highest abstraction level are:
— Phenomenology – Essence
— Hermeneutics – Main interpretation
— Grounded theory – Theory
— Content Analysis – Main category
Although there are many similaritoes there are also distinct differences between various methods within the empirical-holistic paradigm in respect of:
- What kind of knowledge is most interesting? Is it the life-world perspective (as in phenomenology and hermeneutics) or human behavior with possible interactions with other people (as in grounded theory)?
- Should you begin to analyze data during data collection (as in grounded theory) or should all data be collected before analysis begins (as is usually done in phenomenology)?
- Do you think that the researcher, by being aware of their own pre-understanding, can be objective (as in phenomenology) or do you believe that the researcher interprets whether it is the intention or not. You may accept (and even embrace) interpretation if you think it is impossible to avoid this and instead let interpretation drive the process of coding data (as you do in hermeneutics).
Some methods within the empiricalal-holistic paradigm have a clear and well-described method when analyzing collected data. Examples of such method is grounded theory. Other methods are not as well defined, such as content analysis. Within the research methodology phenomenology there are different variants of how to analyze observations. If you say that you have used a phenomenological method, you should also specify in detail how you have done. Sometimes reference is made to previously described analysis models (for example, phenomenological approach according to Giorgi’s methodology).
The chicken or the egg?
It is quite common that researchers have a certain worldview and as a consequence are used to think within a specific paradigm. This may make them prone to stick to aims and research questions that is best answered by methods they are familiar with. A person who is well-placed in, for example, phenomenology is likely to get caught up in new issues that make it possible to choose phenomenology again. Thus, the interaction between aim/research question and selection of appropriate method is not always straightforward. Quite common a researcher choose aim and research question suitable for a specific method rather than the other way around. This is not necessarily bad but it might be good to be aware of the rationale for picking a specific method.
Choosing the best method for your study
The best approach when choosing which method within the empirical-holistic (qualitative) paradigm that suits best for your study is to first first look at the aim of the study. You can do that by looking at the questions posed in the overview below. Based on this choose which method best fit your research question. This is the ideal order of events in the research process. The grouping of empirical-holistic methods below is inspired by Anne Nyström, nurse and doctor of philosophy at Gothenburg University, Sweden.
- Methods focusing on language and communication between humans.
- Content or process analysis* (“What does the choice of words and the organization of the conversation tell us about this group of people?”)
- Ethnoscience (“What can be inferred about people’s world-view from their choice and use of words?”)
- Methods focusing on repeating events / themes / patterns in human life. These methods are looking for patterns describing the lives people live. They focus on human actions or the forms, patterns and structures forming daily life.
- Grounded theory (“What is the pattern describing what is happening…?”)
- Phenomenography (“What are the different ways to perceive…?”)
- Ethnography (“How do people in this culture perceive and manage…?”)
- Content analysis* (What is the pattern emerging around…?)
- Critical research / Feminist Research (“What are the social phenomenons linked to…?)
- (Action research)**
- Methods searching for meaning in people’s experiences or written texts. These methods have a life-world perspective trying to describe how people experience and interpret the world around them. They aim to acquire a deeper description, interpretation and understanding of people’s experiences.
- Hermeneutics (Allows interpretation to be incorporated into emerging patterns of meaning. Acknowledge and allows, even encourage, some degree of interpretation / subjectivity from the researcher.)
- Phenomenology (Recommends not allowing or limiting interpretation to be incorporated into emerging patterns of meaning. Tries to be “objective”.)
- Content analysis* (What is the pattern emerging around…?)
- Life stories
* Content analysis is unfortunately used for two quite different methods. One is a simpler form of method within the empirical-holistic paradigm. The label content analysis is sometimes also used when occurrence of different phenomenon is counted and described with descriptive statistics. This latter form of content analysis resides in the empirical-atomistic paradigm rather than the empirical-holistic paradigm.
** Action research has components of counting numbers and hence is not purely a method within the empirical-holistic paradigm.
Concept analysis focusing on the question “What is the significance / meaning of the concept…?” …is often used in qualitative research. However, concept analysis is not directly an empirical-holistic approach. It lies within the rationalistic-holistic approach (philosophical research).