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.
Thursday, June 14, 2018
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
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.
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.
Monday, May 14, 2018
Theoretical perspectives and research methodologies
In a previous post, I briefly touched on how different theoretical perspectives can guide methodology and ultimately research methods. In this post, I will talk about a wider range of theoretical perspectives, mainly because I need help familiarising myself with them :P
Objectivism
Objectivism posits that reality exists independently of consciousness (i.e. there is an ultimate truth out there to be found).
Positivism
Positivism, like objectivism, posits that reality is external to the researcher and that there should be a focus on facts, rather than on values. Positivism suggests that reality consists of what is available to the senses, and inquiry should be based on scientific observation, which may or may not be theory-laden. Post-positivists, while also agreeing that there is an independent reality to be studied, believe that all observation is inherently fallible.
Interpretivism
Interpretivism looks for culturally-derived and historically situated interpretations of the social life world.
Symbolic interactionism
In symbolic interactionism, meanings arise from social interaction and are handled in, and modified by, an interactive process. Meanings are not fixed or stable, but are rather revised on the basis of experience.
Phenomenology
Phenomenology suggests that attempts to understand social reality must be grounded in people's experiences of it. There is a focus on allowing phenomena to "speak for themselves" and to find the internal logic of subject. To this end, researchers using a phenomenological perspective tend to use interviews.
Realism
The realist perspective states that science paints an accurate picture of the world; however, measuring external reality may be difficult as some observable "facts" may be merely illusions. Realism sometimes lends itself to the pluralism methodology, in which nothing is ruled out. There are different types of realist perspectives. Naïve realism states that the world is largely knowable and is just as it appears to be. Scientific realism suggests that science can find the true representations of the world, though science may be fallible. Critical realism states that the way we perceive the world depends on beliefs and expectations.
Hermeneutics
In hermeneutics, social reality is socially constructed and is considered to be too complex to be understood via observation. As such, interpretation is considered more important in hermeneutics than explanation and description.
Naturalistic inquiry
Naturalistic inquiry looks at multiple constructed realities that can only be studied holistically. It consists of a body of knowledge that describes individual cases.
Critical inquiry
Critical inquiry questions currently-held values and assumptions. It posits that ideas are mediated by power relations and that facts cannot be disentangled from ideology.
Postmodernism
Postmodernism emphasises multiplicity, ambiguity, ambivalence and fragmentation. Researchers using the postmodernist perspective may deconstruct texts to expose how values and interests are embedded within.
Pragmatism
Pragmatism is... well, pragmatic. It basically says that you should use what works.
Research Methodologies
Now that we've gone through a list of ideological perspectives, let's look at some of the methodologies that we can use! This is only a brief list and is by no means exhaustive.
So now that you've heard a few different research approaches, the next question is which approach to use. Firstly, you need to figure out what it is you want to research and what questions you want to answer. For instance, do you want to explore what is happening and figure out if a phenomenon warrants further research? If so, you would use an exploratory study design. Do you want to "draw a picture" of a situation? If so, you would use a descriptive study design, being mindful that these kinds of studies cannot explain why things occur. If you want to know why things occur, you would use an explanatory study design, and if you want to explore people's experiences, views and perspectives, you would use an interpretive study design.
References
Gray, DE 2014, Doing Research in the Real World, SAGE Publications Ltd.
Objectivism
Objectivism posits that reality exists independently of consciousness (i.e. there is an ultimate truth out there to be found).
Positivism
Positivism, like objectivism, posits that reality is external to the researcher and that there should be a focus on facts, rather than on values. Positivism suggests that reality consists of what is available to the senses, and inquiry should be based on scientific observation, which may or may not be theory-laden. Post-positivists, while also agreeing that there is an independent reality to be studied, believe that all observation is inherently fallible.
Interpretivism
Interpretivism looks for culturally-derived and historically situated interpretations of the social life world.
Symbolic interactionism
In symbolic interactionism, meanings arise from social interaction and are handled in, and modified by, an interactive process. Meanings are not fixed or stable, but are rather revised on the basis of experience.
Phenomenology
Phenomenology suggests that attempts to understand social reality must be grounded in people's experiences of it. There is a focus on allowing phenomena to "speak for themselves" and to find the internal logic of subject. To this end, researchers using a phenomenological perspective tend to use interviews.
Realism
The realist perspective states that science paints an accurate picture of the world; however, measuring external reality may be difficult as some observable "facts" may be merely illusions. Realism sometimes lends itself to the pluralism methodology, in which nothing is ruled out. There are different types of realist perspectives. Naïve realism states that the world is largely knowable and is just as it appears to be. Scientific realism suggests that science can find the true representations of the world, though science may be fallible. Critical realism states that the way we perceive the world depends on beliefs and expectations.
Hermeneutics
In hermeneutics, social reality is socially constructed and is considered to be too complex to be understood via observation. As such, interpretation is considered more important in hermeneutics than explanation and description.
Naturalistic inquiry
Naturalistic inquiry looks at multiple constructed realities that can only be studied holistically. It consists of a body of knowledge that describes individual cases.
Critical inquiry
Critical inquiry questions currently-held values and assumptions. It posits that ideas are mediated by power relations and that facts cannot be disentangled from ideology.
Postmodernism
Postmodernism emphasises multiplicity, ambiguity, ambivalence and fragmentation. Researchers using the postmodernist perspective may deconstruct texts to expose how values and interests are embedded within.
Pragmatism
Pragmatism is... well, pragmatic. It basically says that you should use what works.
Research Methodologies
Now that we've gone through a list of ideological perspectives, let's look at some of the methodologies that we can use! This is only a brief list and is by no means exhaustive.
- Experimental and quasi-experimental- RCTs, cohort studies, case-control studies, etc.
- Phenomenological research- Producing "thick descriptions" of people's experiences and perspectives. Often based on small case studies.
- Analytical surveys- Carefully-designed surveys with careful random selection of samples and a tight survey structure.
- Action research- Carried out in close collaboration between researchers and practitioners. There is an emphasis on promoting change. Both quantitative and qualitative approaches may be used. Case studies are the main medium of research.
- Heuristic inquiry- Open-ended, self-directed inquiry in which researchers are immersed in active experience. There is deep, personal questioning. A disadvantage of heuristic inquiry is that it is only weakly generalisable.
So now that you've heard a few different research approaches, the next question is which approach to use. Firstly, you need to figure out what it is you want to research and what questions you want to answer. For instance, do you want to explore what is happening and figure out if a phenomenon warrants further research? If so, you would use an exploratory study design. Do you want to "draw a picture" of a situation? If so, you would use a descriptive study design, being mindful that these kinds of studies cannot explain why things occur. If you want to know why things occur, you would use an explanatory study design, and if you want to explore people's experiences, views and perspectives, you would use an interpretive study design.
References
Gray, DE 2014, Doing Research in the Real World, SAGE Publications Ltd.
Conducting rigorous qualitative research
Qualitative research often gets criticised for being subjective and prone to bias, and those concerns are probably not unfounded. There are, however, ways in which we can improve qualitative research. Here are some methods:
Member checking
In member checking, the researchers return the data (transcripts etc.) and/or the results to the participants to allow them to assess the trustworthiness and ask for input. Member checking has its basis in epistemological foundationalism, which assumes that this method is neutral and can control for bias, allowing for the sorting of the more trustworthy information from the less trustworthy information. However, since there is no "true reality" in qualitative studies to compare to, it is impossible for researchers to truly objectively sort trustworthy from untrustworthy information. Similarly, there is no evidence that member checking enhances credibility or trustworthiness.
There are also practical problems with member checking. For instance, the participant and researcher may disagree with or contradict each other, though this is rarely reported in papers. In case of a disagreement, participants may be inclined to simply agree with the researcher, due to the power imbalance between participant and researcher. Participants may skim over the data, be unable to understand the data, or simply not remember what they said when the data were initially generated, making them unable to offer truly useful insight. Furthermore, political leanings and personal interests may also influence the member checking process, particularly if the topic studied is a sensitive one (e.g. illicit drug use).
Fortunately, all is not lost. Member reflection may not be perfect for verifying results, but it may help to generate additional data. When participants re-read the data, they may generate useful insights. Another positive aspect of member checking is that it may help to ensure that descriptions of the participants do not accidentally reveal who they are.
Calculating inter-rater/inter-coder reliability
This method is exactly what it says on the tin: trying to calculate some kind of inter-rater reliability metric. As simple and objective as it sounds, calculating inter-rater reliability comes with its own disadvantages. Firstly, different coders may perform unitisation (identify blocks of text for coding) differently to one another, making it difficult to compare coders. Secondly, if coders then combine to discuss their differences in results, power differentials etc. between coders may skew the final answers. Thirdly, it is difficult to determine a satisfactory level for inter-coder reliability.
One way to get around problems such as unitisation differences is to use a set of guidelines. Coders can develop a framework and firm coding rules prior to coding. The downside of this is that it may constrain creativity and limit the identification of unanticipated knowledge.
Another way to get around the problems with inter-rater reliability is to use the "critical friends" technique. In this technique, there is critical dialogue between people. Researchers voice interpretations to others who listen and offer feedback. The goal here is to create a "sounding board" or reflexivity, rather than total agreement.
Yet another way to get around the problems with inter-rater reliability is to simply ignore reliability altogether. After all, you will never be able to reproduce the exact same interview.
Universal criteria
Another possible method for improving the quality of qualitative research is to try and adhere to a set of criteria, such as the universal criteria proposed by Tracy (2010). These universal criteria consist of eight main hallmarks: worthy topic, rigor, sincerity, credibility, resonance, significant contribution, ethics and meaningful coherence. Instead of using fixed criteria, an alternative approach is a relativist one: using criteria from open-ended lists so that the criteria are appropriate for the study design.
References
Smith, B, McGannon, KR 2017, 'Developing rigor in qualitative research: problems and opportunities within sport and exercise psychology', International Review of Sport and Exercise Psychology, DOI: 10.1080/1750984X.2017.1317357
Tracy, SJ 2010, 'Qualitative quality: Eight "big tent" criteria for excellent qualitative research', Qualitative Inquiry, vol. 16, no. 10, pp. 837-851.
Member checking
In member checking, the researchers return the data (transcripts etc.) and/or the results to the participants to allow them to assess the trustworthiness and ask for input. Member checking has its basis in epistemological foundationalism, which assumes that this method is neutral and can control for bias, allowing for the sorting of the more trustworthy information from the less trustworthy information. However, since there is no "true reality" in qualitative studies to compare to, it is impossible for researchers to truly objectively sort trustworthy from untrustworthy information. Similarly, there is no evidence that member checking enhances credibility or trustworthiness.
There are also practical problems with member checking. For instance, the participant and researcher may disagree with or contradict each other, though this is rarely reported in papers. In case of a disagreement, participants may be inclined to simply agree with the researcher, due to the power imbalance between participant and researcher. Participants may skim over the data, be unable to understand the data, or simply not remember what they said when the data were initially generated, making them unable to offer truly useful insight. Furthermore, political leanings and personal interests may also influence the member checking process, particularly if the topic studied is a sensitive one (e.g. illicit drug use).
Fortunately, all is not lost. Member reflection may not be perfect for verifying results, but it may help to generate additional data. When participants re-read the data, they may generate useful insights. Another positive aspect of member checking is that it may help to ensure that descriptions of the participants do not accidentally reveal who they are.
Calculating inter-rater/inter-coder reliability
This method is exactly what it says on the tin: trying to calculate some kind of inter-rater reliability metric. As simple and objective as it sounds, calculating inter-rater reliability comes with its own disadvantages. Firstly, different coders may perform unitisation (identify blocks of text for coding) differently to one another, making it difficult to compare coders. Secondly, if coders then combine to discuss their differences in results, power differentials etc. between coders may skew the final answers. Thirdly, it is difficult to determine a satisfactory level for inter-coder reliability.
One way to get around problems such as unitisation differences is to use a set of guidelines. Coders can develop a framework and firm coding rules prior to coding. The downside of this is that it may constrain creativity and limit the identification of unanticipated knowledge.
Another way to get around the problems with inter-rater reliability is to use the "critical friends" technique. In this technique, there is critical dialogue between people. Researchers voice interpretations to others who listen and offer feedback. The goal here is to create a "sounding board" or reflexivity, rather than total agreement.
Yet another way to get around the problems with inter-rater reliability is to simply ignore reliability altogether. After all, you will never be able to reproduce the exact same interview.
Universal criteria
Another possible method for improving the quality of qualitative research is to try and adhere to a set of criteria, such as the universal criteria proposed by Tracy (2010). These universal criteria consist of eight main hallmarks: worthy topic, rigor, sincerity, credibility, resonance, significant contribution, ethics and meaningful coherence. Instead of using fixed criteria, an alternative approach is a relativist one: using criteria from open-ended lists so that the criteria are appropriate for the study design.
References
Smith, B, McGannon, KR 2017, 'Developing rigor in qualitative research: problems and opportunities within sport and exercise psychology', International Review of Sport and Exercise Psychology, DOI: 10.1080/1750984X.2017.1317357
Tracy, SJ 2010, 'Qualitative quality: Eight "big tent" criteria for excellent qualitative research', Qualitative Inquiry, vol. 16, no. 10, pp. 837-851.
Qualitative vs. Quantitative Methodologies
You've probably heard of qualitative and quantitative studies before: quantitative studies tend to be focused on numbers and cold, hard facts, whereas qualitative studies aren't. There are also some other ideological differences underpinning different types of studies, which I'll talk about in this post.
Firstly, some terminology. Ontology is the study of reality, and ontological positions describe what entities exist and what kinds of relationships exist. Ontology can inform epistemology, which is the study of knowledge and how we can truly know things. In turn, epistemology can inform methodology, which is the theoretical and philosophical system that structures the way that research is conducted, and ultimately determines the methods that we use to learn things.
Quantitative paradigm
The quantitative paradigm comes from positivism, which is a "realist orientation" that posits that an ultimate truth exists and can be described as it really is. Furthermore, the truth can be described using facts, which are separated from values. In quantitative studies, the investigator is seen to be able to study phenomena without influencing or being influenced by them.
Qualitative paradigm
The qualitative paradigm, on the other hand, is derived from idealism. According to idealism, reality depends on one's mental structure and activity, and cannot be accessed independent of our minds. As such, there is no single reality; instead, there are multiple realities based on one's construction or interpretation of reality. Our perception of reality is not value-free, and things cannot be described as they really are but only as we perceive or interpret them. The "truth" in qualitative studies is essentially defined as the extent to which a researcher's statements correspond to how people out there really interpret or construct their realities. Generally, the main goal in qualitative studies is not to find the "ultimate truth," but instead to understand a phenomenon better from the point of view of study participants, and to allow readers of the study to make use of these experiences based on their descriptions in the text.
References
Slevitch, L 2011, 'Qualitative and Quantitative Methodologies Compared: Ontological and Epistemological Perspectives', Journal of Quality Assurance in Hospitality and Tourism, vol. 12, no. 1, pp. 73-81.
Firstly, some terminology. Ontology is the study of reality, and ontological positions describe what entities exist and what kinds of relationships exist. Ontology can inform epistemology, which is the study of knowledge and how we can truly know things. In turn, epistemology can inform methodology, which is the theoretical and philosophical system that structures the way that research is conducted, and ultimately determines the methods that we use to learn things.
Quantitative paradigm
The quantitative paradigm comes from positivism, which is a "realist orientation" that posits that an ultimate truth exists and can be described as it really is. Furthermore, the truth can be described using facts, which are separated from values. In quantitative studies, the investigator is seen to be able to study phenomena without influencing or being influenced by them.
Qualitative paradigm
The qualitative paradigm, on the other hand, is derived from idealism. According to idealism, reality depends on one's mental structure and activity, and cannot be accessed independent of our minds. As such, there is no single reality; instead, there are multiple realities based on one's construction or interpretation of reality. Our perception of reality is not value-free, and things cannot be described as they really are but only as we perceive or interpret them. The "truth" in qualitative studies is essentially defined as the extent to which a researcher's statements correspond to how people out there really interpret or construct their realities. Generally, the main goal in qualitative studies is not to find the "ultimate truth," but instead to understand a phenomenon better from the point of view of study participants, and to allow readers of the study to make use of these experiences based on their descriptions in the text.
References
Slevitch, L 2011, 'Qualitative and Quantitative Methodologies Compared: Ontological and Epistemological Perspectives', Journal of Quality Assurance in Hospitality and Tourism, vol. 12, no. 1, pp. 73-81.
Monday, May 7, 2018
Thematic analysis
For my Honours project, I will be doing some qualitative analysis on focus group data. One common form of qualitative analysis is called thematic analysis. Thematic analysis involves looking at themes, which represent some kind of patterned response or meaning within the data set. There is some flexibility in how a theme is defined, and this may depend on the research question. Themes can be either semantic (explicit) or latent (interpretative). They may be determined via an inductive ("bottom-up") process, in which data is coded and sorted without fitting it into a pre-existing framework, or via a theoretical/deductive ("top-down") process, which is explicitly analyst-driven.
There are several steps in a thematic analysis:
Phase 1: Familiarisation with the data
In this step, a rigorous and thorough transcription (for interviews and other verbal data) is done. The transcripts are read at least once. During this time, the analyst takes notes and marks ideas for coding.
Phase 2: Generating initial codes
During this phase, the analyst organises the data into units of meaning, or codes. All of the data extracts are coded for as many themes as possible and collated. Coding can be done manually or via software.
Phase 3: Searching for themes
Once the codes have been generated, they can be sorted into themes. Tables and mind-maps may be used to aid in this process.
Phase 4: Reviewing themes
During this step, the analyst refines themes to create internal homogeneity (i.e. making sure everything within a theme is similar) and external heterogeneity (i.e. making sure different themes have different contents). Firstly, the data are reviewed at the level of coded extracts to see if they form a coherent pattern. Next, the data are reviewed at the level of the whole data.
Phase 5: Defining and naming themes
Once the themes have been reviewed and refined, the analyst has to identify the "essence" of each theme, or the aspect of the data that the theme captures. The analyst then has to figure out how each theme fits into the "story" and if there are any relevant sub-themes. It may help to try and describe the scope and content of each theme in 1-2 sentences.
Phase 6: Producing the report
Once the themes have been found and defined, it's now time to write! However, care needs to be taken to ensure that claims remain grounded in the data, but are still able to go beyond the "surface" of the data.
Pitfalls
There are many errors that can be made while doing thematic analysis, particularly by beginners in qualitative analysis (*cough*me*cough*). Some of these pitfalls are as follows:
Braun, V, Clarke, V 2008, 'Using thematic analysis in psychology', Qualitative Research in Psychology, vol. 3, no. 2, pp. 77-101
There are several steps in a thematic analysis:
Phase 1: Familiarisation with the data
In this step, a rigorous and thorough transcription (for interviews and other verbal data) is done. The transcripts are read at least once. During this time, the analyst takes notes and marks ideas for coding.
Phase 2: Generating initial codes
During this phase, the analyst organises the data into units of meaning, or codes. All of the data extracts are coded for as many themes as possible and collated. Coding can be done manually or via software.
Phase 3: Searching for themes
Once the codes have been generated, they can be sorted into themes. Tables and mind-maps may be used to aid in this process.
Phase 4: Reviewing themes
During this step, the analyst refines themes to create internal homogeneity (i.e. making sure everything within a theme is similar) and external heterogeneity (i.e. making sure different themes have different contents). Firstly, the data are reviewed at the level of coded extracts to see if they form a coherent pattern. Next, the data are reviewed at the level of the whole data.
Phase 5: Defining and naming themes
Once the themes have been reviewed and refined, the analyst has to identify the "essence" of each theme, or the aspect of the data that the theme captures. The analyst then has to figure out how each theme fits into the "story" and if there are any relevant sub-themes. It may help to try and describe the scope and content of each theme in 1-2 sentences.
Phase 6: Producing the report
Once the themes have been found and defined, it's now time to write! However, care needs to be taken to ensure that claims remain grounded in the data, but are still able to go beyond the "surface" of the data.
Pitfalls
There are many errors that can be made while doing thematic analysis, particularly by beginners in qualitative analysis (*cough*me*cough*). Some of these pitfalls are as follows:
- Sorting but not analysing the data
- Using questions (i.e. the questions that participants were asked) as "themes"
- Non-coherent themes
- Overlapping themes
- Insufficient examples in report
- Mismatch between data and claims
- Mismatch between theory and claims
- Mismatch between research questions and the type of analysis
Braun, V, Clarke, V 2008, 'Using thematic analysis in psychology', Qualitative Research in Psychology, vol. 3, no. 2, pp. 77-101
Subscribe to:
Posts (Atom)