Introduction to Qualitative Analysis and writing up your analysis: Thematic Analysis. Dr. Shaun Ruggunan Discipline of Human Resources Management

Transcription

1 Introduction to Qualitative Analysis and writing up your analysis: Thematic Analysis Dr. Shaun Ruggunan Discipline of Human Resources Management

2

3 Why Thematic Analysis Qualitative Analysis is varied and nuanced. Discourse, Narrative, phenomenological but common across most qualitative analysis is some form of Thematic Analysis. So if we get the foundation right then it becomes more manageable regardless of the theoretical approach or method you use.

4 Structure of Presentation The best way to do this is to talk through some of the qualitative projects I have worked on and supervised. So some bias towards my experiences. So not everything can be covered in detail but you will have a sense of how to structure your methodology chapter and organise your discussion.

5 Writing your methods Chapter A key heading in your methods chapter needs to be philosophy of research or something similar. This is missing 90 percent of the time from positivist/quantitative projects esp in management studies. Statistics and statistical analysis are not atheoretical or neutral. Say more than you are adopting a qualitative approach. Its about voice, theory, argument.

6 Use the correct terms Qualitative methods and Analysis has a technical language of its own: USE IT! Common sense versus academic/scientific terminology. For example story telling, interviews, coding, axial coding, hierarchical coding, nodes, patterns, themes have very specific meanings. This means you have to read.

7 Some Terminology that students should use but don t use in writing Data Corpus Data Set Axial coding Open coding Hierarchical coding Latent themes Semantic meanings etc

8 Using the right language So please pay careful attention to the technical language that you use in your proposal, methods chapter and discussion chapter. I d like to see more than generic references to coding and patterns.

9 Presentation of your Sampling Ruggunan,2013 Table 1 Participants Demographics Gender Race Sector 2 female 1 White South African 7 Private Sector 14 male 2 Black South African 9 Public sector 13 Indian South African

10 Sampling example 2 Groh et al, 2011

11 Sampling example 3 Ruggunan 2013

12 Sampling example 4 (Sidat,2007)

13 Writing up your sampling section Must have a narrative. Detailed description as possible. Most journals I have published in have asked for a tabular representation as well. Percentages not always useful with small numbers. No one perfect tabular representation.

14 Thematic Analysis: Method or Tool? You need to know what your data corpus is. Then determine your data set. HRD of South African seafarers: Data Corpus is interviews, websites, labour market reports, focus groups, notes made at maritime conference, youtube interviews with key stakeholders. Data set 1 are all my interviews only, data set two may be my focus groups only etc. Really important for big projects. Eg. Medical lab specialist project.

15 Thematic Analysis Unlike quantitative research, where survey instruments can be administered effectively by third parties or electronically, I suggest that you do as much of your qualitative data collection by yourself. Esp interviews, focus groups Exceptions may be where ethical conflicts exist. Rationale is that it makes subsequent steps easier but not always possible with larger samples, global research sites etc. Use your discretion. Aim is data reduction

16 Step. 1 Familiarise yourself with the data If you have done the data collection on your own you still need to re listen to audio and read transcripts a few times. Easier to do with 15 to 20 interviews than 80 but it needs to be done. If you outsource transcription please double check with audio tape! Biggest project I ve been in had about 86 interviews but there were three of us but had to read all anyway.

17 Step 2 Generate initial codes Code interesting features of the data in a systematic fashion across the entire data set. (semantic or latent) Collate/ organise data relevant to each code. Code manually or use software. Be strategic in your choice. Your interview schedule questions are NOT your codes. You can either go in blind or you can read the literature on your topic first. Related to theoretical approach Eg. Seafarers (blind) medical lab specialists (read theory first first) Data driven or theoretical driven

18 Code for as many potential themes. Seafarer and nationality/race/ethnicity example. Professional status/ hierarchy example Code with context Extracts can be multiply coded Code minority and contradictory features You are trying to establish patterns from your codes.

19 Medical Specialists example And I think one of the biggest [challenges] is the remuneration because the public sector cannot compete with the private sector (Interview: Participant 1) Pay is lower and they [are] struggling to retain people in the public sector (Interview: Participant 3). it is common sense to move out of the public sector, as the private sector pays higher salaries than the public sector. There are attempts by the public sector to improve salaries but at this point it s far, far more lucrative to be employed by the private sector. (Interview: Interview: Participant 5). Coded for 1. pay differentials(semantic), 2. Tension between publicprivate sector (latent)

20 Theory driven coding The code pay differentials eventually developed into the theme REMUNERATION Remuneration was identified as pull factor from the public sector to the private sector. Fitted into the debates on push pull theory in the literature which identifies remuneration as a known push factor amongst health professionals.

21 Data Driven coding Sometimes we don t know much about an empirical case or we want to develop our own model or theory from the data. Seafaring example Seafarers identity instrumental to employment practices. Added new element to diversity perspectives in HRD

22 Step 3.Discovering themes/searching for themes Active process. Please write in an active way. Don t use passive themes emerged. Broader level of analysis. Codes become themes/sub-themes Software/mind maps/ visualisation good for this. Iterative/intuitive Trustworthiness? Multiple coders? examples

23 4. Reviewing Themes You now have candidate themes which have to be reviewed. They are candidates because some may not have enough data to support them for example, or you may want to merge or separate themes or even remove certain themes. You want to reduce data further You want internal coherence in a theme and strong distinctions between themes.

24 Examples of themes

25 Themes examples

26 Themes examples Pull Factors from the public to the private sector (n=16) Reasons to migrate out of the state sector Number of interviewees who cited this reason Percentage of interviewees who cited this reason Remuneration Working conditions Autonomy of work Flexibility of working time Career pathing

27 Step 5 Defining and naming themes So this is about describing the themes in a way that captures the essence of the theme. Don t demand too much from a theme. Maybe Remuneration is too broad a theme and can be defined differently for example? Sub themes can be named here (private remuneration/public remuneration etc) Can you describe your theme in one or two sentences? That s the test.

28 Step 6 Writing the Analysis Best way to learn this is to read peer reviewed qualitative articles. To read journals about Qualitative research. Do not summarise each interview. You are not a journalist so do not report facts only. You need to provide an analytic narrative

29 Writing for a journal Some journals require a findings section then a discussion section. This format may be useful for your thesis as well. Avoid Results opt for findings Under findings you provide the themes you have discovered with examples of extracts that have been coded for that theme. You can do this in a matrix form or a narrative form.

30 Writing your findings Career pathing: Histopathologists have better career pathing in the private sector A common thread through the various narratives of the participants was their implicit and explicit references to career pathing and career mobility. Their narratives demonstrate respondents beliefs that the public sector offers minimal career pathing opportunities compared with employment in the private sector. That s an interesting question my experience has been that unless I wanted to become head of department there was no other career path available to me. Now that I am in private employment, I can decide to have a career track towards partnership, or focus more on marketing of the laboratory, or pursue research based projects on my own. I m still young so I haven t decided how I want to develop but there are for sure many many options available to me here [in private sector] than at my previous employer. (Interview: Participant 4) Long-term career prospects are better in the private sector than in the public sector. As one participant from the public sector expressed: They have frozen posts...the people who qualified are still stagnant, not moving up- no job offers. So, if jobs are frozen, people are going to leave. They are not going to stay at that level- registrar or whatever level because they are now qualified. There is a need for more consultants but they are not opening it up. So there is an exodus at the moment. People are leaving and it is a lot (Interview: Participant 15, 2011).

31 So writing up your findings is more descriptive of your data than your discussion. BUT your discussion needs to be analytical.

32 Discussion of the finding on Career pathing Participants accounts also revealed a concern with career pathing and the lack of career pathing opportunities in the public sector. Allsop et al [31] emphasise that the decision to migrate from one employer to another is often associated with reasons such as the desire to gain additional experience. For example, Maistry (cited in, Hudson, 2011: 22) reports that for a cytologist, returning to South Africa after having worked three years in Saudi Arabia meant taking a few steps back professionally. Hudson [44] adds that one of the reasons that newly qualified specialists move into the private sector or leave the country is the unavailability of consultant posts in the public sector. It is evident from the interview data that histopathologists need to have a sense of a future career path. That path could develop into several routes, including but not only a managerial route, research route, or a more operational route. Histopathologists want career choices or tracks to be available to them. Participants expressed a desire to grow professionally beyond their diagnostics skills and roles. It is frustrating for them that these career pathways are not available in the public sector. This is compounded by their having no control over their careers in the public sector. If histopathologists are to be retained in the public sector, then clear career pathing opportunities need to be made available for them. While in Ashmore s 2013 [36] study medical doctors cited team work, more academically rewarding work and a sense of service, relevance and contribution to the greater public good as reasons why they remain in the public sector, no such sentiments were echoed in the accounts offered by the histopathologists interviewed for this study.

33 Thesis writing Since you don t have variable here as in quantitative research you may want to divide your discussion chapters according to the themes you have discovered. In this case there is no separate findings chapter, but a series of discussion chapters based on your themes. How are your themes similar to and different from the literature? Example of professional status/gendering of work You are doing interpretive work not journalistic reporting.

34 Rigour You have to provide an explanation of rigour in your data analysis process. Meaning versus objectivity or whether participants tell the truth. Triangulation/ verification Credibility, transferability (not always possible or needed), dependability (explain in detail the research process, reflective commentary, confirmability (audit trail)

35 PRESENTING QUALITATIVE DATA TYPE OF QUALITATIVE DATA COLLECTION Quotations Simple Tables Matrices FORMS OF DATA PRESENTATION Maps Diagrams Taxonomies Scales (Single...MDS) Conceptual ETHNOGRAPHIC MODELS Behavioral Conceptual- Behavioral Unstrucutred Direct Observations ** **** **** ***** *** *** Structured Direct Observations **** **** ** ***** **** Key Informant Interviews ***** **** **** *** **** * ***** ** ** Focus Groups ***** **** **** *** ** * **** *** *** PRA Group Techniques **** **** ***** ** *** * ** ** Free Lists ** ***** ** * * Pile Sorts Triads ** ** ** ***** ***** *** Paired Comparisons *** *** ***** *** less complex * less useful most useful ***** more complex

36

37 Matrices

38 Figure 3 Method Matrices Reasons to Smoke Outside the Workplace Type of Information obtained Conclusion Next Investigative Step Key Informant interview general information regarding people's perception of smoking outside the workplace some people clearly see they are addicted, others claim naiveity Observation of the activity Direct Observation varied appearance of the activity, not clearly seen as relaxing when hurriedly done outside the workplace confirms addictive behavior gather more information on places people smoke besides outside the workplace Free Lists names for places where people smoke clearer understanding of reasons people smoke outside the Workplace obtain information on how much people smoke in different locations Pile Sorts Relative frequency of where people smoke People smoke more in confined locations than before and expose others and themselves to side strea smoke See how the frequency of cigarettes smoked has changed as a result of the ban Paired Comparisons Changes in locations of where people smoke as a result of the ban Smoking outside buildings has increased exposure of people to side stream smoke NOT born out by our small sample size, but warrants further investigation Confirmation by Focus Groups, and consideration of cessation attempts Focus Group Discussions Plan an intervention to help those smoking outside buildings to stop