Coding 2: The Mechanics
Phase 2: Hearing What Was Said
6
Coding 2: The Mechanics
Phase 2: Hearing What Was Said
IN THIS CHAPTER we will consider the second phase of data analysis, HEARING WHAT WAS SAID. In this phase you will gain more access to the subjective experience of the research participants, by organizing the relevant text into repeating ideas (Step 3) and the repeating ideas into themes (Step 4).
Step 3: Discover Repeating Ideas by Grouping Together Related Passages of Relevant Text
Working on Each Transcript Separately
As you select the relevant text you will notice that different research participants are often expressing the same idea, sometimes with the same or similar words. These are called repeating ideas.
A repeating idea is an idea expressed in relevant text by two or more research participants.
Repeating ideas are the beginning building blocks from which you will eventually assemble a theoretical narrative.
In Step 3 you systematically search the selections of relevant text for repeating ideas. You first identify the repeating ideas in each separate transcript (from the focus groups or individual interviews, depending on your methodology). Then you combine the repeating ideas from all of the transcripts into a composite list for the entire research sample. As principal investigator, you first do this by yourself, and then combine your work with that of other coders.
1. Begin by opening the file that contains the list of relevant text for Group 1 (or Individual 1). Then create a new file that will become the list of repeating ideas for Group 1.
2. Now highlight and copy the first selection of relevant text from your relevant text file to your repeating ideas file. We call this first selection the starter text.
a. Read through the entire list of relevant text selections, keeping the starter text in mind.
b. Each time you encounter an idea that seems related to the starter text, highlight it and copy it into the repeating idea file. As you do so, you might want to make a note about how the two selections of text seem related. If you made memos, move them with the text selections.
c. Read down the list until you have highlighted and moved all the relevant text related to the starter.
d. After you have grouped together all of the selections that seem related to the starter text, return to your original list of relevant text. Highlight and move the first selection of text that you did not group with your original starter text. This new selection becomes your next starter idea, and you then read down the list of relevant text, highlighting and moving all the selections that relate to this starter. You repeat this procedure until you have grouped together as many of your relevant text selections as possible into your repeating idea file.
Table 6.1 gives you an example of a master list of relevant text, and Table 6.2 shows that same text grouped into two repeating ideas.
Table 6.2 |
||
Haitian fathers |
|
|
Repeating ideas |
||
Repeating Idea #1—Something about religion
(a) The first time I thought about becoming a father was very early in my life. Probably I, I guess because of my upbringing as I was brought up in Church and it was always a serious matter to me. I never went out with a girl without thinking that, this is the girl I could possibly marry. (L, p. 170)
Being a father is a serious matter, because it is connected with being religious.
(d) I was scared to become a father. As a Christian I was afraid not to meet the proper woman to become my wife in order to become a father. When I met my wife and realized that she was a Christian and looking behind at how my father raised us, I decided to become a father at that time. (AG, p. 170)
Fatherhood connected with religion.
Repeating Idea #2—Using one’s own father as a role model
(b) Fortunately, I had my father as an example I would say. He has been with my mother since I was little, and I would say he is still an example for me. (AG, p. 170)
One’s own father as a model to emulate.
(c) My father was a very good father. He is a guy who has justice; you cannot make him tremble in front of situations. (L, p. 170)
Expresses admiration for his father.
(e) Yes, yes when I think of the days of my youth, I can see my father and his dedication. The love that he has shown, and his hard working style and his honesty. All that, they left a serious impression on me. My dream was to look like my father.
Admires and desires to emulate his own father.
Orphan text.
(f) To me it became obvious that I would be a father as a married person. (C, p. 221) Being a father is the obvious and natural consequence of marriage.
Source: Zizi 1996.
As you can see, in Table 6.2, relevant text (a) was our starter text. We grouped (d) with it because they both related to the importance of religion. In this particular transcript, those were the only text selections related to religion. Thus, after grouping them together, we started again with selection (b) as our new starter text, and looked for other selections to group with it. We grouped selections (b), (c), and (e) together because they related to using their own fathers as positive role models.
We refer to selection (f) as an “orphan” because it does not go together conceptually with any other text selections. After we have grouped together as many relevant text selections from all of the transcripts as possible, we will decide whether to include or discard orphans.
We always use rather pedestrian language to name our repeating ideas when we first group text together, and thus named repeating idea #1 “Something about religion.” Then we peruse the text selections looking for an excerpt from the text that we can use as the repeating idea. Using the participants’ own words is more evocative of their subjective experience.
In Table 6.2 we underlined two sentences that we eventually used to name the two repeating ideas: 1. “I was brought up in Church and it was always a serious matter to me”; and 2. “My dream was to look like my father.” These statements capture the meaning of the text in the fathers’ own words. Sometimes you choose the name of a repeating idea as you go through each transcript. You might keep changing the name as you encounter quotes that seem more evocative than your first choice. Sometimes you do not decide on a name until you have gone through all of your transcripts and you have constructed your master list of repeating ideas.
You continue going through your list of relevant text for Group 1 until you have assigned all the relevant text to a repeating idea, or at least as much of it as you can. Because you have highlighted all of the text that you used, you can easily see what text is left over.
Having gotten this far, you will have more of a sense of how the text goes together than you did when you started. Also, several problems may arise, which you can use your newly developed sense of the text to resolve.
Orphans
After you have worked through the list of relevant text, you may be left with some relevant text segments that did not get repeated. You have several choices about what to do with these selections. You can discard the text as unimportant; you can search your transcripts again to try to find text that goes with your solitary text; or you can decide the text is important even if you cannot find other text with which to group it. Sometimes it is important to reflect differences in experience as well as commonalities. Thus, it may be important to report that only one person had a particular experience. Because qualitative research is not focused on quantity, individual differences have an important place in this paradigm.
Ideas Too Broad
When you go back and inspect your list you may find that you have too many items in a group, and can see fine distinctions between items that you originally grouped together. In this case break up your larger group into two smaller groups, each expressing a different repeating idea.
Ideas Too Narrow
Alternatively, you may find that your groups are too small; that you made too many distinctions and have too many ideas. In this case, merge two (or more) groups into a larger one.
Second Thoughts
You may read over your list and find some relevant text that you thought belonged in one place belongs somewhere else. Make the changes. Remember, it’s your theory.
When you have worked through the first transcript, create repeating ideas for each of the other transcripts. You may find that the ideas from one group alter your understanding of ideas from the other groups. Make a note of this in your research journal.
Creating Your Master List of Repeating Ideas
3. We now resume the material on page 55. After you have used as much as you can of the relevant text from all of your transcripts, combine the list of repeating ideas for each group into a file that reflects the master list of repeating ideas for the project. Go through this list idea by idea. When ideas are exactly the same, combine them into your final list. If ideas are similar, combine them as well. In the process you may have to rethink your groupings.
Just as you found orphaned relevant text, you may find a repeating idea in one transcript that does not appear in any of the transcripts of the other groups. Again, you have to make a judgment. If an idea that occurs in only one group seems important for your understanding of the phenomenon, you may decide to keep it. If it does not seem important, discard it. If you do not know what to do, set it aside temporarily and come back to it in later stages of data analysis.
How many repeating ideas should you come up with? There are no hard and fast rules, but we have found that something in the neighborhood of 60 is a workable number, with a minimum of 40 and a maximum of 80. The number that you actually come up with will depend on your cognitive style. If you tend to see fine distinctions everywhere, then your list will probably be large and you will have to reduce it. If you tend to think in broad patterns and see things as similar, then your list will probably be smaller and you will have to expand it. Again, try for about 60 ideas, with the understanding that you can go back and revise if it becomes necessary in later steps.
4. Name your repeating ideas. You may have found evocative quotes to name some of your repeating ideas as you were coding each transcript. Or you may choose the names at the point when you create your master list. Your goal is to choose a short quote that captures the essence of each repeating idea in a dramatic and emotionally vivid way. A good way to find a name is to go through your list and read the relevant text out loud. You may find a phrase that jumps out at you by virtue of its emotional impact, and that will be the name of the repeating idea.
Sometimes you may find that you need to combine two quotes, or give a slight paraphrase. That is fine. If nothing in the text seems to fit, then just give a brief statement of the idea. Remember that you are striving for both emotional impact and accuracy.
For example, we saw in Table 6.2 two repeating ideas composed from relevant text. The first repeating idea, “My dream was to look like my father,” expresses how admiration for one’s own father leads to a desire to emulate him. The second repeating idea, “I was brought up in Church and it was always a serious matter to me,” expresses the connection between religious morality and fatherhood.
The Collaborative Coding Process
We believe that all research should be conducted in groups rather than in isolation, particularly when doing qualitative research. We find that it is ideal to work in groups of four—two students and two more experienced researchers. Each coder goes through the process we have described above. Then the person primarily responsible for the project, the principal investigator, goes over all four sets of coding.
This process makes it more probable that we will not have overlooked any important ideas. Even if one or two of the coders has missed some important repeating ideas, using input from four people usually insures that no important ideas are lost.
After the principal investigator has decided on the repeating ideas from the master list, the results are presented and discussed with the other coders. The other coders may not agree completely with the organization; however, they must be able to see the rationale that the principal investigator used to include each piece of relevant text under each repeating idea. In other words, the organization of the data must be “transparent.”
This process often requires that the data be reorganized. As the master list is discussed, someone on the team will undoubtedly identify text that does not reflect the particular repeating idea under which it has been coded. That text segment will then have to be included under a different repeating idea, or discarded altogether. If it is coded elsewhere, including it with other text segments might mean changing the name of a different repeating idea so that it will be inclusive of the new text segment. This reorganizing and refining of the data requires patience.
When you have finished with this phase, congratulate yourself! You have completed the most difficult and labor-intensive part of the coding process. You have immersed yourself in the text, and acquired an enormous familiarity with the text, almost to the point of memorization. This will prove invaluable as you go on to develop theory.
Step 4. Organize Themes by Grouping Repeating Ideas into Coherent Categories
In the next step, you organize the repeating ideas into larger groups that express a common theme.
A theme is an implicit idea or topic that a group of repeating ideas have in common.
For example, a theme in the Haitian father data is:
Praising aspects of the traditional Haitian father
It includes the three repeating ideas:
“My dream was to look like my father.”
“There is no inch of laziness in my father.”
“I love the way my father treated my mother.”
Another theme is:
Dissatisfactions with aspects of traditional Haitian fatherhood
It includes the four repeating ideas:
“My father never said I love you.”
“Adults do not play.”
“When they say your father is coming, you run inside.”
“My father took care of other children and did not care much for me.”
These themes are shown in Table B.2 of the Haitian fathers article (see Appendix B). You will follow the same procedure for discovering themes that you used to create your master list of repeating ideas.
1. Begin by opening the file, which contains your master list of repeating ideas. Then open a new file that will be your themes.
2. Now direct your attention to the first repeating idea on the printed list, which we will call the starter idea for the first theme.
Read through the repeating ideas list, keeping the starter idea in mind. Each time you encounter an idea that seems related to the starter idea, highlight it and copy it onto the theme list. As you do so, make a note of how it seems related to the starter idea (if you like using memos).
Read the list of repeating ideas until you have selected and copied all the repeating ideas similar to the starter, and therefore to each other. In each case, make a note about the conceptual similarities you identified.
|
(a) “My dream was to look like my father.” (b) “My father never said I love you.” (c) “You have to be there whenever the child needs you.” (d) “We are co-workers in the field of God.” (e) “Adults do not play.” (f) “There is no inch of laziness in my father.” (g) “When they say your father is coming, you run inside.” (h) “My father took care of other children and did not much care for me.” (i) “I love the way my father treated my mother.” (j) “You call your kid and say ‘I love you.’” |
The group of repeating ideas you end up with will define the first theme. The similarities you have recorded in your notes suggest the conceptual basis of the theme.
As an example of how to discover themes, consider the small list of 10 repeating ideas given in Table 6.3. They are the first seven items of Table B.2 in Appendix B, plus three others, given in a scrambled order.
Begin by directing your attention to repeating idea (a) “My dream was to look like my father,” which will be the starter for the first theme. Then, keeping (a) in mind, read down the list until you encounter idea (f) “There is no inch of laziness in my father.” (a) is similar to (f) in that both express the men’s admiration for their own fathers.
Continue reading again until you encounter idea (i) “I love the way my father treated my mother,” which expresses the men’s admiration for their fathers’ behavior toward their mothers, and is therefore similar to the other two. At this point you have come to the end of the list and have discovered the group of repeating ideas that define the first theme. Having selected the repeating ideas for the first theme, highlight and copy them from your list of repeating ideas file into the themes file.
Construct your second theme from the repeating ideas list in the same way. Direct your attention to the first repeating idea that was not included in the first theme. This will be the starter idea for your second theme. Then read through the list, selecting all the repeating ideas similar to the second starter. This group of ideas defines your second theme.
Before you read further, try to construct the second theme, starting with repeating idea (b) “My father never said I love you.”
Here’s what we did. We began with repeating idea (b) in which the men expressed dissatisfaction with their fathers’ absence of affection. We then included repeating idea (e) “Adults do not play,” in which the men expressed dissatisfaction with their fathers’ being strict, stern figures. Next we included (g) “When they say your father is coming, you run inside,” in which the men expressed dissatisfaction about how fearful they were of their fathers. Finally, we included (h) “My father took care of other children and did not much care for me,” in which some of the men expressed dissatisfaction that their fathers had children with many women. This group of four repeating ideas defined our second theme.
Continue going through your list of repeating ideas until you have assigned all of them to a theme. Several questions may be occurring to you as you do this exercise. How do you know when repeating ideas are sufficiently similar to express a common theme? How do you know what that theme is?
There is no formula for answering these questions. Some ideas for themes emerge from your literature review. For example, when we read the research on Haitian families in Caribbean journals, we found that the authors described the Haitian father as a remote disciplinarian, and the Haitian husband as someone who frequently had children with women other than their wives. Thus, when the men in our study complained about these aspects of their fathers (“My father never said ‘I love you’”), the information was familiar to us. Similarly, we are therapists, so when the men also reported admiring their fathers (“My dream was to look like my father”), we were not surprised that they had both positive and negative feelings about them. We used this conceptual framework of emotional ambivalence, of positive and negative feelings about their fathers, to identify abstract patterns among the repeating ideas.
As you go through the procedure, you will grow to understand the research participants and their subjective world. As this happens, you will find that themes will emerge from the data.
After you have grouped together as many repeating ideas as you can, clear up any loose ends in the same way that you did with the relevant text. Deal with orphans by deleting them, incorporating them into a theme you already have, or going back to your raw text to find more repeating ideas that connect with them. If your themes are too broad or too narrow, make changes accordingly. If you have second thoughts about your list of themes, reorganize it until you are satisfied.
How many themes should you have? Although there is no hard and fast rule, we suggest reducing the number of repeating ideas by a factor of 3 or 4, resulting in from 10 to 20 themes, with an average of 15.
3. The next step in the process is to name your themes. While you were grouping the repeating ideas, you had thoughts about the abstract patterns that pulled those ideas together in your mind. Name the themes with an easily understood phrase that expresses this common thread. Keep these ideas simple, and avoid jargon. The research participants should be able to recognize the themes as something they might have said.
For example, as Table B.2 (Appendix B) in the article shows, we named our first theme Praising aspects of the traditional Haitian father. This theme captures how the men admired the way their own fathers fulfilled aspects of the traditional Haitian fathering role.
Similarly, the theme Dissatisfactions with aspects of traditional Haitian fatherhood captures how the men were unhappy with some aspects of the traditional Haitian fathering role. Most prominently they were dissatisfied with their fathers’ emotional distance and sternness, and wanted to treat their own children differently.
As you struggle with naming your themes, you may decide that your repeating ideas need some revision. It is not unusual at this point to go back and change several repeating ideas and relevant text in order to conform to your new understanding of the data. Many students feel very discouraged when this happens. They thought they had completed the repeating ideas phase and moved on to theme creation. Now they feel as if they are back at square one. If this happens to you, do not lose heart. It is actually a positive step, because it means that you are learning about your participants’ subjective experience in a more nuanced way. This is the point of your research!
In our project, we always develop our themes independently and then meet to discuss them as a team. This discussion may result in even more reorganization. Again, we may not ultimately agree on the themes; however, the principle investigator must be able to justify her or his organization to the team.
4. Finally, check your work with a consultant. In our view, this should be a member of the culture you are studying who is not involved in the research project. In the Haitian fathers study, we used a female psychology graduate student whose family was originally from Haiti, and a Caribbean psychologist who was an expert on cross-cultural research on fathers. In our research on gay fathers, we used two psychologists who were established researchers on gay and lesbian families, both of whom were gay.
Using members of the culture is crucial because of the racism, classism, ethnocentrism, and homophobia that are endemic in our culture. You can either ask your consultants to code a small number of transcripts independently and compare your coding to theirs, or you can bring your organization of the data to the consultant, and ask for feedback. As in previous steps, having to explain and justify your work to someone else will improve your thinking. Make whatever changes result from this discussion, and continue until you and your consultant are satisfied.