Content Analysis


Fig. 3.1

The research process of deductive data analysis



Research employing deductive content analysis relies on the same data collection methods and sources that are used in inductive content analysis, for example, interviews, observations, meeting documents, diary entries, historical documents and patient records. Any written material can serve as the input for deductive content analysis (see Chap. 1). As is also the case in inductive content analysis, the proper sample size for deductive content analysis is based on data saturation. Furthermore, the deductive content analysis process promotes returning to the research questions following data collection and analysis (Fig. 3.1). This is important for two reasons. First, it may be possible that the research questions and/or structure of analysis do not accurately reflect the theoretical structure. In this case, the researcher should revise their research question and/or structure of analysis to provide relevant results. A second issue is that the researcher may not have the correct data sources to answer their research questions. In this case, it is important that the researcher notices this immediately after initial data collection so that they can modify the data collection methodology. Researchers can lower the risk for both of these problems—and assure the trustworthiness of the study—by pre-testing the data collection tool and analysis matrix.


3.2.1 Examples of How Deductive Content Analysis Has Been Applied in Nursing Research


A study of adolescents with diabetes found the meaning of disease to encompass the following concepts: threats to mental well-being; threats to social well-being and threats to physical well-being. If a researcher wants to study whether adolescents with asthma have the same experience, they will have to test a previous concept in a new context. This can be achieved in two ways. In one approach, the researcher could create a questionnaire that focuses on the meaning of disease, collect data, and analyse the collected data using statistical methods. This would provide knowledge on the issues that were included in the questionnaire. In contrast to this quantitative approach, a researcher could also choose to create interview questions based on earlier knowledge (i.e. meaning of disease for adolescents with diabetes), interview a group of adolescents and analyse the data using deductive content analysis. This approach may offer the researcher an opportunity to gain other important knowledge from adolescents with asthma because the interviewees are free to answer with their own words and explain how they are feeling. This is one advantage that deductive content analysis has over quantitative research approaches.


Prior knowledge that adolescents with diabetes describe the meaning of disease through the concepts threats to mental well-being, threats to social well-being and threats to physical well-being could be tested in a new context (adolescents with asthma) with the following research questions: (1) Do adolescents with asthma experience threats to their mental well-being?; (2) Do adolescents with asthma experience threats to their physical well-being; (3) Do adolescents with asthma experience threats to their social well-being? (Fig. 3.2). The interview questions should also be based on earlier knowledge. For example, the researcher could ask the interviewees questions like: does having asthma threaten your mental well-being?; what kind of threats to mental well-being have you experienced?; does having asthma threaten your social well-being?; what kind of threats to social well-being have you experienced?; does having asthma threaten your physical well-being?; what kind of threats to physical well-being have you experienced? These questions are quite open and would allow adolescents to freely express their feelings. However, these questions are not appropriate for an inductive approach because they are influenced by the theoretical structure and prior knowledge. As such, the starting point for this hypothetical research would move away from the inductive end of the line presented in Chap. 1 (Fig. 1.​6) and approach the middle point (X).

../images/470474_1_En_3_Chapter/470474_1_En_3_Fig2_HTML.png

Fig. 3.2

The research process of a qualitative study which includes deductive content analysis


The data collection phase also requires the building of an analysis matrix. The collected data will already be structured based on the interview questions. For example, the hypothetical study mentioned above will include data which describe threats to mental well-being. An appropriate analysis matrix for this study—which is based on the same theoretical knowledge that underlies the research starting point, research questions and data collection—is presented in Table 3.1.


Table 3.1

An example of an unstructured analysis matrix














What kind of well-being threats do adolescents with asthma have?

Mental well-being threats

Social well-being threats

Physical well-being threats


The first step of data analysis is the selection of a unit of analysis. As discussed in Chap. 2, this can be one word, one sentence or meaning [7]. Inexperienced qualitative researchers may prefer to select a sentence as the unit of analysis because a sentence is easier to handle than a single word. The researcher will then go through the data to evaluate whether each sentence is related to the research question. All of the instances in which a sentence is related to the research question are recorded in the analysis matrix (see Table 3.2).
Tags:
Apr 18, 2020 | Posted by in NURSING | Comments Off on Content Analysis

Full access? Get Clinical Tree
Get Clinical Tree app for offline access