Thematic Analysis: Definition, Advantages, and Disadvantages
A characteristic feature of the modern stage of psychology is the desire of scholars to develop such research methods that would be relevant to the object under the study, since this science involves not only the very theme but also the facts lying outside this given object. In this connection, thematic analysis acts as one of the most widely used methods to specify and interpret qualitative data. The scientists frequently use qualitative analytic methods when conducting research studies in psychology, among which one can note thematic analysis. On the one hand, some scholars reckon that it refers to the fundamental instrument of analysis. On the other hand, there is also an opposite view, according to which it is not a separate method but a shared skill that serves as a tool across different methods. Thematic analysis is a process applied in the framework of interpretive social research, maintaining a professional search for themes to analyse in the existing data. These themes relate to concepts and express significant information gathered from a text in terms of fulfilling an interpretive task of finding themes and patterns.
It goes without saying that some themes are rather complex to reveal immediately and directly. To identify them, one needs to repeatedly study the text, being harnessed with a researcher’s enthusiasm. At this point, the researchers pinpoint that the identification of certain thematic constructions is easier if one studies the work written, for example, not in a professional journal but in the one that would be comprehensible to the wider audience. The scientists targeting the mentioned audience tend to disclose their thematic prerequisites more transparently rather than their opponents. Thus, the concept of thematic analysis implies holistic and comprehensive research along with the importance of cooperation between different disciplines, especially between the natural and human sciences. A quintessence of various experiences allows discussing a research question from numerous perspectives, which presupposes minimisation of biases and constrained ideas.
Among the benefits of thematic analysis in conducting the qualitative study, one may emphasise that it is not intricate or time-consuming, so there is no necessity to waste extra time on studying. It also provides scholars with the opportunity to select any theoretical framework unlike other methods of analysis that are appropriate for a limited list of theories. Such flexibility presents one of the core benefits researchers obtain while using thematic analysis. This type of analysis implies the utilisation of simple words and avoidance of sophisticated formulas to reveal the results obtained in the course of data collection. The professionals state that resorting to thematic analysis is advantageous if a research study requires collaboration with participants, thus ensuring summary of information from a large text and reflection of its key features. As a result, it is possible to maintain an in-depth description of data that simplifies the search for similarities and differences. It is also beneficial to apply thematic analysis in case there is a necessity to deal with informing policy development and operate large data sets as it provides an opportunity to make information brief and develop categories from data, so it becomes easier to navigate through it.
One should also outline several disadvantages of thematic analysis to create a comprehensive vision of the notion. Rather often, scholars explore solely inferior characteristics of the conducted analysis or research question, which resorts to mistakes made by researchers but not to the method itself. Nevertheless, it is critical to consider the drawbacks of flexibility, namely, that it prevents researchers from revealing a sense of continuity and contradiction. Thematic analysis may be appropriate as far as the documents on which the researcher is working are credible as well. The search for data and its evaluation should accept such criteria as authenticity, reliability, and representativeness. It is barely possible to compile a coding guide that does not include a certain amount of interpretation from the encoders. The coders should rely on their daily experience and knowledge as representatives of their culture to be able to code the material they work with. Even though the possibility to develop numerous interpretations affects reliability negatively, scientists can avoid this complication if they monitor themes and codes tables.
Another noteworthy disadvantage refers to the fact that researchers do not generally recognise it as a separate analytic method even though plenty of them resort to it in their works. With this in mind, studies in which thematic analysis takes places often seem to be atheoretic. Thus, professionals should be more considerable when developing a research question for a qualitative analysis to ensure that it fits thematic analysis and provides them with an opportunity to benefit from this method to the maximum, focusing on well-grounded theoretic assumptions based on the authoritative information.
Stages of Thematic Analysis
The phases of thematic analysis and other qualitative research methods are rather alike. However, there are six specific stages composing thematic analysis that require discussion. At the beginning of the process, a scientist starts searching for and noticing patterns (themes) and topics of interest, but analysis requires him or her to resort to this phase constantly. What is more, scientists have different views regarding the phase, during which they should maintain the engagement with the literature. Some provide evidence to prove that the initial state is the most appropriate for it, because it allows narrowing the field of analysis down while others consider that the last stages provide more benefit, because they can consider new features of the data. Nevertheless, thematic analysis is a recursive process, the maintenance of which consists of the following phases:
- Familiarising with the data. It is possible to obtain the data from the third parties and to collect it without additional assistance. Regardless of the selected option, it is critical to immerse oneself in the data for researchers to understand its content. The specified stage often requires repeated active reading of the data. One should search for meanings and patterns. In addition to the aforementioned, professionals find it beneficial to read all the information at least once before coding, so that they already have particular ideas before they start working. Researchers utilise such an approach even if a detailed analysis is not their major purpose, because it ensures understanding of the topic. It is also beneficial to start taking notes at this stage in order to make the coding process easier. However, this process is time-consuming, which does not appeal to its users. Researchers engage in data transcription if it is initially verbal. At this phase, they create a meaning and achieve a better understanding of the data.
- Generating initial codes. After one has generalised ideas of the data, he or she is prepared to start producing initial codes. The researchers identify a theme of interest and align it with the basic segment while coding depends on the type of themes and a professional’s aim to work with the whole data set of its particular features. The scientists can perform it with the help of a software program with equal attention paid to each data item. They maintain coding with the help of notes written on texts, highlighting potential patterns, or using post-in notes. It is also possible to list codes and then match them with abstracts. The researchers develop codes for a large variety of themes and some related relevant data.
- Searching for themes. Having prepared a list of codes, professional should then put them into potential themes. Here, one should analyse codes and ponder over the possibility to combine them. By writing codes on separate pieces of paper or putting them on a table, a researcher can streamline the process, detecting a relationship between them and gathering data relevant to the given theme.
- Reviewing themes. When professionals outline potential themes, they should review them to ultimately develop the final version. During this process, they may find out that some themes are not actually themes because of the lack of supporting data, similarity with one another, or the possibility to separate them. It is essential to keep in mind that themes should be large, but their content is to be coherent. The researchers divide reviewing in two levels: the first one focuses on the coherency of coded data extracts; the second refers to fixing of inappropriate themes. As soon as they develop the final version of themes, they design a thematic map to navigate through themes and codes straightforwardly. However, if a thematic map does not fit the data set, then researchers should return to the previous task and review coding until the map would acquire necessary characteristics. Thus, one will be aware of all the themes to be analysed, the way they fit together, and information they reveal.
- Defining and naming themes. Having a thematic map, a researcher should identify the essence of each theme and define what it captures. At this phase, one identifies what is of his or her interest and adds a rationale. The identification of subthemes is advantageous for researchers, so they should look for some and reveal the hierarchy of meaning. Scientists need to put the scope and the content of each theme in a couple of sentences and change the initial titles of themes into the names that are suitable for their final analysis.
- Producing the report. The subsequent phase implies that a researcher should complete the final analysis and prepare a write-up report based on the gathered data. The task is to present information in a convincing way, so that readers understand quality and validity of the analysis. The scientists should ensure that the data is in the form of a coherent and interesting story that discusses all the themes (Braun & Clarke 2006). The write-up should also include evidence of themes extracted from the initial data. Researchers can add vivid examples yet avoid unnecessary complexity. An analytic narrative should not only describe the data but also make an argument aligned with the research question.
Abduction Approach
The researchers can use several approaches to conducting analyses, referring to incorporated reasoning with the help of the three methods: induction, deduction, and abduction. While professionals often discuss the first two and use them actively while working, the third approach is less accepted. Abduction is a specific way of searching for explanatory hypotheses along with traditional forms of deductive and inductive reasoning (Danermark et al. 2002). This approach serves as an alternative to the hypothetical deductive method. The task of abduction is not a mere derivation of a probable conclusion but serving as an effective tool for searching for scientific hypotheses that are necessary to explain the available facts. In other words, abduction seeks to identify a universal internal mechanism, through which it is possible to construct a hypothesis that best explains the observed facts.
The logical scheme of abductive reasoning has the following form: there is some phenomenon of A; A would be true if hypothesis B is true; and, consequently, there is reason to believe that hypothesis B is true. In the light of the above reasoning, the abductive argument seems to be similar to a hypothetical deductive reasoning, since it assumes the hypothesis as a premise. However, the course of reasoning in these logical methods turns out to be exactly the opposite. One should emphasise that abduction begins with a thorough analysis and an accurate assessment of established facts that determine the choice of a hypothesis for their explanation. It is not an unerring method of discovering new truths in science but a discovery algorithm, the paramount purpose of which is to search for hypotheses that can contribute to the explanation of these facts.
Abduction is significant since one phenomenon often has numerous explanations, and there is a necessity to define the most appropriate variant for professionals to discuss it above all. However, it is also critical to pinpoint that abduction is more about an attitude than reasoning or a logical order. Abduction is a method of the unfolding interaction that is not fully reliable as it bases on hypotheses that may turn out to be true as easily as they may prove to be false. Nevertheless, there is no other way, perhaps, because deduction does not give new knowledge. Thus, abduction approach encourages professionals to abandon old convictions and develop new ones, thus revealing a valid attitude that aligns with the previously developed ideas.
Theoretical Saturation
The selection of categories in the target sample ends when the saturation point reaches the relative completeness and exhaustion of the theory or descriptive typology. The highest degree of specification of categories, when the provision of additional data fits into the formulated frameworks and does not bring new information, indicates the theoretical saturation. However, in the strict sense, a researcher can never be sure that new categories will not bring new information. In fact, theoretical saturation occurs if there is no need to significantly rearrange the system of categories even in the case of obtaining new data. In accomplishing the mentioned point, a researcher may focus on properties and dimensions, ideas and characteristics, relations between categories and their validity. Only by appropriately addressing all these peculiarities, a professional researcher can conclude that he or she has successfully reached theoretical saturation, and there is no necessity in the further analysis.
Reference
Braun, V & Clarke, 2006, ‘Using thematic analysis in psychology’, Qualitative Research in Psychology, vol. 3, no. 2, pp. 77-101.
Danermark, B, Ekström, M, Jakobsen, L & Karlsson, J 2002, Explaining society: critical realism in the social sciences, Routledge, New York, NY.