Drawing implications from Potter & Hepburn

As preeminent voices in the field of discursive psychology and interaction studies, Potter and Hepburn began their careers researching interviews and interactions. It is through this lens that they offer advice and implications to qualitative researchers on the use of open-ended interview data.

from FreeDigitalPhotos.net

The authors contextualize their argument with 20 years of conversation analytic research on the Q&A format in institutional settings and everyday life. On top of that, researchers are now studying the use of Q&A in the social sciences, turning the "analytic searchlight" on themselves (p. 4).  

Potter and Hepburn want to bring rigor to a form of data collection that has been "too easy, too obvious, and too little studied," and, thus, has been the basis of much poor research (p. 3). They say their goal is to critique the ubiquitous open-ended interview so as to improve it. Considering their backgrounds, can they do this without coming across as theoretically and methodologically imperialist?

To their critics, Potter and Hepburn simply say "...[R]esearchers defending the status quo will need to show that the close dependence of both the form and content of the ‘answer' on the design and delivery of the ‘question' is not of general consequence" (p. 17). Who can argue with that?

A theory-to-practice feedback loop

The chapter is organized into two sections: analyzing interview data and reporting findings from interview data. Personally, I would have preferred more discussion about the design and conduct of interviews themselves, as that is what I'm in the process of doing right now. It seems that as we are interviewing so many procedural and methodological missteps can occur.

I guess it is Potter and Hepburn's belief that the researcher has more oversight over the analysis and reporting stages. We have greater opportunity at these stages to practice reflexivity and exercise conscious control over methodological choices than during the unstructured free-flow of qualitiative interviews. Their point is that theory feeds practice. The more carefully we attend to the challenges in our analysis and presentation of findings, the more thoughtful we will be in our future interview set-ups. Maybe we will reconsider our choice of data generation altogether. (True to their roots, Potter and Hepburn recommend more consideration be given to natural interactions.)

Each section of the chapter addresses four challenges. The first section deals with challenges in reporting and representing data; the second section discusses analytical challenges. As I summarize these, I will note specific implications for my current project.

Challenges and opportunities for reporting interview data

1) Make the interview set-up explicit. Describe how the participants were recruited and the explicit language and categories used to do so. Further, what kind of task, instruction, guidance were the interviewees given? I did not conduct a blind recruitment. I simply sent out a blanket invitation to all the study participants asking them to talk to me about their experience. I have the wording of that email which I can include as an appendix along with the interview protocol that I read verbatim at the outset of each interview.

2) Display the active role of the interviewer. Rather than inserting illustrative quotes from participants only, carefully include extracts that include the entire interactional turn between interviewer and interviewee. This enables readers to see "how the form of the answer may be occasioned by the form of the question" (p. 10). Fair enough. This seems more doable in a dissertation study than in a journal article with space constraints. Potter and Hepburn are building a case for a total departure from how participants' voices are depicted in qualitative studies. The introduction of the playscript format, itself not widely used in research articles, is still inadequate, according to the authors, because it does not capture "delivery of the talk" (p. 10). Are they really suggesting that all qualitative researchers use Jeffersonian transcription?

3) Represent talk in a way that captures action. And there it is: Jeffersonian transcription "allows the identification of a number of potentially consequential interviewer actions" (p. 13). Potter and Hepburn suggest that we give our extracts (not the full transcripts) the Jeffersonian treatment. This is also completely doable in a dissertation study and would likely be appreciated by committee members, so long as an explicit explanation is provided and a key to Jeffersonian notation appears in the appendices. I am currently revising portions of Chapter 3, and I wonder if this isn't something I should consider doing?

from FreeDigitalPhotos.net

4) Tie analytic observations to specific features of interviews. This suggestion is based on the claim that interview data is under-analyzed. Instead of combining interview data with large chunks of analysis in one paragraph, Potter and Hepburn recommend bolstering the analysis with simple formatting changes, such as representing turns on separate lines and numbering the lines.  Again, not hard to do when one is not constrained by space limitations or word counts, as in a dissertation study. But I wonder how these representational practices would be received by journal editors and article reviewers?

Challenges and opportunities during analysis

5) Flooding. This is the phenomenon of the interviewer's agenda creeping into the interview in benign ways, such as follow-up questions and probs formulated on the fly as well as "acknowledgement tokens" and agreement statements which potentially lead interviewees down a desired path. While these interactional events are natural and cannot necessarily be curtailed, they need to be acknowledged or risk analysis "chasing its own tail" (p. 20). I am not sure what the implications of this might be other than to heighten awareness and reflexivity and to perhaps include interpretive commentary to that effect when including extracts with multiple turns.

6) Footing. Footing refers to the various positions or categories that participants may speak from during an interview. In Extract 3 the interviewer poses a question that the interviewee may answer from a personal or institutional standpoint or some combination thereof. Likewise, the interviewer may be positioned as someone who is personally interested in the question or one who is just asking the question so it can go "on the record," like a disinterested reporter. This is fascinating. At this point in the discussion, it is safe to say that the extract Potter and Hepburn have used extensively throughout this article, the one illustrating a "put together on the spot" question about teachers (p. 19), is really just pretty terrible. It is like the proverbial "How did you feel when _____" questions that TV news reporters ask disaster survivors.  Does the question really bear asking? Potter and Hepburn don't belabor that point; they simply call attention to the lack of interviewer engagement, no "ohs" or "hmm-hmms." To an extent, these conversational elements are "subtle, complex and potentially consequential" (p. 24). When they are missing altogether, it suggests that the interviewer perhaps asked a question for which he or she already knew the answer. To be considered a credible social scientist, I had better ask thoughtful questions I genuinely care about, lest I be compared to a "news anchor."

7) Stake and interest. Unlike corporate-sponsored focus group surveys, qualitative research is typically conducted by someone who feels passionately about the topic, and interviewees are recruited because they ostensibly have an interest in the topic too. It is a wise analytic move to pay attention to how interviewer and interviewee manage their stake in the discussion. It seems that stake and interest are the the next logical progression from footing. If I openly acknowledge my interest in the topic, listen attentively, and engage with my participants, I must be reflexive about the consequences of these actions.

8) Cognitivism and individualism. Potter and Hepburn call qualitative researchers to task -- particularly constructivists -- for the ironic "privileging of conceptual mediation over action and the treatment of cognitive language as referential" (p. 28). Qualitative researchers put their interviewees in an impossible "epistemic position" to speak their minds and the minds of others, as if that kind of knowledge can be objectively reported and stated without performing a "range of ambitious cognitive judgments and feats of memory and analysis" (p. 29). Cognitive language (reminding me of the "residue" of positivism from Piantanida and Garman's book) creeps into our interactions, carrying with it the assumption that somehow people have "immediate and privileged access to their own opinions and attitudes" (p. 30). A discourse or conversation analysis approach would obviously be Potter's and Hepburn's desired route to mediate this challenge, but that isn't the only way. I am interested in the analytical tools of Cultural-Historical Activity Theory, which enable the researcher to explore "agentive dimensions" (emotions, identity, motivation) (Roth, 2009, p. 53). In one such study, Roth and colleagues (2004) drew from case study data, including transcripts, to theorize identity formation as predicated upon actions and outcomes.The authors contended that identity is not stable and is the outcome of participation in social activity. The key here is that interview data was not privileged over other forms (videotapes, journal reflections, field notes, and emails), a methodological move that Potter and Hepburn would endorse.

References

Potter, J., & Hepburn, A. (in press). Eight challenges for interview researchers. In J. F. Gubrium, & J. A. Holstein (Eds.), Handbook of interview research (2nd ed.). London: Sage Publications.

Roth, W.-M. (2009). On the inclusion of emotions, identity, and ethico-moral dimensions of actions. In A. Sannino, H. Daniels, & K. D. Gutierrez (Eds.), Learning and expanding with activity theory (pp. 53–71). Cambridge: Cambridge University Press.

Roth, W.-M. (2004). Activity theory and education: An introduction. Mind, Culture, and Activity, 11(1), 1–8. doi:10.1207/s15327884mca1101_1

Drawing implications from Potter & Hepburn

Reading notes for Oct. 20

Reflecting on CDA

It's official.  It there was any doubt I was CDA before, it's now certain: I'm one of them.

 

I approached this week's selections from Rebecca Rogers' An Introduction to Critical Discourse Analysis in Education  thinking, "Gee is my man." But the more I read about Gunther Kress, the more intrigued I became. 

 

What's odd is, Fairclough, Gee, and Kress have been with me all along without me even knowing it! On page 9, Rogers mentions the New London Group and their seminal document A Pedagogy of Multiliteracies: Designing Social Futures. This is a cornerstone piece of my theoretical frameworks, and I am only now making the connection that all three fellows helped author it!  Weird.

The "New London Group" was mostly, if not entirely, a collective of literacy scholars who called for a radical change in literacy pedagogy in response to “cultural differences and rapidly shifting communications media.” Anyone who writes about or researches the "disconnect" between students’ in-school and out-of-school literacy practices tends to cite the New London Group's theoretical overview of how to successfully leverage new literacies in service of educational goals as well as the personal development and overall well-being of students. 

 

The New London Group took pprogressive-minded educators to task for over-emphasizing "situatedness" in instruction, and they put forth the “pedagogy of multiliteracies” to compensate for the limitations of situated practice by adding in overt instruction, critical framing, and transformed practice. Through critical framing, learners “denaturalize” their growing mastery of practice and “constructively critique it, account for its cultural location, creatively extend and apply it, and eventually innovate on their own, within old communities and in new ones” (pp. 86-87).

 

I have referenced the New London Group "manifesto" as a how-to document for student-centered technology instruction; I had never understood it as a "shared vision" for critical discourse traditions, as Rogers notes (p.9). But it makes sense in light of her explication of the generative aspects of Critical Social Theory, where "the end goal is to hope, to dream, and to create alternative realities that are based in equity, love, peace, and solidarity" (p. 5). This is getting at the "design" side, and I think it explains why I am drawn to Kress' vision. 

 

Overall, it was challenging to tease out the distinctions between Fairclough, Gee, and Kress.  It all started to run together at some point, and they all build on the principle of discourse as socially constitutive and tied up in power systems. Discourse is multimodal, with language/grammar being but one of several representational sign systems. And I took Rogers' caution to heart, that it was not necessary to "choose sides" and that a "hybrid" approach was not only possible but even preferred in light of the various constraints imposed, especially within institutional contexts.

 

However, I did feel a tug toward Kress, on whom I will comment a bit further. 

 

Connecting to Kress

As were some of my classmates, I was intimidated by the discussion around semiotics and wondered if I have the intellectual capacity to take on these ideas. But I liked what little I understood about social semiotics as a sort of buffer against pure critique of social and institutional structures.  Social semiotics looks at how meanings are designed and re-designed through interaction, with implications for those (like me!) who are interested in "the relationship between educational practices and productive uses of power" (Rogers, p. 14).

 

I want to learn more about this "difference between design [social semiotics] and critique" (Rogers, p. 14), and I think this is what separates Kress from the others.  It has something to do with the "material" aspects of discourse and what Rogers notes as Kress' "focus on the sign-maker, rather than the sign" (p. 8).  

 

In Kress' chapter, I just really loved his discussion of the student's rendering of cell with nucleus as an "epistemological commitment."  It sounds high-falootin', but what it suggests to me is this total re-orientation in the classroom to what students know and can do as opposed to the typical deficit model that measures students' learning by how well they enact preconceived "standards" of excellent and how well they perform on achievement tests.

 

You see this a lot in literacy instruction, such as when teachers or parents fail to recognize emergent print concepts and letter formation in a pre-literate child's playful "scribbling," or when adolescents' bonafide text creations (fanfiction, YouTube uploads, personal webpages) are dismissed as "spending too much time playing on the computer."

 

I love Kress' emphasis on multimodality as defined on pp. 208-209. , which suggests a major paradigm shift for classroom teaching. Kress says, "To make a sign is to make knowledge" (p. 211). So powerful!! We say we are always looking for "signs of learning" in the classroom, but are we really? Seldom to we create spaces or make room for "spoken signs" and "image signs."

Reading notes for Oct. 20

"I don't know": Reading notes for Oct. 13

Maybe it's a blessing I did not post to my blog last week, because it wasn't until Thursday night's discussion that I began to consolidate my thinking and make some connections across the last chapters of Wood & Kroger and ten Have. I recognize that "consolidating" and "connecting" is the intended purpose of the blog, but even then, I usually seek out someone to talk to as I am immersed in the readings, usually my classmate Renee.

But I didn't have time to chat with my colleague last week, due to a major transcription task in front of me.  Once I removed the earbuds and got the voices of my field informants -- all 11 of them -- out of my head, I drove to class in a mental fog.  I was in no state of mind to discuss, much less write about, the "detailed, painstaking, sometimes tedious work" of producing a warranted and trustworthy research analysis (Wood & Kroger, p. 181).

"I don't know" 
Who would have thought that the innocuous and ubiquitous phrase "I don't know" would snap me out of my transcription-induced funk?

Last Thursday night, some of my classmates expressed skepticism around claims made about the use of "I don't know" in academic discourse. At first, I was reminded of CA's less-than-endearing traits (to me) --  its "obsession" with details and unapologetic refusal to acknowledge the obvious (ten Have, Chapter 3). But the extract in question was from a DA study, not a CA study, and the reaction it stimulated brought our class discussion back full-circle to basic issues and challenges of producing "good" research.  Although it was just a short segment of the full analysis, our diverse reactions to the study provided at least a partial indicator of its power to stimulate meaning-making. 

Audience 
One of the most profound lessons for me in year-long ethnography with Dr. Anders was the idea that we are not speaking for participants, but to an audience. The word used in class last week was "resonance," a criterion for good writing in general, I believe. With 12 hours of qualitative methodology courses under my belt, I can safely say no single scholar has stirred my thinking more on this topic than Laurel Richardson, whose essay "Writing: A Method of Inquiry," I'd already read twice, and now, a third time.

According to Richardson, our "meaning is in the reading." As a qualitative researcher, I must write in such a way as to ensure a connection with an audience, my intended readership, whoever that may be. "...[Q]ualitative work depends upon people's reading it," Richardson tells us (p. 346). 

Applied CA 
Without realizing it, I was primed for reading ten Have's conclusion. Finally!  An acknowledgement (albeit a "cautious" and "agnostic" one) of those researcher-practioners who aim to take up CA to inform or contribute to a specific field or discipline. In fact, I learned from these last chapters that CA's warrant has much to do with the analyst's orientation to audience.

In a summary of ideas proposed by James Heap, ten Have writes, "Whether it is worthwhile to do ethnomethodology (or CA) depends on the value of the news it produces for an audience" (p. 195). Further, the goal of applied CA is to deliver "news" that may foment change within an organization or among a group of participants.

So, CA isn't just the linguistic equivalent of navel-gazing after all! :-)

This idea of "newsworthiness" as a measure of warrantability reminded me of Wood & Kroger's "powerful criterion" of fruitfulness. Fruitfulness has to do with "implications of the present work for other work" (p. 175). Like last Thursday's class discussion, I feel I have performed a 360-degree turn. I am back now to some of my earliest questions, reflections, and reservations. The idea of DA as a mechanism for shining light on "acknowledged social problems" is one I've wondered about off-and-on within this blog. 

"Alternatives"

The price, of course, for openly assuming a stance is increased vigilance in the area of researcher reflexivity. The qualitative researcher is obligated to "bracket," to be openly reflexive about hunches and presuppositions.

Wood & Kroger describe this as taking account of "moral implications" (p. 175).  I like how another author, Patti Lather in "Issues of Validity in Openly Ideological Research," calls it "a self-corrective element" (p. 188). Similar to Wood & Kroger's call for criteria that "transcend" traditional positivist standards (p. 167), Lather offers a "reconceptualization" of validity (p. 188). She writes, "Our best shot at present is to construct research designs that push us toward becoming vigorously self-aware" (p. 190).

I enjoyed how Wood & Kroger recast "reliability" as a matter of repetition, but not in outcomes, which are always subject to interpretation. Instead, repetition as a standard of quality works differently in DA. It's the aggregate of researcher actions, "part of the careful attention to detail and the concern for refinement that are major features of discourse-analytic work" (p. 166).

And validity is not based on black-and-white truths that are somehow captured. Validity is a matter of strength and trustworthiness that is "co-constructed" between informants, author, and audience. Quoting Potter, Wood & Kroger point to the "greater prominence" assigned to reader reaction, "an emphasis that both results from and is encouraged by the greater transparency of discourse-analytic work" (p. 168). 

Crystals

Crystals are typically transparent, sometimes a little cloudy and not without a few imperfections. And that is why I love Richardson's crystal metaphor, recommended as an alternative to the traditional triangulation procedure for demonstrating quality of a study.  I went back to "Writing: A Method of Inquiry" to make sure I understood the metaphor. Could I ferret out any implications for DA? What about those of us who are not writing "postmodernist mixed-genre texts"? Does that mean the process of "crystallization, " as Richardson describes it, isn't a viable alternative within the DA tradition?

I don't know.

Or, maybe crystallization is as much a researcher process as it is product?  We refract and reflect at all stages of research so the final piece of writing, even a conventional research report, is clearer and more engaging.  Richardson describes experimental processes, such as revisioning one's work as fiction so as to see it from different points of view.  This doesn't mean we all hang it up and become fiction writers, she says, it means we take advantage of the "propitious" time we live in and give ourselves up to the "playful pull" (p. 362).

To me, that suggests experimentation in one's writing as a means to locate, condition, and fine-tune one's voice, even if that voice ultimately finds its audience via traditional academic formats. In that regard, then isn't that what this blog is for? A possible venue for "staging a text" and "writing in other ways"?

I don't know. And that's OK.

As Richardson says, "Paradoxically, we know more and doubt what we know."

"I don't know": Reading notes for Oct. 13

Goin' on a pattern hunt!

As part of his Kindergarten homework, Henry had to go on a "pattern hunt." We started the hunt immediately upon pulling up to the curb in front of our house after school. With my phone, I took pictures of patterns he identified in the yard and on the porch.

Henry located numerous patterns of shape, size, and color within the American flag, the picket fence, the decorative lights strung on the porch, the poured-concrete pavers on the ground. Patterns, patterns everywhere!

"Look, Henry!" said his grandmother, who was with us. "Here's one," she said, pointing to a series of concentric circles atop a water meter cover.

Henry frowned, not seeing it. "Granny, that is not a pattern.  A pattern is the same and repeats itself," he instructed, before continuing his search.

Well, of course, it was a sort of pattern we were all looking at in the ground. Patterns involve form and/or structure and occur with varying levels of extensiveness and complexity, thus making them "sometimes difficult to recognize" (Wood & Kroger, p. 117).  Henry was so well schooled in the textbook definition of "pattern," he could not wrap his five-year-old mind around an interesting departure from the rule.

This is the challenge of the novice researcher/analyst: striking the right balance of unschooled intuition and basic analytical knowledge. It isn't easy intuiting significance from interestingly deviant cases if you don't "know" a little about what you are looking at.

Recognizing "patterns of interactions"

At once, I think novice analysts may have an advantage insofar as "unmotivated looking" because we are not deeply schooled in the DA/CA literature base. And, yet, as ten Have explains his strategy, "...[T]he starting point is some 'noticing' in the transcript that something 'interesting' seems to be happening at some moment. From that moment on, the purpose of the strategy is to elaborate and contextualize that rather intuitive moment" (p. 124).

It is helpful, then, to possess "a few basic concepts from the CA tradition to structure one's 'looking'" (p. 121), and ten Have's four, fundamental "organizations" from the CA knowledgebase seem incredibly helpful to me as I move forward.

In a similar vein, Wood & Kroger provide lists of "sensitizing devices," lest we become "loose and undisciplined" during our early explorations of the data (p. 91). They acknowledge that it is difficult to know if an interesting feature hasn't already been described elsewhere by other scholars. This has come up earlier in class discussions: how do we keep up with the vast array of features, concepts, and devices? In this regard, it seems the novice is at a disadvantage if not at least generally familiar with the CA/DA literature. No one wants to reinvent the wheel.

In at least one previous blog post, I have worried about not being able to master the vast array of "concepts and devices" emanating from the traditions of DA. For this reason, I enjoyed their discussion of "useful concepts or sets of ideas" that, in some cases, "transcend any particular tradition" (p. 100). Like ten Have's suggested "organizations," which I think Wood & Kroger lump together as "collected dossiers," the list of concepts in Chapter 8 strike me as something to hang onto as I prepare to embark on my own pattern hunt.

Another go-round with context

Context remains at the center of "lively debate" in DA, this time regarding its role in the analysis and interpretation of data. Where, if at all, does information about settings, circumstances, social roles, demographic variables, and so on work its way in?

Context comes from within and without. Wood and Kroger note the "contextualization of utterances is a procedure that is relied upon by participants themselves and is hearable-visable in the discourse that we are analyzing" (p. 128).

It gets dicey around "extrinsic context" -- how much of setting, institutional orders and participant characteristics, class, gender, race, age, and so on, can we draw on in our analysis? Wood and Kroger recommend a "simple" strategy, which we have heard before in class -- it's only relevant if the participants make it relevant. Otherwise, how do you know where to stop? And heavy doses of researcher reflexivity can help mediate the process of determining what is relevant. Wood and Kroger: "It is crucial to emphasize that the recommendation is not to ignore context, to leave out what is important, but to be very careful about how it is brought into analysis" (p. 129).

This is where I have some questions about the document/text we are to locate for our project. I have looked at one primary document, the School Improvement Plan (SIP), a document every public school must compile and submit to the state each year. I located in the SIP a very small bit about the collaborative teaming that I am currently observing. I have reviewed every relevant page of the school website, but this is understandably geared toward students, parents, and community stakeholders and does not include information about the mission, philosophy, purpose, or background of the collaborative team concept or even site-based management in general. I have also sent out emails to a few team members inquiring about documentation. 

But perhaps I should wait and see what kind of paper trail is circulated or produced from within the team when I am there??

Another interesting analytic move related to context that I would like to learn more about is this idea of "extending the boundaries of the interaction under analysis" and developing "ethnographic knowledge" or "shared knowledge" between myself and the participants.  Wood and Kroger say, "Sometimes ethnographic research is necessary." This suggests a longer engagement in the field than I had originally planned, and I am starting the process of requesting an extension from the principal and the head of research for the school district.

ten Have, P. (2007). Doing conversation analysis: A practical guide (2nd ed.). Los Angeles: SAGE. 

Wood, L. A., & Kroger, R. O. (2000). Doing discourse analysis: Methods for studying action in talk and text. Thousand Oaks, CA: SAGE.

Goin' on a pattern hunt!

DA project update

I have observed twice at my research site "Plan B" (the high school) since Sept. 22. This week brings my third visit to the school and -- finally -- an opportunity to record a full collaborative team meeting between core academic teachers. 

After consulting with my adviser, I have decided to request permission from the building principal to continue fieldwork indefinitely with hopes of turning this project into "something more." I am currently doing a research review for Dr. Allington on the ways in which collaborative dialog, distributed leadership, and shared decision-making support effective literacy instruction. We will see where this goes....

Other than an overarching concern about producing a meaningful transcript in time for the Oct. 13 data session, I would like to share some initial concerns and observations about the site and participants:

1) After listening in on two meetings, I am conscious of the fact that I am really attuned to the content of the talk between teachers and have not really been cognizant or deliberate about talk features and devices. I am growing a little antsy, I guess, because this weeks' chapters are all about pre-analysis and early analysis of discourse, and I am still at the front-end of transcription.  That said, the content of the collaborative team meetings is really interesting (to me) and covers a variety of topics and issues.  I am much more jazzed now about this research plan, than my original Plan A. Funny how things change.

 

2) During the first observation, I sat at the table and chose not to write notes or clack away on my computer because I did not want to put off the group.  Plus, in a way, that constitutes "recording," and I agreed not to record.  So, afterward, I went to the school library and typed up everything that I could remember, first impressions, questions, ideas, etc. On the second observation, I sat away from the table and tapped fieldnotes into my iPad and this proved to be effective, but I was less able to attend to all of the non-verbals, facial expressions, etc.

3) The perfect metaphor for this experience: embedded journalist with a military unit on the front lines of action.  That's how it feels. Some additional observations:

• lots of laughter, humor, camaraderie, "inside" jokes
• lots of micro-level discussion and problem-solving about specific students and student issues (absences, truancy, pregnancy, failing grades, tardies, school transfers, pressures related to family and/or work and financial obligations) So, the somewhat cliched image of educator as "triage expert" comes to mind, but that's my metaphor, not the participants'.
• The bulk of conversation focuses on grades, test scores, passing-failing -- issues of performance as opposed to issues related to actual teaching and learning. And I have heard zero dialog about literacy learning. Perhaps this is the natural outcome of interdisciplinary teaming at a high-needs school, where the problems of students are "shared" between faculty who are making deliberate effort to create a culture of caring. 

4) Some potential problems and challenges pertaining to transcription come to mind:

• lots of overlapping and latching speech
• lots of use of student names.  I can supply pseudonyms for the major participants on the team, but what do I do about the countless student names brought up in conversation?  Last week they probably discussed two dozen different students. Do I insert students' initials into the transcript? Do I use S1, S2, S3, etc.?
• environmental noise (from a photocopier in the meeting room as well as an open door on a hallway filled with students during the passing period) could potentially corrupt some segments of the recordings
• Zan, my pseudonymous friend and primary contact on the team, occasionally addresses me directly to fill in context or to include me in whatever joke is at hand. (I, in fact, know about half of the collaborative team already, so building rapport has proven not to be an issue).  Do I tell her just don't do that anymore?

DA project update

Example of DA study in reading education

I am currently doing a research review of the state-of-the-art in secondary reading instruction.  I am finding that this is variously referred to as "high literacy," "new literacy," "thoughtful literacy," and "higher-order literacy" (Allington & Cunningham, 2007; Langer, 2002). Discussion, or "intelligent literary discourse" (Langer, p. 1), is central to the development of this idea. So, to that end, I had in mind that I would search for a discourse analysis study that focused on some aspect of secondary literacy instruction, but particularly classroom discussion.

I began by searching the recommended discourse analysis journals listed in BlackBoard. I found that my combination of search terms -- literacy (also tried English), high school (also tried secondary), instruction, teachers, and adolescent -- did not yield many results relevant to my research agenda.  Within Google Scholar, I added the search term discourse analysis, and I had more luck finding what appeared to be interesting studies (to me) but nothing that resembled "pure DA." I changed up my approach yet again, going to the combined archives of two top-tier, peer-reviewed journals in my field: Journal of Adolescent and Adult Literacy and Reading Research Quarterly, both published by the International Reading Association. I discovered that if I did not narrow the "Learner" field (as seen in the screen capture below), I got 18 search returns:

However, even among these 18 results, nothing quite resembled the DA studies we read last week. Several of the studies I previewed were of a case study design that deployed discourse analytic techniques.  Other studies claimed an ethnographic orientation while applying discourse analysis methods. Is this the bricolage effect (Wood & Kroger, pp. 25-26)?  In the past, I have read philosophical and scholarly commentary that compares the educator's work to that of  a bricoleur, and I am now wondering if that isn't being reflected in the work of educational researchers within the qualitative tradition, or at least within the DA tradition.

At any rate, one study in particular caught my eye because of its focus on literature-based instruction and classroom talk with pre-adolescents.  I chose this study to critique.

Clarke, L. W. (2007). Discussing Shiloh: A Conversation beyond the book. Journal of Adolescent & Adult Literacy, 51, 112-122. doi:10.1598/JAAL.51.2.3

• Context and motivation-In fall 2003 five fifth-grade students appeared to be vigorously discussing literature in student-led group, but Clarke noticed segments of conversation within the transcripts that went "beyond the surface level of the discussion" (p. 113).  Both Clarke and the students' teacher believed in the value of literature discussion groups for improving levels of student engagement and higher-order thinking and problem solving, but Clarke wanted to open up this "common instructional practice to a critical analysis" in hopes of informing classroom teachers how student-led literature groups might be restructured to better serve the literacy skills and future employment opportunities of "working class students" (p. 113). [The literary text being used is the Newbery-winner Shiloh (1991) by  Phyllis Reynolds Naylor.]

• DA tradition-Claiming "influences beyond this text" (p. 113), Clarke used Critical Discourse Analysis (CDA) to connect the students' discussion of Shiloh to classroom, school, and community discourses. Specifically, Clarke drew upon the three-part analytic framework devised by Fairclough: description, interpretation, and explanation.

• Methods/data sources-As part of a larger, three-year qualitative study, Clarke video- and audio-taped many discussions and supplement these tapes with observation, field notes, student interviews, focus groups, and student writing samples and journal entries.  For the study at hand, he chose to zero in on a single conversation between five fifth graders, which he said represented themes found across other groups. He coded this conversation for turn-taking sequences as well as other features (listed below) and then performed a content analysis based on length of turns and gender.  He wrote "data narratives" (analytic memos?) around each code (pp. 114-115). During the interpretation phase, Clarke widened his analytic net to include all the other data sources from his three-year study, which he said "gives power to this small example as an illustrative case"  (p. 115). During the final part of Fairclough's framework, which is explanation, Clarke accounted for his interpretations by connecting them to broader "situational, institutional, and societal influences" and the existing literature base on dominant discourse by boys (pp. 117-118).

• Language features attended to (theoretical constructs)-Following Fairclough's guiding question -- "What interactional conventions are used and are there ways in which one participant controls the turns of others?" -- Clarke decided to  focused on turn-taking systems. He also "coded" for sidetracking, center staging, giving orders, using insults, disagreeing, ignoring, using reinforcement, community building, giving listening responses, and interrupting. In the final phase of analysis, Clarke connected local themes of male-dominated discourse between students to an "underlying ideology of male discursive power" and "hegemony" (p. 117).

• Claims made as findings-Clarke identified three major themes in his analysis of the conversation extract; all three themes essentially had to do with a gender imbalance and the male-dominated nature of the conversation. This led him to conclude that the literature circles as they currently functioned in the fifth-grade classroom were a staging ground for "perpetuating gendered discursive roles" (p. 120).  

• So what question?-  Allowing for the fact that student-led literature circles in theory can make a powerful impact on literacy development and the socialization of students, Clarke suggests teachers rework their approach by more careful and considerate literature selections (texts that explore gender issues, for example), direct instruction on group processes, and "re-inserting" themselves into groups as a facilitator/coach.

• Comments/reflections -I really wanted to learn more about the use of CDA in reading and literacy education, but I don't think this particular study is an exemplary model.  For one, I don't buy Clarke's  generalizations about "working-class students" and his shaky substantive claims about student-led discussion groups as a "common instructional practice." (This is simply not supported in the wider literature base.) Moreover, he makes questionable methodological moves, such as using "student informants" to contextualize video segments with him in a sort of member-checking process and relying on a massive data set to guide his interpretations. His efforts resemble triangulation; although, he does not refer to it as such. Overall, I was reminded of Wood and Kroger's discussion of striking a balance when drawing on different perspectives, lest you fall victim to the "pitfalls of eclecticism" (p. 25).

Example of DA study in reading education

On serendipity and design: Notes for Sept. 22

As in most research, serendipity is important: it helps to be alert for unexpected opportunities. -- Wood & Kroger, Doing Discourse Analysis

A serendipitous data collection opportunity

This week I will begin observing collaborative faculty meetings held each Thursday morning at a local high school where I use to teach. This is one of two data collection options I am exploring this semester for Discourse Analysis. Unlike my Plan A context, in which I am essentially a researcher-participant, I had to negotiate entry into the high school site, and I am still not entirely certain if and when I will be able to obtain discourse recordings there.

As I have described in previous reflections on research contexts, this opportunity to do fieldwork came about through a chance encounter with a good friend and former colleague who happens to chair the collaborative team. As such, I am just trying "to roll" with this one! I am going to try to assume the stance of an explorer, who might use "whatever data" I can lay my hands on (ten Have, p. 71).

But it's hard not to anticipate what I might find!

Once a perennial low-performer with roughly 80 percent low-SES student population, the school in question began a massive reorganization four years ago upon threat of state takeover. It is now divided into vocational learning communities, and I will be observing collaborative talk between teachers and support staff who work in what is referred to as "the health sciences learning community." Last year, for the first time since the inception of the No Child Left Behind law, the school posted annual yearly progress in all areas. So, to some extent at least, the reforms are working, and by all accounts the school is a more energizing and exciting place to work and to learn than it ever was in the six years I was there.

It is impossible to overstate the level of teacher dis-empowerment that existed when I left the classroom in 2005. Opportunities to work or even to talk collaboratively were few and far between. You might say "professional talk" occurred, but it was more like venting, griping, and "sounding off," usually over rushed 25-minute lunches or while standing in line to make photocopies.  Time was not carved out during actual contract hours for teachers to work face-to-face with their peers. 

This semester in Dr. Allington's REED 605 (Organization and Administration of School Reading Programs), we are studying how schools can transform into places of "thoughtful literacy" through a number of reform measures, including professional dialog and distributed leadership models. This leads me to wonder about the impact of the collaborative team meetings on the quality of literacy instruction at my former school.

Has the emphasis indeed shifted from "minimum literacy competencies" to "higher-order literacy" (Allington & Cunningham, 2007, p. 129)?  Will this be reflected in the collaborative team dialog? Is there a connection between collaborative, problem-solving dialog among teachers and collaborative, problem-solving dialog among the students they teach?


And there I go, formulating ridiculously ambitious questions that cannot possibly be answered within the short scope of this data collection exercise. Worse, I fear I am so blinded by my own narrow interests and expectations that if I were lucky enough to stumble upon some analytically worthy chunks of data, I won't recognize them.  After two years of intense, graduate-level study, far removed from the participants and contexts I claim to "know" so well, will my vision re-adjust to the outside world? Or, am I permanently handicapped by the researcher's lenses? Some form of theoretical myopia?

For these reasons, I took to heart the summary of ethical issues and the final cautions about data sources in Wood & Kroger's Chapter 4. In a case such as mine, where I have "happened upon" discourse in serendipitous fashion, it is best to "open up" the questions and maybe let go of them entirely. I also need to be content with understanding the discourse where it is situated, between professionals around a conference table. As much as I want to explore connections to classroom instruction and teacher-student interactions, this is not the route for doing so.

On a slightly different note, insofar as the actual recording procedure goes, events in the last few days have especially sensitized me to the "ethical and practical concerns that require the modification of ideal practice" (p. 64). I have been advised by the principal/gatekeeper to not record during my first few observations, so as not to interrupt the participants' collaborative flow. I spoke to the chair of the collaborative team about this stipulation, and she remarked that she couldn't think of a single member who would object to being recorded. They are, she said, a fairly "unbridled and brazen bunch." As a former teacher, I can relate to this. Because so much of our work day is routinized, scripted, regimented, and timed, we embrace opportunities for collegial interaction -- perhaps too much! Teachers tend to let loose when in each other's company. And that, of course, may be the very reason why the principal wants me to go slow and make my presence known.

Designing a transcript

Serendipity is great for data collection, but not so much for transcription, I think. A novice who attempts Gail Jefferson's approach of noting every feature "because it's there" will surely go mad before completing a transcript! Somewhere between collection and transcription, a researcher must begin to focus her goals, aims, and objectives.

This week's readings on the topic of transcription gave me a chance to revisit some of the concerns I raised in my earliest DA post: namely, just how is it you prepare a transcript with such attention to detail and NOT have a clearly defined research question?

According to Elinor Ochs, you don't. She says, "...the transcript should reflect the particular interests -- the hypotheses to be examined -- of the researcher" (p. 44).

I can see now that it is less a matter of defining questions as it is continually refining questions, such as the "spiralling fashion" noted by ten Have (p. 69). To that end, Wood and Kroger's advice about "initial readings" was especially helpful for identifying an analytic approach as well as segments of data to analyze; although, it implies an even greater time investment for the transcription process than I had initially calculated. I also will the "strategy of reversal," in the event I end up with an uninspiring data set, which is entirely possible due to my narrowing window for data collection.

Of all the readings, I found ten Have's discussion of the "practical aspects" of transcription to be the most accessible, particularly the sections on timing of silences and formatting issues. His process of going in "rounds," starting with a verbatim rendering that grows increasingly descriptive with the systematic layering of "how's," made a lot of sense, especially in light of the practice sessions conducted in class.

I found the Ochs chapter to be the most challenging. It was the way that she described selectivity as a desirable outcome of transcription and not just an inherent problem of transcription that really made me think! After reading Jefferson, I was struck by Ochs' statement that a transcript should not have "too much information.... A more useful transcript is a more selective one." Over and over again in the literature, we are warned about the impact of researcher bias in the production of transcripts. Yet, Ochs seems to recommend we embrace selectivity -- but with our eyes open. Ochs' less-is-more advice sets up an interesting contrast to Jefferson's "if it's interesting put it in" approach.

Problems of selectivity vary from discipline to discipline, according to the researcher's phenomenon of interest. For example, like the other theorists, Ochs discusses strategies for the representation of verbals and nonverbals, but not in the manner I had come to expect after reading ten Have and Wood and Kroger. Unlike the other theorists who warn against privileging nonverbals, Ochs warns against "verbal foregrounding," which, understandably, poses a problem in her field of study, child language acquisition.

In sum, selectivity is inescapable; it is only problematic in the absence of researcher reflexivity. Regardless of one's field, a lack of reflexivity during the transcription process has "consequences," a point Ochs makes repeatedly. It reveals our biases about spatial organization, contingent relationships between utterances, directionality, verbal and nonverbal behavior, and adult communicative models.

References: 

Allington, R. L., & Cunningham, P. M. (2007). Schools that work: Where all children read and write (3rd ed.). Boston: Pearson. 

Jefferson, G. (2004). Glossary of transcript symbols with an introduction. In G. H. Lerner (Ed.), Conversation Analysis: Studies from the first generation (pp. 13-31). Amsterdam: John Benjamins Publishing Company.

Ochs, E. (1979). Transcription as theory. In E. Ochs & B. Schieffelin (Eds.), Development pragmatics. New York: Academic Press.

ten Have, P. (2007). Doing conversation analysis: A practical guide (2nd ed.). Los Angeles: SAGE. 

Wood, L. A., & Kroger, R. O. (2000). Doing discourse analysis: Methods for studying action in talk and text. Thousand Oaks, CA: SAGE.

On serendipity and design: Notes for Sept. 22