Saturation in qualitative research

And we're back to talking about qualitative research! Sorry not sorry.

What is saturation?

Saturation is commonly used as a determinant of sample size in qualitative research. So what is saturation? It is usually thought of as the point where there is no new codes, or no new data for development of themes, or where the complete range of theoretical constructs is fully represented by the data. There are four main models of saturation, and sometimes more than one model may be combined:

1. Theoretical saturation- Uses the development of categories and theory as criteria for additional data collection. Sampling is guided by the necessary similarities and contrasts required by the emerging theory, in a process known as theoretical sampling.
2. Inductive thematic saturation- Focuses on the identification of new codes or themes and on the number of these themes, rather than their completeness.
3. A priori thematic saturation- Collects data in order to exemplify a theory, rather than to refine or develop theory. Points to the idea of pre-determined theoretical categories.
4. Data saturation- Stops collecting when nothing new is found. Relies on simply identifying redundancy, without referencing theory at all.

Where and why should we seek saturation?

Saturation may have different relevance and meaning in different types of research. In a more deductive study (using theory to analyse data), the concept of saturation may help you to determine how well the pre-determined codes are represented. In a more inductive study (using data to generate a theory), the concept of saturation might involve how well the new codes and themes are identified. On the other hand, saturation might not be a very relevant concept in some other types of research, such as narrative research.

When and how can we achieve saturation?

The question of "when is saturation achieved?" generally depends on the model of saturation chosen. It is important not to decide on saturation too early on, as this may produce superficial results.

Alternatively, instead of asking, "When is saturation achieved?" or "Has saturation occurred?" (i.e. viewing saturation as a process), it may be more appropriate to ask, "How much is enough?" (i.e. viewing saturation as an endpoint). The problem with the "endpoint" model is that looking for a saturation "point" often requires making predictive claims on data that you don't have yet, as you're basically assuming that all of the other potential data out there is just going to be a rehash of what you already have. In fact, some studies consciously sample beyond the saturation "endpoint," just in case. Using a "process" model requires using ongoing judgement to figure out the point of diminishing returns with regard to sampling and discovery of new insights.

References

Saunders, B, Sim, J, Kingstone, T, Baker, S, Waterfield, J, Bartlam, B, Burroughs, H, Jinks, C 2018, 'Saturation in qualitative research: exploring its conceptualization and operationalization', Quality & Quantity, vol. 52, no. 4, pp. 1893-1907.

Aetiology and epidemiology of cerebral palsy

Yep, it's another post about the development of cerebral palsy (see here for more). Since we don't know the "true cause" of cerebral palsy, this post will mainly discuss different risk factors.

Brain damage in cerebral palsy

The cause of brain damage in cerebral palsy is thought to be a combination of cytokine-induced periventricular leukomalacia (PVL) (which I have explained here) and ischaemic damage (possibly caused by asphyxia), complicated by later intraventricular haemorrhage and hydrocephalus (fluid in the brain).

PVL usually affects the brain between 20-32 weeks, and while often associated with diplegic cerebral palsy, it is not exclusively associated with diplegia. PVL may also occur in infants with sepsis, or in young children treated with alpha-interferon for giant haemangioma (tangle of blood vessels).

Acute intrapartum asphyxia is thought to only be involved in 2-10% of cases of cerebral palsy. Some fetuses may already be damaged before entering labour, but a difficult labour may increase the amount of damage. Cooling the baby to reduce metabolic need may reduce early neurodisability.

Birth weight

Birth weight is a risk factor for cerebral palsy. Babies that are small for gestation age are more likely to be damaged by hypoxic-ischaemic events, while babies that are large for gestational age are at a higher risk related to maternal diabetes and obstructed labour. Symmetrical growth retardation (underweight, short, small head circumference) also increases the risk of developmental delay, even when asphyxia does not occur.

It has also been suggested that intrauterine growth restriction, which may result in babies being small for gestational age, might be a result of the brain damage that causes cerebral palsy, rather than being a cause of the brain damage. After an insult to the brain, intrauterine growth restriction of the brain may help to preserve or divert energy to critical areas.

Socioeconomic factors

Cerebral palsy is more prevalent among those of a lower socioeconomic class. It is unsure as to why this may be the case, but hypotheses include possible nutritional deficiencies and/or infections.

Multiple births

Multiple births (twins, triplets, etc.) increases the risk of cerebral palsy. This may be due to a variety of factors, such as twin-to-twin transfusions or in-utero death of a twin, and the increased risk of vascular anomalies of the placenta, prematurity, small size for gestational age (see above), premature rupture of membranes, and/or hypoxia.

Genetic factors

There is some evidence suggesting at least a partial genetic basis for cerebral palsy. There is an increased risk in twins, particularly in monozygotic twins. There is also an increased risk of a sibling has cerebral palsy. Cerebral palsy often appears in familial "clusters" where several members have similar types of cerebral palsy, and there is a greater risk in cosanguineous ("inter-breeding") families. In fact, some families with multiple affected children have been found to have some kind of autosomal recessive disorder, but this only accounts for a very small number of childrren with cerebral palsy.

It is thought that the genetic factors underlying cerebral palsy might impact the intrauterine environment, rather than the fetus itself. Some of the genes that have been investigated are associated with polymorphism in IL-6 and mutations in the prothrombin gene.

Malformations

Up to 12% of children with cerebral palsy have some kind of brain malformation, such as hydrocephalis (fluid in the brain) and microcephaly (small brain). Some of these malformations may be acquired, others genetic. People with cerebral palsy also have a higher likelihood of malformations in the heart, palate, eyes and limbs as compared to typically developing children. (Interestingly, these malformations are more common in people with cerebral palsy who were born at full term, as opposed to preterm.) Some of the brain malformations in cerebral palsy may be due to genetic factors which produce disorders of cortical migration, such as lissencephaly ("smooth brain," i.e. a brain without the usual bumps and folds).

Cerebrovascular events

Cerebrovascular events may also contribute to the development of cerebral palsy. In the 1st and 2nd trimester, the most common type of cerebrovascular event is cortical migration disorder (which is also seen in congenital cytomegalovirus infection- hmm, maybe that's why the Cerebral Palsy Alliance keeps posting about CMV on Facebook...). In the 3rd trimester, the most common type of event is a discrete infarct. Maternal and fetal thrombotic or haemorrhagic disorders may also contribute to these types of events, and ultimately to cerebral palsy.

Placental pathology

There is no single placental pathology associated with cerebral palsy. Furthermore, the extent of damage is important as the placenta has around 30% spare capacity before fetal growth is affected.

Chronic inflammatory placental lesions are one type of pathology that is associated with cerebral palsy. These lesions are associated with PVL and diplegic cerebral palsy and are likely linked with IL-1, IL-6, and TNF-alpha, which damage developing oligodendrocytes. (Such cytokine-induced white matter damage may also be seen occasionally after neonatal meningitis.)

Maternal genitourinary tract infections (e.g. Chlamydia, Trichomonas, UTIs, etc.) during pregnancy may also increase the risk of cerebral palsy, particularly if these infections occur during the first or second trimesters. Gestational age and birth weight are protective factors against such harm. Paradoxically, treating these infections might even increase the risk of cerebral palsy in some cases, perhaps by enhancing cytokine responses, but this is still uncertain.

A note on similar disorders

There are many disorders that present similarly to cerebral palsy, but are actually something different. Some of these disorders, unlike cerebral palsy, have known causes and are treatable. For instance, some disorders of dopamine metabolism may present similarly to cerebral palsy, but can be treated.

References

Eunson, P 2012, 'Aetiology and epidemiology of cerebral palsy', Paediatrics and Child Health, vol. 22, no. 9, pp. 361-366.

Self-determination theory and motivation

Motivation is a state of energy, direction, and persistence. When motivation is authentic, it can boost interest, excitement, and confidence, all of which can in turn boost performance, persistence, creativity, vitality, self-esteem, and well-being. As such, many researchers have been interested in factors relating to motivation and how it can be improved. Self-Determination Theory (SDT) is one such approach to motivation. SDT highlights the importance of humans' inner resources and centres on three main needs of competence, relatedness and autonomy. Let's take a look at aspects of SDT.

Intrinsic motivation

Intrinsic motivation is a natural inclination towards mastery, exploration, and interest. People will only develop intrinsic motivation if an activity holds intrinsic interest for them. There is a lot of variability in intrinsic motivation, which cognitive evaluation theory (a subtheory within SDT) aims to explain.

Cognitive evaluation theory primarily looks at social and environmental factors behind the variability of intrinsic motivation. Cognitive evaluation theory assumes that intrinsic motivation will flourish if circumstances permit, but can be enhanced by other factors. For instance, social-contextual events that conduce feelings of competence can enhance intrinsic motivation, as long as there is also an internal perceived locus of autonomy present. Intrinsic motivation is also more likely to flourish in contexts characterised by security and relatedness.

Self-regulation of extrinsic motivation

Just like how cognitive evaluation theory looks at intrinsic motivation, a second subtheory in SDT, organismic integration theory, looks at extrinsic motivation.

There are many types of extrinsic motivation that can be more or less placed on a scale from least to most intrinsic. At the low end of the scale is amotivation, which is a complete lack of motivation, which may be due to not valuing the activity, not feeling competent, and not expecting the desired outcome (i.e. "why bother trying if it won't work?"). The next type is external regulation, where the action is only performed to meet an external demand, and thus comes with an external perceived locus of causality (e.g. "I'm only going to tutorials to get participation marks.").

As we move through the different types of extrinsic motivation, they become less extrinsic and more intrinsic. Introjected regulation refers to taking in regulation, but not fully accepting it as one's own. For instance, someone might want to avoid guilt or anxiety, attain pride, and so on. Introjected regulation mainly works via contingent self-esteem, and is related to increased effort and anxiety, but a decreased ability to cope with failure. Like external regulation, introjected regulation is also associated with an external perceived locus of causality.

Regulation through identification is the conscious valuing of a goal, so the action is accepted or owned as being personally important. Unlike introjected regulation, regulation through identification is related to increased interest and enjoyment, as well as more positive coping styles.

Integrated regulation is the most intrinsic type of external motivation (i.e. as close as you can get without it being fully intrinsic motivation). Identified regulations are fully assimilated to the self, but ultimately actions are still performed to attain separable outcomes rather than for inherent enjoyment. Hence, integrated regulation is still not fully intrinsic.

Facilitating integration

Greater levels of internalisation and "intrinsic-ness" are related to increased behavioural effectiveness and persistence, as well as increased assimilation within a social group. Sounds good, right? So, how do we encourage internalisation?

To answer this question, SDT focuses on the three needs that I introduced in the opening paragraph: relatedness, competence, and autonomy. Relatedness is of central importance; after all, the main reason why actions are performed is usually because said actions are prompted, modelled, or valued by significant others. Competence and autonomy should also facilitate internalisation. It may also be important for people to grasp the meaning of an action and synthesise it with their other goals and values. On the flipside, if motivation is diminished, it is important to check to ensure that the needs for competence, autonomy and relatedness are not being somehow thwarted.

Psychological needs and mental health

Relatedness, competence, and autonomy are not only important for motivation: they are also important for overall health and well-being. There are many methods in which these needs can be satisfied, and these methods may change over time and culture. The relative importance of each need may also change over time and culture. Well-being will improve so long as these needs are being satisfied, no matter which methods are used to satisfy them.

A focus on intrinsic vs. extrinsic aspirations may also affect well-being. A strong importance on intrinsic aspirations is associated with increased well-being, self-esteem and self-actualisation, as well as decreased depression and anxiety. The opposite is true for a focus on extrinsic aspirations. A person's satisfaction in their life roles has been shown to be attributable to the degree to which the role supports authenticity and autonomy.

References

Ryan, RM, Deci, EL 2000, 'Self-determination theory and the facilitation of intrinsic motivation, social development, and well-being', The American psychologist, vol. 55, no. 1, pp. 68-78.

Ethics in qualitative research

Just like in many forms of quantitative research, there are many ethical dilemmas that you may encounter in qualitative research. This post will start by talking about some different ethical positions before discussing these ethical dilemmas.

Traditional ethical positions

Two of the most commonly-seen traditional ethical positions are utilitarianism and principlism. Utilitarianism, or utilitarian consequentialism, emphasises consequences over intent and aims to choose the action that produces the greatest good for the greatest number of individuals. Principlism, or the duty ethics of principles, judges actions by their intent, rather than by their consequences.

Alternative ethical positions

Virtue ethics

Virtue ethics considers ethical behaviour to be contextual or situational. It emphasises the role of phronesis, or cultivated practical wisdom. In virtue ethics, the researcher's own moral values and skills are emphasised.

Feminist ethics

Feminist ethics, as the name implies, focuses on feminist-informed social values. It emphasises care and responsibility, and the need to expose the diversity of realities. Research informed by feminist ethics aims to be collaborative and participatory.

Culturally responsive relational reflexive ethics (CRRRE)

CRRRE emphasises the need for flexibility as researchers cannot fully understand the perspectives of other cultures. It attempts to understand and affirm other cultures by valuing the dignity, mutual respect and connectedness within and between cultures.

Ethical dilemmas in qualitative research

Informed consent

As qualitative research can be open-ended and a researcher's direction may change somewhat based on new information, it is important to have an ongoing process of construction and negotiation, rather than simply a one-off informed consent form.

Anonymity and confidentiality

It may be difficult or impossible to give absolute confidentiality, especially since details may unintentionally reveal a participant's identity or provide a mistaken identity. Also, some participants may prefer to be named, depending on the research.

Research as therapy

Some participants may find it therapeutic to talk about their experiences. However, it is very important to set boundaries, and clarify the purposes and role of the researcher as compared to therapists. If you are dealing with a sensitive topic, it may be a good idea to make a list of other support services that you can refer to, should the need arise.

The Internet

Internet research has ethical dilemmas of its own. For instance, if you are looking at message boards, blogs, and the like, you may need to think about the need for informed consent from message board users, and so on. You may also need to pay attention to copyright laws. For more information, check out the ethical guidelines from the Association of Internet Researchers.

Protecting the researcher

The researcher may need support and debriefing themselves, especially if they are researching a difficult or sensitive topic.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

Evaluating qualitative research

Even though qualitative research is often seen to be more loosely-defined, it is still possible to evaluate its quality, even though appropriate quality measures may differ slightly between qualitative and quantitative research. Let's take a look.

Reliability vs. Dependability

Reliability, or "repeatability," is important in quantitative research, hence the need for quantitative researchers to provide enough information so that their study can be replicated. Reliability, however, is not necessarily a good measure for evaluating fieldwork, as what people might say in an interview one day might be different to another day. Furthermore, you may get slightly different responses and results depending on the participants. In other words, you cannot step into the same stream twice. A slightly different, but more appropriate metric, is dependability, which represents the stability of the data over time.

Objectivity vs. Confirmability

As there is subjectivity required in processes such as coding, qualitative researchers cannot really strive for complete objectivity. Instead, confirmability may be a better metric. Confirmability assures that the data, interpretations and outcomes are rooted in the context and people studied, rather than in the personal biases of the researchers. In order to strive for confirmability, the decision-making trail, purposes and assumptions should all be made explicit. It may also be helpful to have a supportive but "critical friend" to be a sounding board as you interpret the data.

External validity vs. Generalisability

Generalisability is qualitative research's answer to external validity. Just like external validity, generalisability means that the findings should be applicable to other settings, and theories should be able to be expanded and generalised. Qualitative research also aims for generativity, which refers to the ways in which the work enables someone to see or act upon phenomena. Unlike in quantitative research, where external validity is usually judged by concrete factors such as sampling methods, the responsibility for judging the generalisability of qualitative studies rests on the readers of the study.

More about validity...

Validity refers to the degree to which the researchers' claims correspond to reality. Valid studies measure what they intend to measure. There are several different approaches towards the concept of validity in qualitative research.

• Parallel perspective / transactional notion of validity- This approach focuses on establishing a match between the realities of the researchers and the researched. Techniques used include triangulation and member checking (allowing the participants to read the transcripts and results and provide feedback).
• Diversification perspective / transformational approach- This approach focuses on "catalytic validity," or the degree to which the research process energised participants and gave them a better knowledge of reality. Ultimately, this approach focuses on whether the research made a difference.
• Letting go perspective- This perspective abandons the notion of validity completely and focuses instead on characterising traits of the research. A few sets of criteria for looking at these traits have been devised, though there is also the option of developing the criteria while engaging in inquiry.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

Presenting the results of a qualitative study

Once you have performed your analysis, it is time to present it! There are many ways to present qualitative data, from the traditional ways (e.g. writing a report), to more creative (e.g. artistic performances). This post will only cover a few of them.

General tips

When presenting your qualitative study, you should describe and justify your methods, from sampling to data collection and analysis. You may want to consider validity criteria, depending on your topic. When presenting your results, it is usually a good idea to interweave participants' quotes with the your observations, so that participants' and researchers' voices can be balanced.

Traditional or realist tales

Traditional or realist tales aim to foreground the voices of participants, as participants are seen to have "experiential authority."

Confessional tales

Confessional tales are highly personalised tales that foreground the voice and concerns of the researcher. They take a reflexive stance and reveal the research process itself.

Autoethnography

In an autoethnography, researchers tell stories about their own experiences, aiming to relate their personal experiences to cultural experiences.

Artistic methods

There are many artistic methods that have been used to represent qualitative data. These include, but are not limited to, poetic representations, ethnodrama, ethnographic nonfiction, and musical performance.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

Analysing qualitative data

Once you have collected your data, you need to analyse it! I have already touched on two types of qualitative analysis (content and thematic analyses) in an earlier post, but there are many more types of analysis than just those two. In this post, I will talk about some of the other types of qualitative analysis, but this list is by no means exhaustive.

Hierarchical content analysis

Hierarchical content analysis, like most other types of qualitative analysis, requires some kind of coding. In this type of analysis, codes are usually short, simple and precise key words. These key words are chosen to represent segments of primary content and "essence" of the data. Once codes have been identified, the researcher then identifies patterns and the way that the patterns interplay with one another, make comparisons and contrasts, divide the data into categories, and so on.

Hierarchical content analysis can be summarised in six main steps:

1. Immersion
2. Searching for, identifying and labelling themes
3. Connecting and ordering themes
4. Cross-checking to verify that all themes and categories are represented
5. Confirmation (possibly by another investigator)
6. Presenting results (usually in table form)

Grounded theory analysis

In grounded theory analysis, researchers generate theory from the data, rather than imposing theory on the data. Grounded theory analysis emphasises analytical techniques, such as constant comparison, in which emerging data is interpreted in relation to data that has already been collected. There are various kinds of grounded theory analysis, such as Glaserian, Straussian, and constructivist. Below are the main steps in Straussian grounded theory analysis:

1. Open coding with short, specific codes. Intense line-by-line coding is performed. Concepts are examined in terms of their properties and arranged. During this stage, it is important to remain open-minded.
2. Axial coding. Categories are related to sub-categories and a core category (or perhaps categories) is identified. During this stage, researchers may develop a conditional matrix, which visually depicts the relationship between concepts and conditions under which they occur. Researchers using the "paradigm model" may focus their attention on the context and "interactional" strategies that underpin the core category.
3. Selective coding / theoretical integration. During this stage, there is further linkage of categories to the core category in the pursuit of a coherent story. This stage continues until there is evidence of theoretical saturation, which in the case of grounded theory analysis, means that there are no new ideas or insights that can assist in the development of the theory.

Thematic analysis

I have already covered thematic analysis here. In thematic analysis, there is an emphasis on interpretation as researchers identify, analyse, interpret and report patterns. Writing up the data is also considered to be an essential part of the analysis. Here are the six main steps of thematic analysis:

1. Immersion
2. Generating initial codes
3. Searching for and identifying themes to create initial "thematic maps"
4. Reviewing themes
5. Defining and naming themes
6. Writing the report (which is another opportunity to refine the analysis)

Interpretative phenomenological analysis

Interpretative phenomenological analysis aims to understand and "give voice" to the concerns of participants, and contextualise and make sense of such concerns. The focus is on the understanding and experience of the participants. The main steps in interpretative phenomenological analysis are as follows:

1. Searching for themes in one individual's case (rather than across all cases)
2. Identifying themes
3. Connecting themes
4. Producing a table of themes
5. Continuing analysis with other cases
6. Writing- translating themes into a coherent account

Narrative analysis

Narrative analysis, as the name suggests, focuses on stories, which are a primary source of qualitative data. Narrative analysis aims to work with the whole story intact and examine how stories are put together. There are many types of narrative analysis:

• Thematic narrative analysis: Examines the core pattern by looking at the content
• Structural analysis of narrative types: Examines how stories are put together
• Performative analysis: Examines how talk is relationally and interactively produced and performed as a narrative
• Analytical bracketing: Analytically moves back and forth between the "whats" and "hows" of the narrative

Discourse analysis

Discourse analysis has a broader focus on language, and aims to attend to the "business" that is done in talk. There are several types of discourse analysis. Foucauldian analysis understands discourse as a process of constructing objects and subjects, and is primarily concerned with ways in which discourse constructs subjectivity, selfhood and power relations. Critical discourse analysis examines the social and political context of discourse.

Visual analysis

Visual narrative analysis focuses on the production of images, the images themselves, and how different audiences interpret images.

Computer-assisted data analysis

Computers can also be used to assist with coding. I'm not sure whether this truly counts as a separate type of analysis, or rather just a tool to aid in the other types of analysis.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

Using interviews in qualitative research

Since interviews are so widely used in qualitative research, I'm going to look at them further here.

Types of interviews

Interviews can be seen as a "joint construction about selves and the social world," as both the researcher and the researched work together to learn about the phenomenon of interest. Interviews can be completely unstructured, with only a broad, open-ended question to keep the phenomenon of interest in mind. They may instead be semi-structured, in which there is a pre-planned guide to direct the interaction, but most questions used are open-ended and allow for a variety of responses.

Focus groups are an offshoot of interviews that I will cover in a little more detail, given that I'll be running focus groups for my Honours project. Instead of being one-on-one like a traditional interview, focus groups involve a small group of around 4-8 participants, allowing participants to bounce ideas off one another in a supportive environment. Usually, no more than 10 participants are involved, as all participants need to have the opportunity to share their views. Sometimes, homogeneity between participants is desired in order to capitalise on shared experiences; other times, heterogeneity may be preferred.

There are also sub-types of focus groups. In concerned focus groups, participants have a stake in the topic. In naive focus groups, on the other hand, participants do not have a particular commitment to the topic.

Of course, there are many other types of interviews outside of traditional one-on-one interviews and focus groups. For instance, interviews may be conducted over the computer or phone. Interviews may also be conducted while moving around, which might be useful in cases where you want the participants to "show you around their world."

Conducting interviews

Unless you are conducting a completely unstructured interview, chances are you will need an interview guide. The number of questions should be kept as small as possible in order to avoid over-questioning your participants. Questions should be kept open-ended and free of jargon or awkward phrasing that might make interpretation difficult. The order of questions is also important: broader, less-invasive questions should be asked before more invasive questions; questions addressing a similar theme should be grouped together; and final questions should invite the participant to fill in any unexplored gaps. Once you have created your interview guide, you should test it on a friend or a colleague.

Before the interview, you should remind your participants and test your recording equipment. If this is one of a series of interviews with the same participant, you should read over previous transcripts to refresh your memory. You should also consider what style of dress would be most appropriate for the interview, as certain styles of dress may put off or intimidate particular groups of participants.

During the interview, after checking the recording equipment again, you should explain who you are and what your study is about. You should also emphasise how data will be kept private and anonymous. While actively listening to your participants, you should record non-verbal communication, biographical objects that your participant may have brought in, and other characteristics that are important for your research. You may also consider using aids yourself (visual or otherwise) to help stimulate participant recall. Throughout the interview, remind yourself of your research questions, but make sure not to ask them all at once. Once the interview is finished, be sure to thank your participant for their involvement in your research.

After the interview, write down some notes about the interview while it is still fresh in your mind. You should also thank participants formally and consider asking them for their feedback. Later on, you can start transcription and analysis. There are many transcription methods that you can use, but I don't know enough about them to discuss them here.

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

Considerations in conducting qualitative research

Yep, it's yet another post on qualitative research. This post might be a bit all over the place (since my notes are a bit all over the place), but I'll do what I can to keep things organised.

What kinds of questions are asked in qualitative research?

Qualitative research often asks the "why" and "how" questions. As qualitative research is often open-ended and exploratory, questions may change depending on participants' experiences and responses. It may help to generate a "statement of purpose" that provides a framework or "anchor point" while not tying you down to a specific question or set of questions.

How might a researcher's perspective influence the type of research performed?

As discussed here, how we perceive reality and knowledge informs methodology, which in turn informs the actual methods used in research. Let's see some more examples of how this happens.

Phenomenology describes things as they are experienced, and requires the suspension of taken-for-granted assumptions. Phenomenologists often distil the phenomenon down to its "essence" by using a technique called free imaginative variation, in which parts of a phenomenon are changed to see if the phenomenon is still identifiable.

Grounded theory is a methodology for using qualitative data to construct theories that are grounded in the data. Researchers using grounded theory will often aim to continue until the point of data saturation (when nothing new is found). Grounded theory researchers develop abstract ideas and seek specific data to refine emerging categories. They may select cases that are the most likely to produce relevant data for testing any emerging theories.

Critical/openly ideological research is conducted with the central intention of emancipation: enabling people to gain enough knowledge and power to be in control of their own lives. Knowledge is seen to be structured by existing sets of social relations. Researchers conducting this kind of research may perform participatory action research, which seeks to actively involve participants as part of a commitment to change.

What is the role of the researcher in qualitative research?

No matter what kind of research is performed, researchers need to keep reflexivity in mind. Reflexivity is a self-aware analysis of the dynamics between the researcher and the researched. A reflexive research journal may help in this regard.

The role of the researcher may change depending on the kind of research. For example, in some kinds of studies, researchers may simply observe a phenomenon from an outsider's perspective. In other kinds of studies, researchers may become immersed in the phenomenon (e.g. attempt to become a member of a particular culture) in order to observe from an insider's perspective.

How is sampling performed?

Unlike many quantitative studies, where the sample size can be calculated from the outset using the statistical concept of power, qualitative studies allow for the flexibility to make sample size decisions throughout the course of the study. Generally, 4-40 participants are sought, but it depends on the type of study. There are many ways in which participants may be chosen: using certain predetermined criteria, looking for "typical" cases, looking for "deviant" cases, asking participants to refer other suitable participants, convenience sampling, and so on.

How are case studies used in qualitative research?

There are several ways in which case studies can be used in qualitative research. In the intrinsic approach, a researcher may seek to obtain a better understanding of a particular case, and in the instrumental approach, a case is studied to provide insight or redraw a generalisation. The multiple/collective approach can be used to jointly study a number of cases. Choosing participants for a case study is done in the same way as choosing participants for any other kind of qualitative study.

How do qualitative researchers make use of the literature?

Qualitative researchers use the literature in various ways. Firstly, they can use it to define the phenomenon of interest and identify research gaps. Secondly, they can use it to determine appropriate methodological choices. Thirdly, they can use it to compare and contrast the results from their own study (or studies).

How long does qualitative research take?

Obviously, this depends on the kind of research that you are doing. Bear in mind, though, that qualitative studies can take a lot of time: the simple act of transcribing a 1hr talk can take 5 hours or so (and that's not even getting to data analysis)!

References

Sparkes, AC & Smith, B 2013, Qualitative Research Methods in Sport, Exercise and Health: From Process to Product, Taylor & Francis Ltd, London.

The COREQ checklist for qualitative studies

Checklists have been developed to evaluate the quality of pretty much any kind of study, and qualitative studies are no exception. The Consolidated Criteria for Reporting Qualitative Research (COREQ) checklist is a checklist that can be used to evaluate qualitative studies, such as focus group studies. It is divided into three domains.

Domain 1: Research team and reflexivity

This domain contains stuff such as personal characteristics of the research team, which may help in evaluating potential personal bias. The relationship of the researcher(s) with the participant(s) is also considered.

Domain 2: Study design

This domain evaluates participant selection, sample size, setting, presence of non-participants who may influence responses, questions and prompts in data collection, and the duration of interviews or focus groups. Domain 2 also evaluates how well the researchers have clarified their theoretical framework.

Domain 3: Analysis and findings

This domain evaluates how well researchers have clarified the process of coding and identification of themes, backed up their assertions with quotations from different participants, and assessed consistency where multiple coders were involved.

References

Tong, A, Sainsbury, P & Craig, J 2007, 'Consolidated criteria for reporting qualitative research (COREQ): a 32-item checklist for interviews and focus groups', International Journal for Quality in Health Care, vol. 19, no. 6, pp. 349-357

Content analysis and thematic analysis

I haven't written in a while, mainly because writing about stuff that I still don't fully understand just seemed way too overwhelming. But since I don't have anything pressing to do right now, I can't make any more excuses for myself :P

What are content and thematic analyses?

Content analysis is a general term for a number of strategies used in analysing text. It is mainly used to code and categorise large amounts of textual information, and may pay attention to certain aspects such as trends, word patterns, word frequency, and relationships between words. In content analysis, the characteristics of the document are described, including aspects such as: "Who says what to whom, with what effect?" Content analyses generally describe the data in conceptual form.

Thematic analysis is a method for identifying, analysing and reporting patterns in text. The main goal in thematic analysis is to search for and identify common threads that extend across an entire interview. Thematic analyses generally use minimal description, while interpreting various aspects of the topic.

Both content and thematic analyses are largely based on the factist perspective, in which data are seen to be more or less accurate and reflective of reality.

Modalities

Content analysis and thematic analysis are associated with two modalities: inductive and deductive. An inductive modality is used when no previous studies exist, and requires the collection of rich data for development of a theory. In contrast, a deductive modality is used when testing a theory in a different situation, or comparing categories at different periods. In contrast to using an inductive modality, using a deductive modality provides results that are less "rich," but more detailed.

Data analysis

The first step of either content or thematic analyses is to transcribe the data. The transcripts should then be read several times. Both types of analysis require breaking the data down into codes. However, content and thematic analyses differ somewhat in what kind of content is considered, and how themes are developed.

In content analysis, researchers usually choose between latent ("hidden, below-the-surface") content and manifest (overt) content. Themes are reached based on the frequency of occurrence of codes in the data.

In thematic analysis, researchers usually consider both latent and manifest content (rather than choosing one or the other). The importance of themes depend on their relevance to the research question. In thematic analysis, data analysis continues right through writing up the data, which is considered to be the final opportunity of data analysis.

Evaluation of the analysis

Credibility, dependability, confirmability and transferability are aspects that may be considered when evaluating the quality of a content or thematic analysis. It is also important to consider whether or not new insights have been provided by the analysis.

References

Vaismoradi, M, Turunen, H & Bondas, T 2013, 'Content analysis and thematic analysis: Implications for conducting a qualitative descriptive study', Nursing and health sciences, vol. 15, no. 3, pp. 398-405.