JULIE E. ROBBS & AMBER BARNHART. HOW DO PEOPLE ... LeBaron, 2002; Koschmann, Zemel, Conlee-Stevens, Young, Robbs, & Barnhart,. 2003; Glenn ...
TIMOTHY KOSCHMANN, ALAN ZEMEL, MELINDA CONLEE-STEVENS, NATA P. YOUNG, JULIE E. ROBBS & AMBER BARNHART
HOW DO PEOPLE LEARN? Members' Methods and Communicative Mediation
Abstract. We are concerned with how learning and instruction are accomplished as interactional achievements, that is the practical details of how participants actually go about doing learning and instruction on a moment-to-moment basis. We focus in this chapter on what we term a problernatizing move, that is a form of social action that has the effect of calling something previously held to be so into doubt. Drawing conceptually and methodologically on Ethnomethodology and Conversation Analysis, we examine problematizing moves in two settings-a problem-based learning (PBL) tutorial meeting conducted face-to-face (F2F)and a distributed PBL (dPBL) meeting mediated through a chat interface. We note that even with the constraints on communication imposed by the mediating technology, the methods employed by members to problematize a problem resemble those seen in F2F meetings. We argue that contrasting members' methods across settings employing different forms of communicative mediation can be instructive with regard to understanding both the effects of the mediation and the nature of the methods themselves.
A book recently published by the National Research Council (Bransford, Brown, & Cocking, 1999) bears the hopeful title, "How People Learn" and provides a summary of research in psychology and education over the last century. Missing from this account, however, is a careful description of how learning is actually accomplished as an interactional achievement within instructional settings, that is, how participants in such settings actually go about doing learning. Presumably, competent participants in such social scenes (i.e., members) have practical procedures (i.e., methods) by which they recognize and display that learning is being done. Ethnomethodology (EM) is a discipline that takes up such matters as its topic of study. The eponymous methods studied by ethnomethodologists might be described as "sense assembly procedures" (Garfinkel & Sacks, 1970, p. 343) in that they are the means by which members make sense of the world around them and their role in it (Garfinkel, 2002; Heritage, 1984). Members' sense-making methods constitute a repertoire of otherwise taken for granted social practices. They provide the basis for our common sense understandings of the world and are the means by which the actions of others acquire their definite sense. Because they are taken for granted, however, they have tended to fall outside the purview of traditional educational inquiry. This oversight, we argue, has served as a barrier to understanding the interactional practices through which instruction is produced and has biased our conceptualisation of learning itself, treating it as an exclusively psychological matter. In a project that has now spanned more than a dozen years, we have been engaged in studying how members do learning in a particular setting,
KOSCHMANN ET AL.
problem-based learning (PBL) tutorials in an undergraduate medical education curriculum. Problem-based learning has been described as a "collaborative, casecentered, and learner-directed method of instruction" (Koschmann, Kelson, Feltovich, & Barrows, 1996, p. 96). Though this description nicely specifies the instructional principles upon which the method is based, it provides little insight into the practicalities of how PBL is actually produced in the nonce. Our research has been designed to address the question, what methods do members use to constitute an activity recognizably and accountably as PBL. Numerous reports have come out of this project (cf., Koschmann, Glenn, & Conlee, 1997, 2000; Koschmann & LeBaron, 2002; Koschmann, Zemel, Conlee-Stevens, Young, Robbs, & Barnhart, 2003; Glenn, Koschmann, & Conlee, 1999; Glenn & Koschmann, in press) and our descriptive approach has evolved over time. Early on, we focused on the talk leading up to the production of learning issues (Koschmann et al., 1997). When students come to recognize a deficiency in their understanding of some matter related to a case, they may take up the matter as a learning issue (LI) to be independently researched following the tutorial meeting (Barrows, 1994). Learning issues are recorded on a whiteboard, so it is a relatively easy matter to trace backwards through the stream of interaction to locate where the recorded issue was initially raised. We termed the intervals of talk leading up to the production of a learning issue Knowledge Assessment Segments (KASs). In communication studies, a segment refers to an intermediate-level unit of discourse, larger than a two- or three-utterance sequence such as the extensively documented IRE sequences (c.f., Sinclair & Coulthard, 1975; Mehan, 1978; Cazden, 1988; Lemke, 1990; Fox, 1993; Wells, 1999) and smaller than episodes of talk such as lessons, classes, or tutorial meetings. We defined a KAS as, "a topic-delimited segment of instructional discourse in which participants raise a topic for discussion and one or more members elect to display their understanding of that topic" (Koschmann et al., 1997, p. 2). The production of a LI is predicated on three conditions: "there must be a recognizable knowledge deficiency, the students must see the missing knowledge as relevant to or necessary for the eventual practice of medicine, and, finally, there must be consensus about the timeliness of undertaking the study" (pp. 1-2). There remains, of course, the task of providing a more practical account of how this is routinely accomplished in interaction and this was our focus of initial study. As we became more deeply involved in this work, however, we began to find our initial analytic strategy to be too restrictive and our labelling to be unsatisfactory. Though the production of learning issues is a crucial aspect of the method, there is much, much more going on in these meetings than simply producing a list of items for later study. We also began to find the notion of assessment to be a bit problematic. Though something very much like assessment is entailed in producing a LI, it doesn't necessarily always occur whenever participants display an understanding with respect to some matter. The term assessment also has two
H O W DO PEOPLE LEARN?
usages potentially relevant to tutorial discourse. In educational psychology, assessment usually denotes measurement of a subject's understanding using some pre-defined instrument. In communication studies, on the other hand, assessment is used to describe action, vocal or otherwise, that is treated interactionally as a display of judgement with regard to some matter (cf., Pomerantz, 1984a; Goodwin & Goodwin, 1987). To dispel any possible ambiguity and also to mark a broadening of the scope of our analytic interests, in Koschmann et al. (2000), we adopted the label Knowledge Display Segments (KDSs) to describe our object of study. In so doing, we no longer limited ourselves to studying intervals of talk that resulted in the production of a LI, but instead sought out fragments in which one member raised a topic for discussion and others elected to display their understanding(s) with regard to the that matter. We focused in particular on the tutor's role in facilitating the production of KDSs. In related work (Glenn et al., 1999; Glenn & Koschmann, in press), we examined how diagnostic theories were occasioned within tutorial meetings. The two activities, learning issue production and theory generation, both involve a display of understanding and both are a collaborative achievement. In Koschmann and LeBaron (2002), we introduced the notion of learner articulation as an analytic category that brings together all of these elements (i.e., knowledge assessment and theory generation, knowledge display and the integration of understandings across participants). Following on Koschmann et al. (1996), we defined learner articulation as having two aspects: "the act of giving utterance ... to force a cohesive explanation" and "the action of jointing or interrelating of concepts and relationships" (p. 93). The second aspect of learner articulation highlights the interactional nature of knowledge display-the need to produce collective understandings by negotiating expressed individual understandings. We argued that learner articulation is not simply a lexical matter but also an embodied phenomenon and we extended our analysis to include the ways that participants used their hands, bodies, and material environments as resources in joint sense-making. Our analytic strategy was one of identifying instances in which speakers employed gestures coupled with talk to display their understanding and create connections to the displayed understandings of others. Learner articulation provides a broad and useful means of conceptualising what participants in PBL tutorials are doing, but it falls short of serving as a member method by which the activity of PBL is recognizably produced. Learners gather in tutorials not necessarily to articulate understandings so much as they gather to engage in a form of joint problem solving. While learner articulation may be occurring, therefore, it is not the object of the activity per se from the perspective of the participants. To understand the activity as problem solving, therefore, we needed to go beyond learner articulation to understand how problems themselves are produced as interactional achievements. Koschmann, Kuutti, and Hickman (1996) observed that many current theories of learning assume a precipitating breakdown in understanding as a starting point for learning to occur. For the American pragmatist philosopher John Dewey, for example, learning is initiated in situations that have turned problematic. Dewey (193811991) defined inquiry as the process whereby the problematic aspects of the situation are ultimately resolved. He made clear, however, that he was not talking
KOSCHMANN ET AL.
about problems as posed instructional exercises. Dewey wrote, "A problem is not a task to be performed which a person puts upon himself or that is placed upon him by others" (p. 111). Instead, it is a matter for learners themselves to discover. For that reason, Dewey described inquiry itself as "a progressive determination of a problem and its possible solution" (p. 113). Koschmann (2001) argued that a key contribution of PBL is that it affords a setting rich in opportunities for doing inquiry in precisely the sense described by Dewey. Dewey's useful conceptualisation, however, does not explain how the "progressive determination of a problem" is actually accomplished interactionally. In more recent work (Koschmann et al., 2003), we have attempted to address this issue by documenting how some matter comes to be treated as problematic by participants in a PBL tutorial meetings. We focused in particular on what we termed the problematizing move, that is a form of social action (e.g., utterance, gesture, facial expression) that has the effect of calling into doubt some matter previously taken as so. Problematizing moves perform two kinds of work: they direct attention to some potentially problematic matter and, at the same time, project some form of collective action with regard to that matter. A problematizing move, however, does not in and of itself produce a problem in the sense described by Dewey until it is taken up as such by all the parties to the conversation. In prior writing (with the exception of Koschmann et al., 2003) we have focused on the methods members employ in face-to-face (F2F) tutorial meetings to negotiate mutually-endorsed understandings and evaluate the adequacy of these understandings. In our most recent work, we have begun to examine alternative models of PBL, models in which the constraints that tutorial participants convene in the same place or even at the same time are relaxed. Such approaches are sometimes described as "distributed" (Koschmann, 2002) versions of PBL to differentiate them from the more conventional, F2F variety. To do distributed PBL (dPBL), some form of mediating technology is inevitably entailed. A logical extension of our work done to date might be to ask how is problematizing accomplished in non-F2F PBL tutorial meetings? In this chapter, therefore, we will carefully examine the methods employed in problematizing some matter in a dPBL meeting mediated through a computer-based, textual (chat) interface. Before doing so, however, we will re-examine a fragment of interaction in a F2F tutorial meeting that has been described in a number of earlier publications (Koschmann et al., 1997, 2000; Koschmann, 2001; Koschmann & LeBaron, 2002) this time focusing on the methods employed to problematize some matter. We will conclude by discussing how all of this might be related to overcoming some of the biases and barriers to better understanding how learning occurs in settings of collaboration and the role of technology therein. 2. DATA 2.1 Fragment 1: "What would be the risk?" 2.1.1 Setting PBL involves small teams of students (five or six is considered optimal) working with a faculty tutor who learn in the process of working through a collection of
HOW DO PEOPLE LEARN?
clinical teaching cases. Within tutorial meetings, data are gathered, hypotheses generated and tested, and conclusions drawn in an interactive manner similar to that used by medical practitioners. Curriculum designers use various means to simulate encounters with patients. Trained patient surrogates, known as "standardized patients" (Barrows, 1987), are used for some cases. For others, a paper-based simulation, the Problem-Based Learning Module (PBLM), was designed to allow for free inquiry, providing responses for any question, examination, or laboratory test a physician might request for an actual patient (Distlehorst & Barrows, 1982). In the implementation of PBL that we studied, the tutor guided students through their discussions of teaching cases by modeling reflective reasoning Information and ideas that emerged during discussion were organized and recorded on a whiteboard by a student designated the scribe. The whiteboard was sectioned to hold information of specific types--emerging evidence with regard to the case (Data), hypotheses of underlying cause (Ideas), matters for further study (Learning Issues), and developing plans for future inquiry about the patient (Actions). In the fragment analysed here, the medical students (all identified by pseudonyms) were enrolled in the second-year of the problem-based learning track. The tutor ("Coach") was highly experienced and widely recognized for his skill in teaching in collaborative settings. During the segment of talk represented below, these students were in the end of a Gastrointestinal/Endocrine/Reproductiveunit that occurred late in their second year of medical school. The group was introduced to this case by performing a history and physical on a standardized patient. The interval of talk analysed here took place during a follow-up meeting and the students were engaged in the final stages of processing the patient problem. They sat at a table containing notes, textbooks, and materials relevant to the case under discussion. The group used a white board to record notes pertaining to the patient's status, diagnostic theories, and learning issues. At the time the fragment begins, the students have collected most of the information usually derived from the patient history and physical and they are discussing possible lab tests.
2.1.2 The Case The case was authored by a medical school faculty member and was based on actual clinical records. It involved a 32-year old female patient (L.M.) who presented with a complaint of abdominal pain of two weeks duration worsening in the last two days. This was associated with chills, nausea, and malaise. On questioning, the patient also revealed experiencing a burning sensation on urination and a small amount of yellowish vaginal discharge. An abdominal exam revealed right lower quadrant tenderness and the pelvic examination indicated cervical tenderness. During the pelvic examination, cultures of the cervix were taken which revealed infection with Chlamydia, a sexually transmitted bacteria that can commonly cause pelvic inflammatory disease (PID). Pelvic inflammatory disease is a consequence of untreated sexually transmitted diseases of the lower genital tract. It is postulated that the infection ascends to the upper genital tract (uterus, fallopian tubes and ovaries), causing clinical symptoms of fever and abdominal pain. Advanced cases of PID can result in abscesses in the fallopian tubes or ovaries; these may not respond to
antibiotics and may require surgical drainage of the abscess. An aspect of the final sequelae of PID is that scarring may occur after infection and inflammation resolves, which can produce infertility. As the fragment begins, the group was discussing how to further investigate L.M.'s abdominal pain using radiographic imaging, given their hypothesis of PID. Computerized axial tomography (CT) scans are often performed on patients with abdominal pain to look for causes of the pain. In patients with PID, CT scans are useful to look for complications such as abscesses, and reactive inflammation of adjacent organs. CT scans employ a thin beam of X-rays that are generated to pass through the patient's body; the radiation that passes through is picked up on the other side by an electronic radiation detector. Since the x-ray tube and the detector are mounted on opposite sides of a circular gantry, rotating the gantry as the measurements are being made every few degrees and using a computer to reconstruct that image will result in a cross-sectional image of that part of the patient. The amount of radiation patients are exposed to in an abdominal CT scan is around 10 mSv (millisieverts). The millisievert is a scientific unit of measurement for radiation dose. This is around 3 times the amount of radiation the average person in the U.S. receives each year from naturally occurring background radiation (from radon, cosmic radiation etc.). This is much more than the amount of radiation exposure from one standard plain film chest x-ray (0.lmSv). Different tissues and organs have varying sensitivity to radiation exposure, for example, organs that contain more rapidly dividing cells or the rapidly dividing cells in an embryo or fetus are more sensitive to harm from radiation exposure. In this case, there is a question of whether or not L.M. might be pregnant. The amount of radiation from an abdominal CT scan might be harmful to a developing fetus, and most physicians would prefer to avoid exposing the fetus to that risk if at all possible. A glossary of some of the medical terminology related to the case discussion can be found in Appendix D. 2.1.3 Analysis In describing the activity seen here as problem-based learning we must be clear about what we are taking to be "the problem." There are, in fact, an assortment of problems to consider. There is the problem that originally brought the patient into the physician's office, that is the patient problem. As a case, this encounter with the patient represented a clinical problem for the attending physician. When the case is selected and used for instructional purposes, it becomes a teaching problem. As the tutorial members work through the details of the case, they may display incomplete, discrepant, or otherwise inadequate understandings which, when taken up, are treated as problems of understanding. There are forms of interactional work involved, however, in producing problems of understanding and we were interested in documenting the methods members use to accomplish this. Ethnomethodology defines a topic of study (i.e., members' methods) and a locus (i.e., within practical activity), but does not specify any particular methodology by which it should be studied. Conversation Analysis (CA) is an area of specialization within ethnomethodological research that focuses on the methods members use in
HOW DO PEOPLE LEARN?
interaction to make sense of each other and to, in turn, be seen and heard as sensible (Heritage, 1984). CA furnishes a methodological framework for the rigorous study of unfolding interaction (c.f., Psathas, 1995). Studies of member methods always begin with extensive fieldwork. It is difficult, however, to study the details of conversation in situ as it is being produced. Observational notes, compiled in the field, are often not an adequate basis on which to construct an analysis. Conversation analysts, therefore, usually work with detailed transcripts built from field recordings, either audio or video. In this study, we have both incorporated findings from the CA literature and employed CA methods and transcription conventions. Our analytic strategy in this study was to collect examples of what appeared to be problematizing moves, that is actions that have the demonstrated effect of (1) indexing some matter as potentially problematic and, at the same time, (2) marking it as actionable, i.e. projecting some form of collective action with regard to the matter at hand. Analysis of the materials gathered in the field was done in phases of individual and group effort. Preliminary analyses were presented in collaborative "data sessions" to extend and enrich the analysis presented here. In describing what might count as "ethnographically adequate description," McDermott, Gospodinoff, and Aron, (1978) argued that analysts must provide readers with sufficient ethnographic material to enable the reader to independently evaluate the proposed analytic claims. To this end, we provide in Appendix B a transcript of the full exchange upon which our analysis was based. Prelude to a problematizing. The fragment begins with one of the students, Joel, making a proposal based on prior reading for how the case should be assessed. Patrick immediately offered a concern regarding the proposal: Excerpt 1.1 1 Joel: 2 3 4 5 Patrick: 6 7 Jackie: 8 Patrick:
They did talk about doing a CT along with along with it if you feel there are abscesses but it's low yield unless you feel there are abscesses= =Think you can [get can a lot of risks doing a CT 1= L ~ n dan ultrasound can be used for that reason] = to the pelvis.
Patrick's utterance indexes a potential problem ("a lot of risks") but lacks specificity. Offered as a question, Joel i.esponded to it as such and then asked Patrick to better articulate his concern ("What would be the risk?'). Jackie (lines 13, 15) interjected that there would only be a risk under certain circumstances. Patrick reformulated his "risks of doing a CT" to the "danger of X-raying the ovaries and that." There was difference of opinion, therefore, as to whether there was a safety concern at all and, if so, under what circumstances. If it constituted a problem of understanding, however, it was not one for which the group had yet established a sense of collective ownership. At this point, Coach intervened:
KOSCHMANNET AL. Excerpt 1.2 20 Coach: 21 Jackie: 22 23 Patrick: 24 Joel: 25 Jackie: 26 Patrick: 27
Is there a ?risk to CT? OHmO (0.9) Sure. l t l s an ?x-[ray. L~erah h t ' s an X-ray, there's always a risk to an X-ray.
Coach's question was produced with an accompanying gesture similar to that of a crossing guard stopping traffic. His question had a similar effect on the flow of the conversation-it momentarily arrested the discussion and set it in a particular direction. Because of his role as tutor, his question elevates Joel's concern to a matter warranting the attention of the collective. But the nature of the danger alluded to by Patrick had yet to be articulated. In the discussion that followed, Coach probed the students to get them to articulate their respective understandings, focusing on the differences between CT scans and X-rays. Joel, in formulating a position with respect to this matter, raised a question of his own: Excerpt 1.3 40 Joel: 41
?What is the dosage (0.4) relative ( ) from uh normal X- ray to a C T i
Joel's question reshaped the issue under discussion, from that of some largely unspecified risk associated with CT scans to the amounts of radiation received from CT scans and conventional X-rays. Jackie (lines 47-48) indicated that she did not remember the "relative dosage." After some additional discussion on the part of Coach, Jackie, Joel, and Patrick, Joel offered an answer to the question he had raised in the prior excerpt: EXCerDt 1.4 75 Joel : 76
I understand that the CT is ?just about equivalent to an X-Jray.
Joel's position, carefully qualified ("I understand that"), constitutes a display of understanding. In producing this formulation, Joel set the scene for a problematizing move. Problematizing the problem. Though Joel's display of understanding was performed by an individual speaker, its production was a collaborative effort. It built upon Patrick's original expressed concern and Coach's leading questions. Most directly it tied back to his own question in line 40-41. The problematizing move was also produced collaboratively: ExCerDt 1.5 75 Joel: 76 77 Coach:
I understand that the CT is ?just about equivalent to an X-Jray. Is it?
HOWDO PEOPLE LEARN? 78 79 80 81 82
That's what my understanding &is I- I'm not ) r ~ * just m saying ( Melissa: LWhy don't we just put it up as a learning issue. Joel: >Let throw that [up< Jackie: L~eah.
Coach's query directs attention to Joel's expressed understanding. This move has particular force owing to Coach's recognized expertise and his role as a tutor. It assesses Joel's understanding without providing an explicit assessment. Schegloff, Jefferson, and Sacks (1977) described how participants in a conversation display a preference for speaker self-repair over listener correction. In this case, Coach's question creates a space for Joel to do self-repair without specifying exactly what needed to be repaired or offering an alternative formulation. Joel's response acknowledges that his prior formulation has been challenged as potentially problematic, but rather than offer an alternative formulation instead reiterates his qualifying condition, "That's what my understanding is." This is an example of telling "how I know" (Pomerantz, 1984b). By formulating this as only an "understanding," Joel explicitly marks his proposal as defeasible. Coach's query is seen and heard as directed to Joel. This is made visible by the fact that Coach turns to look at Joel while issuing the question. It is also evident from the adjacent positioning of the query and the "cohesive" (Halliday & Hasan, 1976) deployment of anaphora. Despite the fact that the question is so marked as Joel's to answer, Melissa responds in overlap with Joel (line 80) and proposes that the matter be taken up as a LI. Melissa's "it" is subtly different from Coach's. Coach's "it" refers to Joel's displayed understanding, Melissa's to the underlying issue to which Joel's display constituted a response. Melissa's proposes uptake of this matter as a collective concern. It treated the matter as actionable and proposed a specific course of action. Joel and Jackie both ratified this proposal. Postscript to a problematizing. In response to Melissa's proposal, Alice rose from her seat and recorded "CT vs. x-ray risks" on the board as a LI. By this action, the group formally agrees to address the problematic issue by investigating it after the meeting. Though this would appear to resolve the matter for the purposes of the current discussion, Coach pushed it a bit further: Excerpt 1.7 83 Coach: 84 85 Joel: 86 87 Joel: 88 89 Coach: 90 91 Joel:
>I was going to say< how sure are you on a scale of zero to [ten. Lnot (0.6) Three? (0.8) Think we oughta make a ?learning issue out of it. r ~ h e hheh heh t ~ a(maybe we ought to)
Coach again withheld assessment of the correctness or incorrectness of Joel's original formulation (lines 75-76). Instead he spoke to the way in which Joel
KOSCHMANN ET AL.
qualified his answer ("my understanding is") and pressed Joel to quantify his level of certainty. Joel confessed that his level of certainty with regard to this matter was low. The students had now displayed an understanding (such as it was), had acknowledged that their understanding was incomplete, and had resolved to redress that deficiency. One might think that it was time to move onto new things, but Coach was apparently not quite finished: EXCerRt 1.8 92 Jackie: 93 Melissa: 94 Jackie: 95 Coach: 96 Jackie: 97 Coach: 98 Jackie: 99 Coach: 100 101 102 103 Jackie:
know [that ultrasound isn' t radiation L (Joel) this is from the ( 1
Right?= =No, h l trasound isn ' t radira: tion. L~verybodyknows that. But if you're gonna get [hesitate what I ' m gonna I'say, doing a scan, because the woman might be preqnant then I think you oughta know if the CT scan is to be of concern or ?not when doing it on a pregrnant woman. L~i::~ht.
In this final excerpt, Coach provided a justification for his original question, "Is there a risk to CT?'(line 20). He offered an argument that the matter would be of importance to the students in their later clinical work. In particular, Coach is modelling a form of professional practice-whenever a clinician makes a decision based on some belief, it is necessary to assess the certainty with which the belief is held. 2.1.4. Discussion In Koschmann et al. (1997), we described this same fragment in terms of the work required to produce a learning issue. In Koschmann et al. (2000), we documented the tutor's role in this process. In Koschmann and LeBaron (2002), we examined the ways in which the participants in the 'What would be the risks?' fragment used gesture in articulating their understandings. Finally, in Koschmann (2001), this fragment was analysed as an example of Deweyan inquiry enacted. Here, we revisit it once again to document the work involved in producing a problem in understanding. Collaborative effort was required to build toward a display of understanding and collaborative work was needed to produce a problem of understanding. Coach's query (line 77) implicitly places Joel's expressed understanding in doubt. Like Patrick's initial expressed concern, however, Joel's problematized understanding was still an individual matter, since there was no evidence of group ownership. Melissa's proposal that it be taken up as a learning issue transformed it into a collective concern which was immediately ratified by the other members of the group. We defined the problematizing move as an action that has the "demonstrated effect" of (1) indexing some matter as potentially problematic and (2) projecting some form of collective action. In this case, the problematizing move was accomplished jointly by Coach and Melissa and evidence of uptake was provided by the other students through their endorsement of the matter as a learning
H O W DO PEOPLE LEARN?
issue. We now turn our attention to a problematizing move performed in a CMC discussion. 2.2 Fragment 2: "Salicylate toxicity" 2.2.1 Setting The students participating in this PBL tutorial were third-year medical students doing a clerkship rotation in Family and Community Medicine. The six-week clerkship begins with a short orientation and training period at the main campus, followed by a five-week preceptorship experience in which each student is assigned to a practitioner in the community. The students provide patient care in both ambulatory and hospital settings under the direct supervision of the preceptor. The preceptorships occur in different communities scattered across the state. Because most students enrolled in the clerkship are located off campus, it would be inconvenient to convene F2F tutorial meetings during their clerkship. 2.2.2 Mediating Technologies Students in the Family and Community Medicine clerkship are supplied with laptops and have dial-up access to the Internet at their preceptor sites. Weekly PBL tutorial meetings are conducted using off-the-shelf CMC software (Microsoft NetMeetingTM). NetMeeting offers a chat facility as well as a shared whiteboard. Groups of four to six students interact with a faculty tutor to explore selected patient cases and identify learning issues. Between weekly meetings, students independently investigate these issues from their preceptor sites. Students post information gathered about the learning issues to a discussion forum using another off-the-shelf collaboration tool, WebCT. When the group reconvenes the following week, the learning issues are reviewed, the case is closed, and a new case is opened. Interaction among participants was constrained by the way the technology was organized for use. Typical steps in posting a message were as follows: (1) A participant composed a message by typing it into the message field provided by the chat software. (2) Once a participant completed a message, he or she posted the composed text to a server for distribution to recipients. (3) Once the message was received by the server, it distributed the message to recipients at different locations. The posting and distribution of messages in this manner created certain interactional difficulties for participants. For example, a message distributed by a server may have been received at a participant's computer while the participant was engaged in typing his or her own message, while reading a previously posted message or while doing anything else, as long as the computer was connected to the chat session. Because participants did not see messages as they were being composed, something like an overlap could occur between the receipt of a sender's message and a recipient's actions. In such cases, the sender of a message did not have access to the activities of recipients while the message was being composed or sent, or knowledge of when or if recipients actually received his or her message. How participants oriented to and responded to the arrival of a new message is an important interactive
KOSCHMANN ET AL.
feature of chat room interactions (cf., Herring, 1999). In a recent paper, Garcia and Jacobs (1999) observed that the organization of turn taking in computer-mediated communication is different from that in spoken conversation. They note, for example, that the notion of "transition relevance place," so central to the Sacks, Schegloff, and Jefferson's (1977) "simplest systematic" for coordinating turns at talk is lost in what Garcia and Jacobs describe as "quasi-synchronous CMC" (p. 339). They further contend that the meaning of pauses and the methods available to participants for self-repair are also different in this mode of discourse. We made certain presumptions about the messages posted and received in these tutorial meetings. Previously posted messages were resources to which all participants in the chat had access. This does not presuppose that all messages were in fact relevant to each recipient or were resources to which recipients necessarily oriented in the production of their own messages. It only suggests that posted messages constituted a corpus that participants could use as resources in the production of subsequent messages. Each posted message was designed to be intelligible to recipients. This may seem a bit obvious, but we expand the sense of intelligibility to include, in addition to specific content, some sense of the intended recipient(s) as well as the relation a given message may have had to previously posted messages. From this presumptive base, it is possible to inspect posted messages as they were received at the server and draw some tentative conclusions. As the following excerpts show, even in a quasi-synchronous electronic chat environment, clinical problems and learning issues, which are central to problembased learning approaches (Barrows, 1994), are not given ready made but are jointly negotiated and socially-organized achievements. 2.2.3 The Case The case under consideration concerned a patient, R.W., a 55 year-old farmer, who was brought to the Emergency Department by ambulance after his wife found him wandering, cold and confused, outside their home earlier that morning. In the ambulance, an intravenous line was started, R.W. was given oxygen, and his blood sugar level was checked and found to be elevated. R.W. had no prior history of diabetes (high blood sugars). On physical examination, he was noted to be breathing with very deep breaths, despite not complaining of any shortness of breath or difficulty breathing. This type of breathing is termed "Kussmaul's respirations" and is associated with conditions that cause an imbalance in the body's pH. The students ordered a laboratory test, termed arterial blood gases (ABG), that measures the pH, oxygen and carbon dioxide content in the arterial blood. Two such tests had been done on R.W., one when he arrived at the emergency room and another a few hours later. Since R.W. was noted to have high blood sugar levels (glucose) and had Kussmaul's respirations, students had initially hypothesized he might have a condition called diabetic ketoacidosis (DKA). This is a complication of uncontrolled diabetes that results in the disruption of normal aerobic metabolism, and a switch to anaerobic metabolism. Normal pH is around 7.4. Anaerobic metabolism, however, produces ketone bodies which can ultimately turn the blood pH acidic. An ABG reading made when the patient was admitted to the hospital, however, showed a
HOWDO PEOPLE LEARN?
slightly alkalotic pH of 7.44. R.W. had been taking large amounts of prescription and over-the-counter painkillers. The students, therefore, also ordered a toxicology screen to see if R.W. had toxic levels of acetaminophen or aspirin (salicylates). The test indicated that R.W. had toxic levels of salicylates in his bloodstream. An overdose of aspirin may initially cause the blood pH to appear normal because of respiratory alkalosis and a metabolic acidosis. As the toxicity progresses, however, a patient with salicylate toxicosis is unable to compensate for the metabolic acidosis by lung or kidney mechanisms and severe acidosis results. Indeed, this was what was seen in the second arterial blood gas measures for R.W. It is at this point that we begin our analysis of the interaction. While there was agreement among participants that a severe acidosis was occurring, one participant remained uncertain as to whether the change was due to primarily metabolic or respiratory mechanisms. Clinically, it is important to understand which mechanism is responsible for the body's acid-base imbalance because treatment depends on first normalizing the pH, then on correcting the primary cause of the imbalance. Because of this, the plans for treating R.W.'s problem would differ depending on the primary mechanism involved. 2.2.4 Analysis As in the 'What would be the risk?' fragment, our interest here is in members' methods for problematizing a matter, though in this case in a computer-mediated tutorial meeting. The same analytic strategy was employed for collecting examples of problematizing moves, that is actions that have the demonstrated effect of (1) placing some matter into doubt and (2) projecting some form of projective action with regard to that matter. Interaction is automatically transcribed in CMC, simplifying the work of the analyst. The transcript so produced, however, is not like a conventional CA transcript, such as the one found in Appendix B. Conventional CA transcripts capture details of delivery (e.g., timing, intonation) not relevant to CMC. Other aspects of a conversation analytic approach to studying interaction do remain relevant, however. A shared objective for all CA studies is to give an account of the order and placement of each contribution to a conversation. An omni-relevant question is, why this and why here? Sequential organization is a crucial resource for meaning making for members and analysts alike. This is true not only of F2F conversation, but also of computer-mediated discourse. The meaning of any particular utterance/contribution/message can be displayed, acknowledged, and, in some cases, shaped in the unfolding sequence that follows it. As Heritage (1984) described it: [C]onversational interaction is structured by an organization of action which is implemented on a turn-by-turn basis. By means of this organization, a context of publically displayed and continuously up-dated intersubjective understanding is systematically sustained. (p. 259)
This is also true of messages produced in computer-mediated exchanges. Because of this, the methods for studying the sequential organization of talk in F2F conversation are also useful in studying computer-mediated exchanges. Other
researchers have already employed CA methods in studying CMC (cf., Arnseth, Ludvigsen, Wasson, & Morch, 2001; Garcia & Jacobs, 1999). A transcript of the complete fragment can be found in Appendix C and a glossary of some of the medical terms and abbreviations can be found in Appendix D. Putting forward a candidate matter of concern. In the first excerpt, a participant put forward a matter of concern to her and solicited consideration of this matter from other participants in the PBL tutorial. As is evident from the transcript, participants resisted its uptake as a matter of concern for the group. Excerpt 2.1 1 3:40:58 2 3
S1: S2 :
I was impressed with the depth and sources you had for your posted learning issues. What did you think of what others posted? nice job everyone. Did everyone find mine? everyone did a good job but I still can't really figure out why the initial ABG was alkalotic- can anyone help me out I think everyone did a good job . . . I certainly enjoy reading posts like Sl's, where the info is related back to the case. Strong work Sl! lab error, S2 ;)
In lines 7 through 10, S2 used the tutor's invitation to evaluate previously posted learning issues (lines 1 through 4) as a way of introducing a matter for the group to consider. S2's message oriented to the tutor's invitation by offering a cursory evaluation of previously posted learning issues followed by an initial formulation of a matter as problematic. S2's actual formulation of the problematic matter is notable for a number of reasons. First, S2 localized the problem as a deficit in her knowledge or understanding: "I still can't really figure out why the initial ABG was alkalotic" (lines 7 through 9). In asking for assistance from the group as a whole, she constituted members of the group as resources that were to be called upon to help her, i.e. "can anyone help me out" (lines 9 and 10). This formulation of the query implied that recipients were able to understand the nature of the problem presented by S2, assess the extent and limits of their own and S2's understanding of the problem based on the formulation of the query, and respond either to resolve the issue or to take up the matter for further investigation for how it might lead to the formulation of a clinical problem. Second, the laboratory finding is problematic because the group has previously committed to a diagnosis of DKA, but S2 did not explicitly implicate this diagnostic theory in her message. This excerpt ends with S4's post at line 15 which displayably oriented to S2's request since it was explicitly addressed to S2 and constituted the first identified uptake of S2's problem. However, the formulation of the post suggests that S4 was actually downgrading the significance of S2's message by invoking "lab error" as
HOW DO PEOPLE LEARN?
the explanation and then offering the gestural "wink" marker. S4's message was framed as a possible reason for "why the initial ABG was alkalotic" but the wink marker, i.e. ;), suggested that the response was not to be taken seriously. This "joking" response was in the form of a candidate resolution of the difficulty. However, the wink marker indicated to S2 and to other recipients that the post was not to be taken as a serious response to the query but was to be seen as a comment on the query itself, implying that S2's query was of a sort that warranted this kind of non-serious response. S4's downgrade also could be seen to imply that responses from other group members were not warranted. S4's downgrading response treated S2's problem as a problem that did not warrant uptake by the group. Thus we see that uptake of a member's issue by the group as a whole is not always a simple matter. Two possible explanations for the group's reluctance to take up the matter are suggested by the postings. First, the fact the S2 tried to divert the group from attending to its own accomplishments could be seen as abrupt and unsympathetic to the effort others expended in producing previously posted resolutions of learning issues. The resistance of the tutorial group to abandon its concern with its own accomplishments also suggests that these accomplishments, incremental or partial though they may have been, were very important to the way the group understood itself. It could have been very difficult for this group to maintain itself as a group without an acknowledged sense of accomplishment. A second reason that might account for the group's unwillingness to take up S2's question in a serious way, and one that constituted S4's response, may be that an anomalous lab result will not always warrant further scrutiny. Perhaps S2 had not provided enough information about the possible consequentiality of the anomaly to pique the interest of other members of the group or make evident its nature as a problem. We hold as a matter of course that for an anomaly to be taken up as a problem, it needs to be recognized by the participants themselves as significant, relevant and problematic for the entire group. In other words, the fact that a tutorial member did not understand an anomalous laboratory result did not necessarily mean that the laboratory result was problematic for the group. The problem could have been attributable to the tutorial member herself. Of course, a combination of these reasons could also account for the observed reluctance of members to take up S2's request for help. The problematizing move. The response S2 received suggests that a matter of concern to an individual participant was not always considered a matter of concern for the group. Thus, in order for S2's concern to be taken up by a group of this sort, additional work needed to be done. In the next excerpt, we see evidence of the work done to transform S2's initial query into a matter for the entire tutorial group to investigate. This involved a reformulation of the initial query and its subsequent uptake by other members of the group. As the transcript demonstrates, the tutor's endorsement of the matter contributed in a large way to the group's uptake of the matter.
S1: S1: S2 : S1: S1: T: S5:
you points S5. you won points does everyone think we're dealing with DKA? I thought S6 may have been excessive yes, DKA I posted the initial and repeat ABGs, what do all think of S2's first and second question? I'm a little confused by that pH as well S2.
When there was no subsequent uptake of her initial inquiry by any other members of the tutorial group, S2 produced a second query. S2's second question makes explicit the matter being placed in doubt (the diagnosis of DKA). By addressing her question to "everyone," S2 elevates the matter to one of collective concern. Her message was formulated in a way that called on recipients to weigh in either in support of or in opposition to the proposed candidate cause (see Pomerantz, 1984b). 2.4.3 Uptake of the matter by the group. Uptake is an artful, interactional achievement and is often delicately done, especially if it had been resisted previously by members of a group. In this case, we can see how uptake was done in ways that also displayed specific alignment with S2. Excerlst 2.3 22 3:44:18 23 24
I posted the initial and repeat ABGs, what do all think of S2's first and second question? I'm a little confused by that pH as well S2. salicylate toxicity causes a=resp alkalosis initially he has over compensated. now is entering resp. alk. I sort of think this is a mixture of ASA toxicity and DKA . . .can that be possible? So is it mixed metab acidosis and resp alkalosis? yes yes yes ...and salicylates can cause this
As in the previous fragment we see an example of the faculty member pursuing a "teachable moment." The tutor here first oriented the attention of the group to the anomalous lab finding, posted on the NetMeeting whiteboard. By introducing additional clinical data, the tutor endorsed Sl's uptake of S2's problem of understanding, thus indicating this was a matter for the group to take up.
HOW DO PEOPLE LEARN?
Once the tutor had endorsed S2's queries by posting additional clinical data, she then called on all participants for an evaluation of S2's queries (lines 22 to 24). S5 addressed a message explicitly to S2 (lines 25 to 26) in a way that specifically aligned with S2 and the tutor's endorsement of the anomaly S2 had identified. By addressing his message to S2, S5 may have been doing work to explicitly display alignment with and to further endorse the uptake of S2's queries for how they stood for his own problem of understanding the matter and thus pointed to possible problems for the group to take up, especially given the difficulties S2 had encountered with respect to getting the group to take up her queries. Other members of the tutorial group took up discussion of the anomaly once these endorsements from F and S5 were made available to the group. S4, S1 and S5 engaged in an interchange in which various ways of accounting for the clinical anomaly were put forward (lines 27 through 34). With these possible candidate causes for the observed anomaly now available to members of the group, the tutor called on the group to weigh in with respect to a particular candidate cause that would account for the observed anomaly (lines 33 and 34). What follows is a series of alignments in lines 35 through 37. S4 did more than just align with the tutor's position, however. S4 expanded on that alignment by citing a specific basis, i.e. "salicylates", for the candidate cause put forward by the tutor. Building on this, the tutor queried the group to describe salicylate toxicity and this became a new topic of discussion.
2.2.5 Discussion There are many differences between the 'What would be the risk?' fragment and the 'Salicylate toxicity' fragment and not all are due to the mediational circumstances under which the described events were conducted. In the latter fragment, the students are at a later stage in their training and the discussion is more sophisticated. More is left to inference and the matter in question is much more complex, having to do with dynamic processes in multiple, unfolding systems. In the first case, considerable work went into formulating a display of understanding before the problematizing was undertaken. In the second fragment, the matter being placed in doubt was first implied and only named as the problematizing move was produced. In 'What would be the risk?', the matter problematized became a learning issue for the group; in 'Salicylate toxicity,' the problem of understanding became a catalyst for a different form of collective action-the reformulation of a diagnostic theory. In both cases, however, a similar form of work was required to produce a problem of understanding. That is, in both cases, some matter being called into doubt was referenced by the participants and some form of collective action was projected and undertaken. As the analyses suggest, even in highly mediated forms of interaction that are constrained by the affordances of the technology used, members of PBL tutorials need to enlist the interest, involvement and support of other members of the group to elevate a matter from a personal or idiosyncratic problem of understanding to a problem of understanding for the group as a whole. As with the face-to-face interaction examined earlier, the work done to get participants in distributed learning to take up a matter as a collective problem of understanding involved calling on
participants to take positions with respect to the matter at hand. When taking positions, members displayed the scope and limitations of their understanding of or knowledge about the matter at hand. The surprise finding, therefore, is that even with the documented barriers to communication imposed by the mediating technology (e.g., the lack of "coherence" [Herring, 19991, the absence of prosodic and gestural cues, disruption of ordinary turn-taking strategies as documented by Garcia & Jacob ), the interactional methods employed by members to problematize a problem (at least with respect to the initial steps of establishing a problem) resemble those seen in F2F meetings. This is an important finding and one with important implications for instruction. It should, for example, come as welcome news for those who would like to pursue the use of PBL in distributed environments. As with the face-to-face interaction examined earlier, this interaction showed a clear orientation among participants to learner-directed instruction. In fact, it was a student who put forward a matter of concern to the group and it was the same student who persisted in putting the matter forward when other students were reluctant to address the matter. The tutor in this instance worked with the students to allow them to work out for themselves the relevance of the matter. An important feature of the tutor's work involved withholding explicit evaluation of student responses to S2's queries. By withholding her evaluation of the responses to S2's queries, the tutor provided the students with the opportunity to develop their own positions with respect to matter S2 put forward. By explicitly calling on students to develop their own positions with respect to S2's queries and Sl's response, the tutor could be seen to be implicitly calling on the other students to offer their own evaluations. Rather than enact a conventional IRE sequence (c.f., Sinclair & Coulthard, 1975; Mehan, 1978; Wells, 1999) in which the teacher puts forward an inquiry, the student responds, and the teacher evaluates the response, it was the students themselves who produced all three components. By making students accountable for the evaluation of the responses they themselves put forward, the tutor displayed an orientation to the interaction as learner-directed instruction. Likewise, by accepting that accountability within their interaction, both the students and the tutor constituted their work as learner-directed instruction.
3. BARRIERS AND BIASES TO UNDERSTANDING COMPUTER-MEDIATED INTERACTION We begin from the premise that the greatest barriers to understanding instructional interaction are conceptual and methodological. This applies both to instruction carried out F2F and under distributed and technologically-mediated conditions. The disabling bias in educational research has been and continues to be in treating learning and instruction as exclusively psychological matters. Instruction is produced through the joint interactional work of teachers and learners. It stands to reason then that if instruction is, first and foremost, an interactional matter, then it should be studied as such. Further, if learning is something that is produced in and by instructional discourse, then evidence that learning is being done should be made available within the same discourse. Herein lies the importance of documenting
HOWDO PEOPLE LEARN?
members' methods of sense making in instructional settings, for it is in and through these methods that instruction and learning are both achieved and produced. The PBL literature consists almost exclusively of outcome studies designed to determine whether or not PBL works (Albanese, 2000; Colliver, 2000). These assessments of efficacy, however, take the treatment variable (i.e., the instructional intervention) as a given. No one to date has undertaken the difficult task of documenting the actual practices by which PBL is recognizably and accountably produced (Hak & Maguire, 2000). The work described here endeavours to remedy this omission. As such, it represents a departure from prior research on PBL. As our studies shift from PBL in F2F settings to distributed implementations of PBL, we have begun to examine members' methods for producing PBL in tutorial meetings mediated by technology. In this way our work speaks not only to what it means to do PBL, but also represents a novel approach for studying the effects of communicative mediation. Rawls in her introduction to Garfinkel's Ethnomethodology's Program (2002), wrote, "methods used by persons to create the orderliness of ordinary social occasions are constitutive of those occasions" (p. 6). She went on, "Constitutive practices are those which can only meaningfully exist if they are made recognizable by those who practice them" (p. 6). Garfinkel, in establishing EM as a new field of inquiry, took understanding of the orderliness of ordinary social occasions as the central question. Rawls, in fact, defined Ethnomethodology as, "the study of the methods people use for producing recognizable social orders" (p. 6). Documenting the methods by which recognizable orderliness is produced is by no means a simple matter, however. As the overview of our earlier work provided in the introduction would attest, an adequate characterization of members' constitutive practices can be elusive. The evolution of our descriptive framings from Knowledge Assessment Segments to problematizing moves reflects our ongoing struggle to produce a usable account of what participants in a PBL tutorial are actually doing. Numerous challenges lay ahead. As we have described, the task of producing adequate description of the interactional practices by which members produce instruction is neither trivial nor straight-forward. We also have yet to develop a practical methodology for applying the findings of descriptive research to pedagogical design work (c.f., Koschmann, Zemel, & Stahl, 2004). Nonetheless, it is only when we become more articulate about the methods members use to produce what recognizably passes as instruction that we will be able to make meaningful advances toward designing artefacts to support collaborative forms of instruction. As the analysis presented here demonstrates, contrasting members' methods across settings employing different forms of communicative mediation can be instructive with regard to understanding both the effects of the mediation and the nature of the methods themselves. Such studies may also contribute in the end toward answering the more basic question, how do people learn.
KOSCHMANN ET AL.
ACKNOWLEDGEMENTS Support for the development of the distance learning program described here came through a grant from the U.S. Dept. of Health & Human Services, Bureau of Health Professions. Partial support for the first author while preparing this paper was provided through a grant from the National Science Foundation (EHR 01-261 04).
NOTES 1 A summary of the PBL process as practiced at this institution can be found in Koschmann et a1.1996). An expanded description can be found in Barrows (1994). The problem-based learning track was discontinued in the fall of 2000.
These transcription conventions, developed by Gail Jefferson, are summarized in Appendix A. 3
By "demonstrated effect" we mean that through an analysis of the unfolding sequence, an utterance (or succession of utterances) can be demonstrated to do the work of problematizing in the way described here. This is never based solely on semantic content of the utterance, since evidence must be provided of uptake of the matter by the speaker's interlocutors (see Sacks [I9921 on the analysis of sequence in talk). 4
Digitzed video for this fragment can be found on the CD-ROM that accompanied Koschmann and LeBaron (2002).
5 Though not necessarily novel in other respects. There have been a number of studies carried out from an ethnomethodological perspective in instructional settings (e.g., Fox, 1993: Lynch & Macbeth, 1998; Ford, 1999) and there is a strong tradition of doing ethnomethodologically-informed research in work related to CSCW (c.f., Button & Sharrock, 2000; Clarke et al., 2003). We know of no other work, however, that focuses specifically on the details of how instructional interaction is produced under different communicative circumstances.
REFERENCES Albanese, M. (2000). Problem-based learning: Why curricula are likely to show little effect on knowledge and clinical skills. Medical Education, 34,729-738. Arnseth, H., Ludvigsen, S., Wasson, B., & Morch, A. (2001). Collaboration and problem solving in distributed collaborative learning. In P. Dillenbourg, A. Eurelings, & K. Hakkarainen (Eds.), Proceedings of the First European Corzference on Computer-Supported Collaborative Learning (pp. 75-82). Maastricht: Maastricht McLuhan Institute. Barrows, H.S. (1994). Practice-Based Learning: Problem-Based Learning applied to rnedical education. Springfield, IL: Southern Illinois University School of Medicine. Bransford, J., Brown, A., & Cocking, R. (1999). How people learn: Brain, mind, experience, and school. Washington, D.C.: National Academy Press. Cazden, C. (1988). Classroom discourse: The language of teaching and learning. Portsmouth, NH: Heinemann. Colliver, J. A. (2000). Effectiveness of problem-based learning curricula: Research and theory. Academic Medicine, 75,259-266.
H O W DO PEOPLE LEARN?
Dewey, J. (199111938). Logic: The theory of inquiry. In J. A. Boydston (Ed.), John Dewey: The Later Works, 1925-1953, Vol. 12. Carbondale, IL: SIU Press. Distlehorst, L.H., & Barrows, H.S. (1982). A new tool for problem-based self-directed learning. Journal of Medical Education, 57, 466-488. Fox, B. (1993). The human tutorial dialogue project. Hillsdale, NJ: Lawrence Erlbaum. Garcia, A., &Jacobs, J. (1999). The eyes of the beholder: Understanding the turn-taking system in quasisynchronous computer-mediated communication. Research on Language and Social Interaction, 32, 337-368. Garfinkel, H. (1967). Studies in ethnonzethodology. Cambridge, U.K.: Polity Press. Gartinkel, H. (2002). Ethnonzetltodology's program: Working out Durkheirn's aphorism. Lanham, MD: Rowman & Littlefield Publishers. Garfinkel, H. & Sacks, H. (1970). On formal structures of practical action. In J.C. McKinney & E. Tiryakian (Eds.), Tl~oreticalsociology: Perspectives and developrrzents (pp. 337-366). NY: Appleton-Century-Crofts. Glenn, P. & Koschmann, T. (in press). Learning to diagnose: Production of diagnostic hypotheses in problem-based learning tutorials. To appear in M. Maxwell, D. Kovarsky, & J. Duchan (Eds.), Diagrlosis as cultural practice: An account of the power of language in diagnosis. The Hague: Mouton. Glenn, P., Koschmann, T., & Conlee, M. (1999). Theory presentation and assessment in a problem-based learning group. Discourse Processes, 27, 119-133. Goodwin, C. & Goodwin, M. (1987). Concurrent operations on talk: Notes on the interactive organization of assessments. Papers in Pragmatics, 1, 1-52. Hak T., & Maguire, P. (2000). Group process: The black box of studies on problem-based learning. Acadentic Medicine, 75,769-72. Retrieved October 27, 2000 as: http://www.academicmedicine.org/cgi/content/ Halliday, M.A.K. & Hasan, R. (1976). Cohesion irz English. London: Longmans. Heritage, J. (1984). Gat-finkelarzd Etlzrto~rtetl~odology. Cambridge, MA: Polity Press. Herring, S. (1999). Interactional coherence in CMC. Journal of Computer-Mediated Conznzunicatiort, 4(4). Retrieved April 15,2003 as: http:Nwww.ascusc.org/jcmc/vol4/issue4/heing.html Koschmann, T. (2001, March). Dewey's contribution to a standard of problem-based learning practice. First European Conference on Computer-Supported Collaborative Learning (EuroCSCL), Maastricht, Netherlands. Retrieved April 5, 2001 as: http://www.mmi.unimaas.nl/euro-cscVPapers/9O.pdf Koschmann, T. (Ed.)(2002). Studying collaboration in distributed PBL environments. Distance Education, 23(1). Koschmann, T., Glenn, P., & Conlee, M. (1997). Analyzing the emergence of a learning issue in a problem-based learning meeting. Medical Education Online, 2(2). Retrieved on October 15,2002 as: http://ww~v.n~ed-ed-onlinc.ordrcs00007.ht1n Koschmann, T., Glenn, P., & Conlee, M. (2000). When is a problem-based tutorial not tutorial? Analyzing the tutor's role in the emergence of a learning issue. In D. Evensen & C. Hmelo (Eds.), Problern-based learning: A research perspective on learning interactions (pp. 53-74). Mahwah, NJ: Lawrence Erlbaum Assoc. Koschmann, T., Kelson, A.C., Feltovich, P.J., & Barrows, H.S. (1996). Computer-supported problembased learning: A principled approach to the use of computers in collaborative learning. In T.D. Koschmann (Ed.), CSCL: Theory and practice of an ernerging paradigm (pp. 83-124). Hillsdale, NJ: Lawrence Erlbaum Associates. Koschmann, T., Kuutti, K., & Hickman, L. (1998). The concept of breakdown in Heidegger, Leont'ev, and Dewey and its implications for education. Mind, Culture, and Activity, 5, 2 5 4 1 . Koschmann, T. & LeBaron, C. (2002). Learner articulation as interactional achievement: Studying the conversation of gesture. Cognition & Instruction, 20, 249-282. Koschmann, T., Zemel, A., Conlee-Stevens, M., Young, N., Robbs, J., & Barnhart, A. (2003). Problematizing the problem: A single case analysis in a dPBL meeting. To appear in B. Wasson, S. Ludvigsen, & U. Hoppe (Eds.), Designing for change. Amsterdam: Kluwer Academic Publishing. Koschmann, T., Zemel, A,, & Stahl, G. (2004, June). The video analyst's manifesto (or The implications of Gartinkel's policies for the development of a program of video analytic research within the learning sciences). In Y. Kafai, W. Sandoval, N. Enyedy, A. Nixon, & F. Herrera, Proceedings qftlte Sixth International Cortfererzce qf the Learning Sciences (pp. 278-285). Mahwah, NJ: Lawrence Erlbaum Associates.
KOSCHMANN ET AL.
Lemke, J. (1990). Talking science: Language, learning, and values. Norwood, NJ: Ablex. McDermott, R., Gospodinoff, K., & Aron, J. (1978). Criteria for an ethnographically adequate description of concerted activities and their contexts. Semiotica, 24, 245-275. Mehan, H. (1978). Learning lessons. Cambridge, MA: Harvard University Press. Pomerantz A. (1984a). Agreeing and disagreeing with assessments: Some features of preferredldispreferred turn shapes. In J.M. Atkinson & J. Heritage (Eds.), Structures of social action: Studies in Conversation Analysis (pp. 57-101). Cambridge, UK: Cambridge University Press. Pomerantz, A. (1984b). Giving a source or basis: The practice in conversation of telling 'how I know'. Journal ofPragnzatics, 8,607-625. Psathas, G. (1995). Conversation analysis: Tlie study of talk-in-interaction. Thousand Oaks, CA: Sage. Sacks, H. (1992). Lectures on conversation. Oxford, U.K.: Blackwell. Sacks, H., Schegloff, E., & Jefferson, G. (1974). A simplest systematics for the organization of tumtaking for conversation. Language, 50,696-735. Schegloff, E., Jefferson, G., & Sacks, H. (1977). The preference for self-correction in the organization of repair in conversation. Language, 53,361-382. Sinclair, J.M. & Coulthard, M.C. (1975). Towards an analysis of discourse; The English used by teachers and pupils. NY: Oxford University Press. Wells, G. (1999). Dialogic inquiry: Towards a sociocultural practice and theory of education. NY: Cambridge University Press.
HOWDO PEOPLE LEARN?
Appendix A: Transcription conventions
I Marks the beginning and end of temporal overlap among
Brackets Equal sign
utterances produced by two or more speakers. Indicates the end and beginning of two sequential 'latched' utterances that continue without an intervening gap. Measured in seconds, a number enclosed in parentheses represents intervals of silence occurring within (i.e., pauses) and between (i.e., gaps or lapses) speakers' turns at talk. A timed pause of less than 0.2 sec.
Indicates a fallinu pitch or intonational contour at the
Rising vocal pitch or intonational contour at the
Exclamation point Comma
I conclusion of a t i h constructional unit (TCU).
I conclusion of a TCU. An inverted question mark (i) I represents a half rise. I Marks the conclusion of a TCU delivered with emphatic and animated tone. Indicates a continuing intonation with slight upward or downward contour, as in the enunciation of an item in a not yet completed list, occurring (generally) at the end of
a - TCLJ .
Greater than/ Less than signs Degree signs
Arrows Breath sounds
I I I I
An abrupt (glottal) halt occurring within or at the conclusion of a TCU. A colon indicates sustained enunciation of a syllable vowel, or consonant. Longer enunciation can be marked using two or more colons. Portions of an utterance delivered at a noticeably quicker (> c) or slower (c >) pace than surrounding talk. Marks speech produced softly or at a lower volume than surrounding talk. Represents speech delivered more loudly than surrounding talk. Underscoring indicates stress on a word, syllable or sound. Marks a rise (1')or fall (J)in intonation. Audible expulsion of breath (linguistic aspiration) as in laughter, sighing, etc. When aspiration occurs within a
KOSCHMANN ET AL.
Appendix B: "What would be the risk?" Tape #95-001.1 (1:55:25 to 1:57:46)
Patrick: Jackie : Patrick: Joel : Joel : Jackie: Patrick: Jackie:
(?I Patrick: Coach: Jackie:
They did talk about doing a CT along with along with it if you feel there are abscesses but it's low yield unless you feel there are abscesses= =Think you can rget can a lot of risks doing a CT L ~ n dan ultrasound can be used for that reasonj = to the pelvis. (0.5) No why. (2.5) What would be the rjsk. W u h only if it was [ectopic. L~eah ( . ) (but probly) rror if she was =nantl
=Well even even ( . ) well would you have ( . ) danger of X-raying ovaries Oand that0 Is there a ?risk to CT? OHmO
HOWDO PEOPLE LEARN? Patrick: Joel : Jackie: Patrick: Coach: Patrick: Coach: Patrick: Joel : Joel : Joel : Patrick: Jackie :
Coach: Jackie: Coach : Jackie: Coach: Jackie: Joel : Coach: Joel : Jackie : Coach : Patrick: Joel : Jackie: Joel : Jackie: Coach: Joel : Coach : Joel : Melissa: Joel :
(0.9) Sure. ltfsan ?x-rray. L~erah
Lit's an X-ray, there's always a risk to an X-ray. (0.5) I mean ( . ) what is the risk of a CT is there a difference between X ( . ) uh CT and an ordinary X-ray? (0.2) Yea:h (0.2) uh C-tee 1:s ( . ) urn:: ( . ) in uh p1a:ne (0.2) Yu:h So: I would think that the CT (1.0) would be: (0.6) instead of just a plain fi:lm (0.4) would be more X-rays being used .hhhh (0.5) *What is the dosage (0.4) relative ( . ) from uh normal X- ray to a CTi [~CTis *serial CT* Ois ito serial X-rays >is it not