Coding and analysis of qualitative data
The analysis of qualitative data is normally seen as strenuous by most researchers. This is because analysis of qualitative data is a continuous process that begins from the minute the data is collected and continues throughout the research. It is therefore not a distinct procedure that can be separated from the rest of the processes involved in the research process. This is different from the analysis of quantitative data which is normally carried out as a separate procedure. In addition, the analysis of qualitative data cannot be left in the hands of expert analysts who had no part in the data collection process. Researchers interested in qualitative research have to collect data and analyze the collected data by themselves. This is because one of the main objectives of qualitative research is to provide the researchers with a deeper understanding of the subjects of their study and how these subjects view the world and the problem under investigation. As a result, researchers have to make use of their personal experiences with the subjects of the study; the settings under which the data were collected, and any secondary documents that were used to collect data. Indeed, analysis of qualitative data is considered to be the most difficult element of qualitative research studies. This owes to the complexity of the data involved in such studies. Because qualitative studies deal with opinions and views of the participants rather than their numbers, the data collected often differs and varies widely depending on the social, cultural, gender, psychological, and emotional orientations of the participants. To analyze such differing data, the process of coding becomes crucial to the researcher. Whereas coding and analysis of qualitative data are not tantamount, coding is an important aspect of analysis and must be done before any analysis can be conducted.
Coding of qualitative data entails the classification of the raw data into categories in an attempt to find existing relations among the categories. Seidel and Kelle argue that coding involves “noticing relevant phenomena; collecting examples of those phenomena; and analyzing those phenomena in order to find commonalities, differences, patterns, and structures,” (1995, p.56). The creation of classifications however prompts the production of a theoretical scheme that is suitable to the data. The scheme assists the researcher to derive questions from the data, making comparisons of the data, making alterations to the categories, eliminate some categories which are unnecessary, and arranging the categories in a hierarchical manner. It is important to recognize two separate, although interrelated, stages of data coding: one stage concentrates on the meaning of the data within the research, while the other concentrates on the meanings of the data that may be useful to external readers (Gough and Scott, 2000).
Codes used in qualitative research can either be objective or heuristic. Objective codes are used to compress illustrations of the facts portrayed in the data. The code words are therefore used as a replacement for the data collected and the analysis can concentrate on the codes alone rather than on the text. Once this is done, the researcher can then follow conventional distributional analytic procedures to test the hypotheses. However, in order for this to be done the researcher should be able to have confidence in the code words. This can be achieved in three ways: ensure that the portion of text replaced by a code word is a clear-cut occurrence of the true representation of the codeword; ensure that the code words are used consistently in the text, and ensure that every occurrence of the representation of the code word is identified. On the other hand, heuristic code words are used by the researcher to collect observed things so he can subject them to further investigation. Heuristic codes assist the researcher to restructure the data and gaining diverse opinions of the collected data. They make possible the detection of new information because they change and develop as the research process continues. In the process, “heuristic code words change and transform the researcher, who, in turn, changes and transforms the code words as the analysis proceeds,” (Seidel, 1998, p.14).
The coding of qualitative data often results in compressed data that is easy to manage and analyze. Categories created during coding are not viewed in isolation from each other. Instead, each category, though different from the rest, is interrelated and interlinked with all the other categories. This helps the researcher to make sense of the data in the light of the problem under investigation keeping in mind the social viewpoints of the participants. When researchers create categories, they are in the process of making resolutions about the organization of the data in a useful manner that will assist the analysis process. The categories, therefore, have to be incorporated into the broad analytic framework. Coffey and Atkinson argue that “codes are links between locations in the data and sets of concepts or ideas, and they are in that sense heuristic devices, which enable the researcher to go beyond the data,” (1996, p.37). There are two analytic processes that are fundamental to the process of coding although their character varies according to the types of coding adopted. The first involves making comparisons between the data and categories, while the second one involves making inquiries from the raw data. The names given to categories can be derived from the concepts and keywords that the researcher comes across while conducting a literature review. Alternatively, they can come from the terms and expressions utilized by the participants of the study.
Researchers can approach the process of coding in two ways. The first approach involves first collecting data then coding. The researcher first collects the data from the informants without having any predetermined list of categories. Once the data has been collected, the researcher then goes through them trying to find the suitability of each data to the study’s context and the relationships that exist between the data. The researcher then creates categories from the data and fits each data into its most suitable category. This approach is the basis of the grounded theory that was initiated by Glaser and Strauss in 1967. The second approach is the reverse of the grounded theory. The researcher first creates a list of categories that he wants to conduct the study with. The researcher then proceeds to collect data from the informants while in the process trying to fit the collected data into the predetermined categories. The predetermined list of codes is often created from the theoretical framework of the study, the literature review, research hypotheses, and questions (Basit, 2003, p.145).
Research proposal: Diversity management in American healthcare organizations
Ethnic diversity is one of the issues that the majority of healthcare organizations in the United States have to tackle on a daily basis (Dreachslin and Hobby, 2008). This is due to the increasingly diverse nature of the American population that has resulted from high numbers of immigrants from different parts of the globe. Many communities in the U.S. are continuously experiencing an upsurge in members of ethnic minority groups such as African-Americans, Chinese Americans, Japanese Americans, and European Americans among other communities. Diversity has indeed affected every institution in the nation with the healthcare sector being one of the most affected sectors. Diversity in healthcare organizations affects not only patients but also the staff and managers of such organizations (Whitman et al., 2008, p.26). Differences in language, cultural beliefs, religious beliefs, and attitudes affect the access of healthcare services, the way in which healthcare services are dispensed, and ultimately the outcome of healthcare services. Hobby and Dreachslin argue that “when diversity issues are not understood, valued and appreciated for their impact on the delivery of patient care, the healing process and communication/trust, they become contributors to disparities and unequal medical outcomes,” (2007, p.6). It is against this background that this current study is based. The study will examine the socio-economic, economic, and cultural factors which affect the delivery of healthcare services across different ethnic groups in the United States.
The study will aim to address the following questions:
- Does the level of education affect a patient’s access to healthcare services?
- Does the level of income and wealth affect a patient’s access to healthcare services?
- Does the type of occupation affect a patient’s access to healthcare services?
- How do contradictory cultural beliefs affect the delivery of healthcare services?
- How does the American healthcare financing system affect patients of ethnic minority groups?
A qualitative study will be conducted to examine the issues raised above. The justification for conducting a qualitative study rather than a quantitative study lies in the objectives of the study. The study is not interested in finding out how many socio-economic, economic and cultural factors affect the delivery of healthcare services. It is also not interested in finding out how many patients from ethnic minority communities are negatively affected by the American healthcare system. Instead, the study wants to understand deeply the issues that patients from ethnic minority communities go through when they try to access healthcare services. It wants to gain a deeper understanding of the factors that affect the delivery of health services to ethnic minority populations and how such effects come about.
The participants of the study will be drawn from four different ethnic communities in New York State: Caucasian, African American, Mexican American, and Chinese American. Stratified sampling will be used to select the participants to ensure that the four ethnic communities are well represented. Stratified sampling will also be used to ensure that the participants belong to different socio-economic statuses. For instance, it will ensure that members with different education levels, different English language proficiency, different income and wealth levels, different occupations, and different health insurance packages are all represented.
Data collection techniques
To conduct the study, two major techniques will be used: in-depth interviews and focus group discussions. The in-depth interview will be conducted with the use of a semi-structured questionnaire. The questionnaire will contain both closed-ended and open-ended questions to allow the researcher to gain more information necessary for the study from the informants. The interview will not be conducted in any organized manner. Instead, the researcher will choose to ask questions in an order that he seems suitable to the informant depending on the direction the interview will take and on the responses given by the informants (Banister et al., 1994).
Focus groups are “an informal assembly of participants whose opinions are requested about a specific topic,” (Crabtree and Miller, 1999, p.56). Focus groups as a method of data collection will classify the participants into smaller groups according to their ethnic origin. This means that the Caucasians, African Americans, Mexican Americans, and Chinese Americans will be grouped separately. The researcher will then conduct discussions with each of these groups to enable him to gain a deeper understanding of the factors that affect each of the participating ethnic communities as far as access to health services is concerned.
In-depth interviews and focus groups are relevant to this study in a number of ways. First, the two methods require that the researcher should have a healthy relationship with the informants. The informants should be able to trust the researcher while the researcher should have respect towards the informants irrespective of their conflicting beliefs (Coolican, 1994). This is because in-depth interviews and focus groups discussions are conducted through a close and personal interaction between the researcher and informants. Second, the two methods allow the researcher to clarify any vague responses given by the informants. They also enable him to dig deeper and probe further when he feels that the responses given are short or incomplete and that the informant is holding back useful information. The informants also have the opportunity to provide additional information that they feel would be appropriate to the study (Patton, 1980). This is useful in any qualitative study because its main objective is to understand the thoughts, feelings, experiences, and opinions of the participants. This can only happen through a deep and extensive interaction between the researcher and the informants.
Rigor, validity, and reliability of the methods
Rigorous research is one that uses tools and techniques that are fit to meet the study’s objectives (Bernard, 1996, p.10). The tools used in this study will guarantee the rigor of the study in various ways. First, the methods used to collect data, that is in-depth interviews and focus groups will produce data that is appropriate enough to answer the research questions. This is because, in both methods, the researcher will be actively present while gathering data from the informants. The researcher will therefore have the power to direct the interviews and discussions in a manner that will produce relevant information for the study. Second, the methods used to collect data will produce the suitable level of detail necessary for answering the research questions. In-depth interviews and focus groups allow an extensive exploration of the issues under investigation through clarification and probing of information. As discussed earlier, the researcher can probe the informants for additional information about a particular issue. The informant is also able to provide more information than initially enquired. As a result, the two methods will enable the researcher to gain as much information as needed for the study. Lastly, the rigor of the study will be ensured by explaining in clear and simple language to the colleagues of the researcher and other users of the study’s findings of the process of the study. This would entail explaining why and how the data collection tools were selected and the entire procedure followed in collecting, coding, and analysis of the data (Ryan, p.5).
The meaning of reliability and validity of qualitative research differs from that commonly used in quantitative research. This conflict in the applicability of these two terms has been addressed by several scholars. Stenbacka (2001), for instance, argues that “the concept of reliability is misleading in qualitative research. if a qualitative study is discussed with reliability as a criterion, the consequence is rather that the study is no good,” (p.552). In qualitative research, reliability and validity are gauged by measures such as credibility, confirmability, consistency, and dependability. Credibility and confirmability are measured by the degree of the truth with which the researcher reported the data collected from the informants (Strauss and Corbin, 1990, p.250). Qualitative research requires that the researcher should have an open view of the opinions of the informants even when their outlook contradicts the researcher’s outlook. It is not his duty to report what he thinks should be the responses of the informants but rather what exactly the responses are. This requires the researcher to identify with the informants and actually empathize with them by sharing in their own world. The two methods that will be used to conduct the study will provide the researcher with this opportunity because the researcher will have time to actively interact with the informants. This will ensure the credibility and confirmability of the study. On the other hand, the consistency and dependability of qualitative research depend on the procedures used to code and analyze the data. In this study, the same method will be used to code and analyze all the data collected to ensure that the results of the study are consistent and dependable.
Data analysis approaches
The analysis of the data collected in the study will be conducted using several steps. The transcripts from the interviews and focus group discussions will be analyzed through thematic content analysis using a mixed coding chart. The themes will be derived from the research questions and conceptual framework. In the first phase, data reduction, the transcripts will be read and the text coded sentence by sentence in order to identify the main themes presented by the informants. In the second phase, data display, the themes identified by the informants will be classified into a conceptually clustered matrix (Miles and Huberman, 1994). The cross-case analysis will then be used to establish any existing relationships between the themes and to identify converging (themes that are commonly identified and shared by the different groups of participants), diverging (themes that are commonly identified by the informants but whose application differ across the different groups of participants), and marginal themes (themes that are sporadically identified by some of the participants). The final phase will involve drawing conclusions, making inferences, and providing recommendations to the healthcare organizations based on the findings. In this stage, the interrelationships between the converging, diverging, mirror, and marginal themes will further be examined and studied again in order to identify the major factors that influence the delivery and access of healthcare services by members of the ethnic minority communities. Conclusions will then be drawn from these interrelationships and recommendations made appropriately in order to solve the problem of inequality in the healthcare sector.
Triangulation of the findings
Triangulation involves the use of multiple methods, data sources, and analytic techniques to enhance the reliability and validity of research findings. Golafshani (2003) states that “engaging multiple methods and data sources will lead to more valid, reliable and diverse construction of realities,” (p.604). In this study, triangulation can be done in several ways. First, the researcher can increase the sources of the data by using participants from different states other than New York, by collecting data from other partners of the healthcare system such as medical personnel and managers rather than just patients. Secondly, the researcher can use other data collection methods in addition to the two discussed earlier. Observation can be a useful method of collecting data whereby the researcher observes the interaction between healthcare providers and patients from different ethnic communities. This will enable the researcher to see for himself exactly how ethnicity influences the delivery of healthcare services instead of just relying on the informants’ words.
Ethical issues involved in qualitative research
The nature of qualitative research methods is that the researcher must almost always have close interaction with the informants. This can pose a number of social or psychological problems to the informants. As a result, various ethical principles must be upheld when conducting qualitative studies (Richards and Schartz, 2002). These principles include:
- Informed consent – the researcher should provide the potential participants with detailed information concerning the nature of the study, its benefits, potential harm, and the rights of the participants. Based on this information, the participants can choose to participate or withdraw from the study. The consent given by the participants can be in written, verbal, or taped form and it binds the informants to the study.
- Confidentiality – the researcher should not record any personal information of the informants that could damage their reputation or cause them any problem. Information that should be made anonymous include names, addresses of location, and sensitive medical information.
- Protection from harm – the researcher should ensure that the benefits of participating in the study far outweigh the risks. Even then, it is the responsibility of the researcher to protect the participants from inevitable harm, for instance, psychological distress when dealing with a sensitive issue.
- Feedback of findings – it is the responsibility of the researcher to provide the informants with a copy of the study’s findings once the study is completed. The researcher should also recognize the important role played by the informants in the success of the research and should thank them for their participation (Crabtree and Miller, 1999).
- Banister, P., Burman, E., Parker, I., Taylor, M., and Tindall, C. (1994). Qualitative methods in Psychology: A research guide. Buckingham, UK: Open University Press.
- Basit, T.N. (2003). Manual or electronic? The role of coding in qualitative data analysis. Educational Research, 45(2), 143-154.
- Bernard, H.R. (1996). Qualitative data, quantitative analysis. Cultural Anthropology Methods Journal, 8(1), 9-11.
- Betancourt, J.R., Green, A.R., Carrillo, J.E. and Park, E.R. (1999). Cultural competence and healthcare disparities: Key perspectives and trends. Health Affairs, 24(2), 499-505.
- Coffey, A. and Atkinson, P. (1996). Making sense of qualitative data. London: Sage.
- Coolican, H. (1994). Research methods and statistics in Psychology (2nd ed.). London: Hodder and Stoughton.
- Corbin, J. and Strauss, A. (1990). Basics of Qualitative Research: Grounded Theory, Procedures and Techniques. Newbury Park, CA: Sage.
- Crabtree, B.F. and Miller, W.L. (1999). Doing qualitative research. Thousand Oak: Sage Publications.
- Dreachslin, J. and Hobby, F. (2008). Racial and ethnic disparities: Why diversity leadership matters. Journal of Healthcare Management, 53(1), 8-13.
- Golafshani, N. (2003). Understanding reliability and validity in qualitative research. The Qualitative Report, 8(4), 597-607.
- Gough, S. and Scott, W. (2000). Exploring the purposes of qualitative data coding in educational enquiry: insights from recent research. Educational Studies, 26, 339-354.
- Hobby, F. and Dreachslin, J. (2007). Diversity and disparities: Parallel challenges for 21st century healthcare. Bridges, 13(3), 5-6.
- Miles, M.B. and Huberman, A.M. (1994). Qualitative data analysis: an expanded sourcebook (2nd ed.). Thousand Oak: Sage Publications.
- Patton, M.Q. (1980). Qualitative evaluation methods. London: Sage.
- Richards, H.M. and Schartz, L.S. (2002). Ethics of qualitative research: are there special issues for health service research? Family Practice, 19, 135-139.
- Ryan, G.W. What are standards of rigor for qualitative research?
- Seidel, J., and Kelle, U. (1995). ‘Different functions of coding in the analysis of textual data’. In: Kelle, U. (Ed). Computer-aided Qualitative Data Analysis: Theory, Methods and Practice. London: Sage.
- Seidel, J. (1998). Qualitative data analysis.
- Stenbacka, C. (2001). Qualitative research requires quality concepts on its own. Management Decision, 39(7), 551-555.
- Strauss, A. and Corbin, J. (1990). Basics of qualitative research: Grounded theory procedures and techniques. Newbury Park, CA: Sage Publications, Inc.
- Whitman, M.V., Davis, J.A., Walker, A., Barrett, C.C. and Battista, A. (2008). Cultural and linguistic competence in healthcare: The case of Alabama General Hospitals. Journal of Healthcare Management, 53(1), 26-40.