Coding 1 Codingindexingcategorizing N B confusion because used

Coding 1

Coding/indexing/categorizing N. B. confusion because used in quantitative data where it means putting numbers to answers. n “indexing” “categories” “codes” “themes” = linking chunks of data (text) as representative of the same phenomenon. n Not necessarily to count them (cf. Content analysis) 2

Analysis. Bryman suggests these stages Stage 1 n Read the text as a whole, Make notes at the end n Look for what it is about n Major themes n Unusual issues, events etc n Group cases into types or categories (may reflect research question – e. g. male and female) 3

Stage 2. Read again Mark the text (underline, circle, highlight) n Marginal notes/ annotations n Labels for codes n Highlight Key words n Note any analytic ideas suggested. n 4

Stage 3. Code the text n n n Systematically mark the text Indicate what chunks of text are about – themes – Index them. Review the codes. Eliminate repetition and similar codes (combine) Think of groupings May have lots of different codes (Don’t worry at early stage – can be reduced later) 5

Stage 4. Relate general theoretical ideas to the text. Coding is only part of analysis n You must add your interpretation. n Identify significance for respondents n Interconnections between codes n Relation of codes to research question and research literature. n 6

Thematic Coding n Grounded Theory (Glaser and Strauss + Corbin + Charmaz) Interpretative Phenomenological Analysis (Jonathon Smith) Template analysis (Nigel King) Framework analysis (Ritchie and Lewis) n All are types of thematic analysis. n n n 7

How is coding done? Text In a village like this. . . the young fellows in the village don't seem to have much difficulty when they're out of work – a fortnight and they're back again – word of mouth, I'd say. It’s a different, tricky situation that I'm in – I just can't say, “Oh, I heard there's a job going on building site, I’ll go and have a go for it. ” I wouldn't be able to do that. Code Age contrast Residence focus Young find work easily Word of mouth Contrast situation Constrained 8

Applying the codes to the data n Need to take code and its definition and apply in standard way to the text. Identify chunks of text to which code applies n Can be phrases, sentences, several sentences or even paragraphs n Coded passages may overlap n 9

Questions to ask n n n "What is going on? What are people doing? What is the person saying? What do these actions and statements take for granted? How do structure and context serve to support, maintain, impede or change these actions and statements? " (Charmaz 2003: 94 -95) 10

What can codes be about? Lofland suggests: 1. Acts – usually brief events 2. Activities – of longer duration in a setting, people involved 3. Meanings – what directs participants’ actions? a) b) 4. 5. 6. What concepts they use to understand their world What meaning or significance it has for them. Participation – People’ involvement or adaptation to a setting Relationships – between people, considered simultaneously Settings – the entire context of the events under study 11

What can codes be about? 2 Strauss suggests n Conditions n Interactions n Strategies and tactics n Consequences What happens if… Sabatier in Policy context suggests: n Causal adequacy n Financial resources n Legal/bureaucratic powers or constraints n Political/interest group support n Official/bureaucratic commitment n Social/economic environment 12

What can codes be about? 3 Mason suggests n Literal - words, dialogue used, actions, settings systems etc. n Interpretation - implicit norms, values, rules, mores, how people make sense of phenomena n Reflexive - researcher’s role in the process how intervention generated the data. 13

Ways to identify themes Ryan and Bernard (2003) n Repetitions n Indigenous typologies (in vivo) n Metaphors and analogies n Transitions (pauses, sections) n Similarities and Differences n n Linguistic connectors n n Constant comparison Because, before, after, next, closeness, examples Missing data (what is omitted) 14

Coding supports 2 forms of analysis Retrieval n Using the coding frame n 15

1. Retrieval n n n Retrieve all the text coded with the same label = all passages about the same phenomenon, idea, explanation or activity - Literally cut and paste Used envelopes/files - Now done using software Enables cross case comparison on same theme. 16

2. Using the coding frame n Use the list of codes to examine further kinds of analytic questions, e. g. relationships between the codes (and the text they code) n grouping cases n 17

Data driven or concept driven? n n n n Inductive or deductive Most qualitative analysis does both i. e. start with some theoretical ideas these derived from literature, research brief/questions, interview schedule and discover new ideas, theories, explanations in the data. Strauss - sociologically constructed codes vs. in vivo codes 18

Example ‘Dancing’, ‘Indoor bowling’, ‘Dances at works club’, ‘Drive together’ ‘Joint activities ceased’, ‘Joint activities continuing’ ‘Loss of physical co-ordination’, ‘Togetherness’, ‘Doing for’, ‘Resignation’, ‘Core activity’ Descriptive codes Categories Analytic codes 19

Example showing coding marks 20

Code list = Code scheme = Coding frame = Template List of codes with definitions n Separate from the documents n May be hierarchical Used: n To apply the code in a consistent way. n To share codes with others, especially in a team n 21

Code Definitions Typically records: 1. The label or name of the code. 2. The name of the researcher. (Not needed if you are working alone. ) 3. Date when coding was done or changed. 4. Definition of the code. Analytic idea it refers to. 5. Other notes about the code, e. g. 1. 2. ideas about how it relates to other codes a hunch that the text could be split between two different codes. 22

Coding hierarchy Codes can be arranged in a hierarchy e. g. with these codes from a study of friendship n n n n n Close, generalised friendships Sporting friendships Sports club members Work friends Making new friends - same sex Making new friends - different sex Losing touch with friends Becoming sexual relationships 23

Example code hierarchy n Friendship types n n Close, generalized Sporting • Club • Non-club n n Work Changes in Friendship n Making new friends • New same sex friends • New different sex friends n n Losing touch Becoming sexual relationships 24

Code Big or Small Chunks Pro Wide/ high Maximize usefulness of level code - applied to enough chunks to justify recontextualization. Avoids prejudicing later analysis Con Few episodes can be identified to match code. Includes lots of less relevant material Coding vague Narrow/de Greater differentiation. Important contextual tailed Clear definition. Easier data may be lost. to identify chunks in text Loss of meaning. Too many codes to remember 25

Example showing analysis n n n One of a set of interviews by Wendy Hollway and Tony Jefferson. On fear of crime Will use some of this for a group work exercise. Part of interview with: Barbara 65, F, White, Retired nursing auxiliary, Interview covered, Husband's death, ill health, sister - prison, stealing & drug taking, tenants association. From low crime area. 26

INT So you say - well 2 of those things happened after - when you've been talking to this accountant friend of yours. How did it come up? I mean that's er, you'd been alone for quite a while. . BARBARA They'd been burgled. INT Right. BARBARA And they got through a little window like this. Actually 'e'd got a young lad with 'im. And er, Margaret's engagement ring and she says "that was the one thing - that was the one thing, it grieved me more than anything" she said. "They could 'ave the television, the lot" she said. But the fact that they took 'er engagement ring… INT Yeah. BARBARA That upset 'er. And er, we were just talking in general and - and it came up and I says er, "I've got a chain on my door. " And 'e says er, "it's not strong enough that, Barbara. " He says "you really want something else on" and 'e went - his daughter lived up Stokebridge and 'e went to a little shop up there, or something. And got me that chain… 27

BARBARA …And 'e put it on and you can lock it. If you put it on as you're going out, er, its 'ook, and then you 'ave to unlock it to let it drop. INT Ah ha. BARBARA When you come in. INT Oh right. BARBARA You know, you can push the door and it - oh and it is strong as well. INT Ah ha. And the 4 locks on the back? Do they date back further? BARBARA Oh God, yeah. INT So you had lots of security even when your husband was alive? BARBARA Oh yeah, mmm. Mmm. Em, I've got one of those dead locks at the top. INT Yeah. BARBARA You know, they're just a hole in the door and they're not from outside, they're only from inside. And even that locks wrong way. You 'ave to turn it that way to unlock it. (laugh). 28

Notice Interviewer and respondent names are in capitals n Wide margins and space and a half between lines n Use of contractions n Place names and people’s names anonymised n 29

Read through About neighbour being burgled n Lost TV etc. and engagement ring n Old and new security on front door. n Replaced by friend. n 30

Mark up text n Annotations and codes. 31

32

33

Coding Frame n Crime experienced (the type of crime participants discuss having experienced themselves or by their friends and neighbours). n n Burglary Vandalism Violence But these descriptive. Be analytic. E. g. n n Low level (not reported etc. ) Significant (with emotional impact) 34

Coding Frame, cont. n Security measures (What measures people have taken to protect themselves, their property etc. both in the past and more recently). n n n n Chain Dead lock Burglar alarm Safe Car alarms Personal Alarm Stay in Walk with others 35

Coding Frame, cont. n But these descriptive. Be analytic. E. g. Physical, technology n Behavioural n Psychological (lights on timer etc. ) n 36

Coding Frame, cont. n Feelings about experience of crime Frightened n Hurt by loss (especially personal items) n 37

Descriptive vs Analytic/theoretical n Descriptive Just what the people said n What happened n Their terms n n Analytic Use social science theory n Groups codes together n Use terms the respondents don’t or wouldn’t n 38