A productive beginning

In October, when I first got the request for a public abstract, IRB approval, and my response to questions, I was thrilled, and went into a frenzy of activity, digging further into discourse and linguistic analysis and survey design. The questions were exactly the questions I would have run into at some point, so to get them so early was immensely helpful. I therefore felt mentored and scaffolded by my program officer, Elliott Douglas (@ElliotPDouglas) and the anonymous reviewers on my panel. 

I reached out to some of my advisory board members, who responded quickly, despite the short timeline. Special thanks to Jennifer Bekki, PhD for her detailed feedback and great suggestions about a resource for survey design:

McCoach, D. B., Gable, R. K., & Madura, J. P. (2013). Instrument development in the affective domain. Social and Corporate Applications: Springer.

I began my first step, which was to attempt to adapt the agency toolkit for my analysis. 

Konopasky, A. W., & Sheridan, K. M. (2016). Towards a diagnostic toolkit for the language of agency. Mind, Culture, and Activity, 1-16.

Abigail presented similar work in a session at AERA 2017, and it resonated with me. Adapting this was productive and fun, if slightly speculative work at the outset. It was a bit of a leap of faith, as this was a new-to-me focus on qual analysis. 

Code	Description	Value  1	Value  0	Value  -1
SelfAs	Speaker uses "I" or "we" plus a verb and an object or condition, placing the self or team as the actor. In contrast, when the speaker is the object, s/he gives agency to others, such as by fulfilling their request. Leave blank for instances where generic "you" is used or no subject	Speaker uses "I" placing self as actor	Speaker uses we, placing team as actor	Speaker places self as object, gives agency to other
Mit_Gen	Mitigation via generic form. Speaker uses a generic "you" or places self among many others.	Not present	Present	NA
Mit_Mod	Mitigation via modality. Speaker uses verbs that indicate a lack of control over the situation, such as "had to' in place of "did"	Not present	Present	NA
Mit_Adv	Mitigation via adverbial form. Speaker modifies action with "just" "so" or "because" or similar words, indicating that external circumstances affected their actions	Not present	Present	NA
Mit_Hed	Mitigation via hedging. Speaker modifies statement with diminishing terms, or modifies with suppose, guess, or similar.	Not present	Present	NA

 

Part of my phase 1 work is to reanalyze existing data, including some of my dissertation data! When I did my dissertation, very few social scientists took the 3-paper dissertation route. As a result, I ended up with ~150 pages of interpreted transcript. While this looks impressive, it was essentially unpublishable work. And very little of that qualitative work has ever made it out of that dissertation. So my first step was to begin coding some of that work. 

I was surprised at how well it initially worked. Because of my past as a geologist, I tend to look at longer timescales (weeks, months, years) of learning trajectories, using the tools of interaction analysis and conversation analysis. ‚Äã

---

Jordan, B., & Henderson, A. (1995). Interaction analysis: Foundations and practice. Journal of the Learning Sciences, 4(1), 39-103.
Goodwin, C., & Heritage, J. (1990). Conversation analysis. Annual Review of Anthropology, 19(1), 283-307.

What surprised me was the fractal quality I uncovered. Just as in geology, where the structures observed at the landscape scale are often also findable under the microscope in thin section, here, the patterns I saw over months were also findable in just a few turns of talk.

I reached out to another advisory board member, Dr. Tryphenia Peele-Eady, who directed me to a resource, and quickly diagnosed a dissatisfaction I was having: looking only at the functional linguistics did not fit what I was trying to accomplish. I needed a sociolinguistic approach. 

Together, we refined the coding scheme and conducted some analysis, which we submitted as a conference proceeding. This process of attempting to clarify framing agency also helped clarify for me that it really is a productive construct. While the two teams we analyzed both made decisions, only one team made designerly decisions. This approach easily differentiated between the two teams. 

Gee, J. P. (2014). An introduction to discourse analysis: Theory and method: Routledge.

In my proposal, I named several "peer thought partners" who "have a record of pushing my thinking constructively, and who are already familiar with my research." I reached out to one of them, Brian Gravel, and we worked on some analysis, thinking about the ways agency might play out in heterogenous design, a topic on which we are presenting at AERA 2018. We did some productive pushing and prodding on the construct of agency, which we hope to present at ICLS 2018. 

Finally, I just want to reflect briefly on this experience so far. It has already pushed my thinking. I feel so privileged to have this opportunity, this push to do work I really want to do. In the busy-ness of academia, it is too easy to set aside some of the good but hard work for the right now work. I am so lucky to have this grant that pushes this good but hard work into the right now.