This is the anonymized transcript of my Frontiers in Education (FIE) 2014 conference talk. The paper title is "Using Realtime Transcription to do Member-Checking During Interviews" and the authors are myself (Mel Chua) and Robin S. Adams. Since the paper was about realtime transcription, I did not use a slide deck. Instead, I projected a CART feed (live captions) for my own presentation as I spoke, so that my audience could see me demo what I'd written about. The transcript below is therefore both (1) what was said verbally, and (2) what was shown on the screen. All names, except mine and Lynn's (who requested to be identified in this transcript), have been changed.
MEL CHUA: Okay. This is actually a two-slide presentation. Oh, hi, everybody, I'm Mel. I'm from Purdue and I'm doing that dissertation thing and I do qualitative research, so I do lots of interviews. But there's one little bit of a wrinkle -- I'm deaf, so talking is hard.
[new slide explaining the CART acronym]
One thing I use in my classes for accessibility is something called CART. It stands for Communication Access Real-Time Translation. And that's what it looks like. And it's basically a stenographer who comes and types super, super fast on a magic chording keyboard what people are saying, which is one of the reasons I'm wearing a microphone and a microphone will be passed around the room. What ended up happening was I just used CART for my interviews. And one of the side benefits of CART is you get a transcript right while you're talking.
[new slide with the agenda in bullet-points]
What I'm going to be presenting is what it looks like, and what are some of the implications of CART, because it actually has some interesting implications for research methodology and subject positionality and the kind of interactions you have during the conversations.
I thought that the best way to do that would actually be to show you what this is. And so everybody, say hello to Becky. (Note: name has been changed.) Becky is my captioner for today.
[switches from slides to a live Streamtext of the CART for the talk -- the realtime transcript of the event scrolls on the projector for the remainder of the talk]
THE AUDIENCE: Hello, Becky.
MEL CHUA: Becky, everyone says hi. Do you want to say hello to everyone?
BECKY (typed on the screen): (Hi, everyone!) (How are you today?)
MEL CHUA: Yeah, sometimes people ask me, "What speech recognition software are you using?" It's not a software, it's a person. So that's the point. I wanted to show people a little bit about what it looks like and what can happen when you do this kind of thing during an interview. Tom was kind enough to volunteer to do a mini demo. (Note: name has been changed.) Hi, Tom.
TOM: Hi, Mel.
MEL CHUA: Tom, can you tell me a little bit about the balance you strike between research and how you do including diversity in the classroom?
TOM: And including diversity in the classroom?
MEL CHUA: Yeah.
TOM: That's a great question. I just started a new job at a teaching university in [SCHOOL NAME] so I don't have structured expectations to teach. I teach 9 credits, 3 courses this semester. That's considered on the light side compared to some of my colleagues. So I have a structural place for teaching.
And I have a structural blessing -- an institutional blessing to do research but I don't really have a structural place to do it. So ways that I kind of -- I incorporate strategies by -- I strategize by collaborating with other universities that give me a structure. I have two days off. Not really off. But I'm not teaching classes. For two afternoons, Monday and Friday.
And that really -- on Friday I focus on research. Monday, teaching. And how that -- and I bring myself to foster diversity in the classroom. By trying to be attentive to those who -- I'm in a [small school in a US state]. It's a pretty homogenous population so I do notice -- and we are very male dominated. I do notice when those -- come in my classroom that may not feel like they identify with everyone else. So there were some strategies with stereo -- protecting against stereotype threat that I try to incorporate in the classroom. Is that about time?
MEL CHUA: Yeah, that's awesome. Thanks, Tom. And so if you hold on for a moment, you might notice we actually have a transcript up already. One thing we can do right now is scroll back a bit and say, Tom, it's member check time. Hang onto this for a moment. [hands Tom the microphone]
MEL CHUA: I'm going to scroll through real quick some bits of what you said, and if you see something that seems interesting and pops out to you and you want to talk a little more about it, stop me and then just say that.
TOM: Okay. And by the way, this is completely unrehearsed. So you know. (Audience laughter) Okay. I'm looking. I start off pretty descriptive. Probably to gain comfort with the question. And so I talk with what I know. Right off the bat. So yeah, go down a little bit... and if we looked up, I'm -- I have like some -- maybe go up -- back up, I'm sorry. It's like right in between. In an in-between spot.
Yeah, where I say I have a structural place for teaching and I have a structural blessing -- an institutional blessing I'm correcting myself there to do research but I don't really have a structural place to do it. I kind of -- I have some pause about that. Because I'm thinking, oh, what if this were to get out. And how would this reflect on the university that hires me and feeds my wife and kids and me.
But I'm okay with it. But it does give me some pause whenever I see it. So we're good. Do you want me to keep going? Okay.
MEL CHUA: Thanks, Tom. Keep in mind this was a quick demo. In an actual research interview you would go much longer and much more in depth. Just from this you can see a couple of implications. First of all, member checking can be done in the same session as the interview, so the dropout problem that you have -- it can get at that a little bit.
Second, the positionality. So instead of subject interviewer, Tom became sort of a co-analyser or co-researcher and his reflecting on his own words and so forth. While it was still fresh in his mind. And that's actually something that a lot of participants do in interviews anyway. In Holstein and Gubrium's "The Active Interview" from 1995. [pause to let captioner catch up] Wow. Okay. They talked about indigenous coding, which is when people are in the middle of an interview and they say things like, "Oh, just like I said before." Or, "This is a good example of..." and they start analyzing and reflecting on what they have said.
But the difference is that what Tom was able to do, you saw that he went back up to a portion of his transcript and then quoted the exact words he said instead of having to remember it. So it's indigenous coding. But it's grounded in the direct verbatim words of a transcript.
Another thing this does is it makes transcription more visible as a methodological choice we're making. A lot of times we just go, oh, transcripts, transcription, transcription. But that's not actually the case. The choices we make can have a big impact on way we analyze and way we present our findings. This makes it much more visible.
There are some downsides. You've got to set this up in advance. Because it's a person. You have another person's schedule to juggle, but it's much like if you were doing a foreign language interview and needed a Spanish translator or something, not that different.
Cost-wise it's about for -- shall -- it's about like paying an undergrad to transcribe thing except it's faster, much faster.
Some people feel a little weird when doing this. I will sometimes use Google Docs and have a transcriber write into a Google Doc so we can correct it, correct typos in the middle. Some people report it's a little distracting to see their words pop up so they don't look at the screen. Everyone responds to it in different ways.
And I wanted to close off by there's a few folks in the room that's been subject to this particular method and I wanted to give them a chance to speak to what that's actually like.
Dave. I asked him before [the talk if he'd like to speak]. (Note: name has been changed.)
DAVE: One thing I notice is when you were going through this, you kind of switched from -- from this descriptive mode to then reflecting on what you're describing, and I found that in participating in this, I would -- in particular when we would go back to the transcript in a follow-up session, I would look at things that I said and just enter this reflecting... "Hmmm, why did I use that word, or why did I use that phrase, and do I actually think that, and how do I feel about this being transparent and out there for the world to see?"
It's certainly a sense of unease at times, but I actually found it really useful for development in my own thinking to look at my words both as they were appearing and then a week later or so or two weeks later and reflecting why in the world I would say certain things. Yeah, it's... terrifying, also. (Audience laughter)
MEL CHUA: I'm continuing to experiment with this stuff, I'm happy to talk with people about it. Robin Adams, my advisor and co-author, is sitting right there and can also speak to what this method. She's been on both ends of the microphone.
I thought I would leave some time for questions because this might be the first time a lot of people have seen this.
AUDIENCE QUESTIONER: Thank you. Does it lend itself available for all types of analysis for example analysis where you do have (Audio cutting in and out) -- analyzers and those are kind of getting lost, as well?
MEL CHUA: Are you still getting audio okay?
BECKY (captioner, typed on the screen): (It was cutting in and out a little bit).
>> MEL CHUA: Okay. I guess that would answer my question.
So that's a great question. And actually one of the things CART does is that it makes transcription very visible as a deliberate choice of methodology, so it's probably not the right choice for super precise verbal protocol analysis type stuff. Also, when I do this, I always make a backup audio recording just this case something cuts off or I want to go back and make sure that's exactly what was said. Because, yes, you do lose some of the precision just like you would with any type of interpreter or translation type thing. I think of this as sound detection. So it's not appropriate for everything. But when it's just communicative, [it works].
AUDIENCE QUESTIONER: (Speaker off mike).
MEL CHUA: Yeah; yeah. The backup gives me a good idea if there's a part that was super fast or I had a really weird, you know, Russian author name that we need to track, that kind of thing. Any other questions?
AUDIENCE QUESTIONER: So in interviews when people hear voices on audiotapes a lot of times people are like, "That's what I really sound like?" or on TV, they are embarrassed, like, "That's what I look like?" Here you see a little bit of that effect, I guess like, "That's what I really said." Is that hard to get past? Is that part of the protocol, is that what you're analyzing?
MEL CHUA: [to Lynn Andrea Stein, another audience member] Do you want to answer that question? (Note: Lynn is Lynn's actual name; she requested to be identified in this transcript.)
LYNN: I knew that was my question. (audience laughter) Mel and I started using this protocol when she was interviewing me for [a research project], and I can't read the transcription. And we actually switched to using a method in which I just type to Mel and we type back and forth, because as good as Becky and her colleagues are, it's exactly that I can't stand to hear my voice, I can't stand to see somebody else's transcription. I worry a lot about precision, so for me, because I can type fast enough that we can have a real-time conversation, this method was really difficult. And part of me is grateful -- I love the real-time conversation and analysis of the conversation as it goes. And I also wish that I could tolerate this method because I think it's great and I just can't get myself to do it.
MEL CHUA: Thank you, Becky I think we need to switch over to the next person. Thank you, again. (Audience applause)
BECKY (captioner, typed on the screen): (Thank you!)