Exogenous and Endogenous Educational Design

NOTICE: This is LONG...it is an outline of the article.

Reconstructing context: Negotiating the tension between exogenous and

endogenous educational design.

Tabak, I. (2004). Reconstructing context: Negotiating the tension between exogenous and
endogenous educational design. Educational Psychologist, 39 (4), 225-233.

Description
When design and intervention are central to the research process, context as a construct is problematized. How we define context can facilitate or impede our ability to construct rich and vertical accounts of learning. A design stance may predispose us to less profitable notions of context. Tabak presents the advantages of design-based research for understanding how to enact novel forms of learning and for understanding the means through which this learning occurs.

SUMMARY
Data Presented and Supporting Literature:
Two issues pertinent to defining context (p. 225)

• One has to do with demarcating the relevant scope and bounds of the research context because each “context” is itself embedded in yet other broader contexts that impinge on the setting and research results (Barab&Squire, 2004;Cole, 1996, p. 134).
• Second has to do with the underlying metaphors that lend meaning to the term.
• focus on this second issue with respect to design- and intervention-based research

In educational psychology, the thinking that is part of the “contextual turn” includes an appreciation for the complexity of psychological phenomena.

• studies reveal that motivational goal orientations are not stable personality traits, but are rich and malleable constructs that are influenced
  by numerous factors such as task structure and teacher press (e.g., Ames, 1992; Wolters & Pintrich, 1998).

Design-based research methods is a term used to describe a particular stance taken to design- and intervention-based studies of learning in naturalistic settings. This approach, invariably referred to as design experiments, design research, teaching experiments, or development research (Edelson, 2002), is gaining momentum in a number of educational research areas(see recent special issues: Barab, 2004; Kelly, 2003).

What Are Design-Based Research Methods? (p. 226)

• Design-based research methods incorporate both design and empirical research with the goal of developing models and understanding of learning in naturalistic intentional learning environments.
• The design aspect involves designing an intervention that refines a new form of learning to articulate and advance a particular
  position on learning.
• These can include learning materials and curricular structures, activity structures, and instructional strategies
• These designs are based on principles drawn from theory, the literature in the field, and prior research.
  
• The empirical component of design-based research methods studies learning and instruction as it occurs in naturalistic
  settings infused with these designs.
  
• The research includes a series of iterative cycles of design, enactment, retroactive analysis, and redesign.

Why Are Design-Based Research Methods Pertinent to the Study of Classroom Contexts? (p. 226)

• The study of classroom context requires the investigation of more than one variable at a time.
• Classroom context requires a qualitative and inductive component in the research program.
• A study of classroom context should attempt to answer the “how” and “why” questions in addition to the “what” questions.
• The study of context requires that the researcher be present in the classroom.

Deconstructing the “Design” in Design-Based Research Methods (p.227)

• Exogenous design refers to instructional materials, activity structures, or instructional strategies that have been developed
  for the purposes of the research. Exogenous design predominantly reflects the researchers’ or “outsider’s” voice (cf. Fishman, Marx, Blumenfeld, Krajcik, & Soloway, 2004).
• Endogenous design refers to the set of materials and practices that are already in place in the local setting. It also refers to
  those devised by the local participants “in-action” as part of the enactment.
• The boundary between exogenous and endogenous design can be vague.

In Design-Based Research (p. 228)

• Designers often deliberately under-specify the design.
• Teachers adapt designs to suit the needs of their particular students.
• Too specific might constrain teachers and not allow for localization.

• The exogenous design also acts as a communication tool (Ball&Cohen, 1996; Schneider, Krajcik,&Marx, 2000) expressing the goals and pedagogical approach that are embodied in the design. Therefore, designers might choose to omit actions and events that are essential for sustaining classroom life but not central to the innovation and key ideas of the design.

• Together, the exogenous and endogenous designs compose the fabric from which the enactment is formed. Therefore, any explanations of how learning occurred must include elements from both the exogenous and endogenous design.
• Design-based research favors an ecological approach that views designed settings as interacting systems (Cobb et al., 2003).

Study/ Research Conducted
NEGOTIATING THE TENSION BETWEEN EXOGENOUS AND ENDOGENOUS DESIGN:AN EXAMPLE (p. 229)

• The case presented illustrates how design researchers must make decisions whether to press for a particular enactment of
  an exogenous design or embrace a novel enactment. It further illustrates how such decisions can interact in constructive and
  destructive ways with the endogenous design and with our ability to derive veridical accounts of classroom learning.

Background

• The example draws from the Biology Guided Inquiry Learning Environments (BGuILE) project, a science education reform initiative attempting to support inquiry-based science in middle and high school biology classrooms and understand the development of scientific reasoning in classrooms from a sociocultural perspective (Reiser, Tabak, Sandoval, Smith, Steinmuller, & Leone, 2001; Tabak, 1999).
• In answering these questions
  
• First, we craft classroom environments that are intended to realize idealized visions of inquiry-based science.
  
• Second, we study the confluence of our research questions in diverse classrooms.

Evolution of Exogenous and Endogenous Designs (p. 230)

• 1996 study cycle.
• Exogenous design in a K–12 classroom (we worked with a 9th-grade introductory biology class). It was also the first time that the participating teacher, Ms. Patrick, taught the unit.
  
• The computer-based investigation was launched with a whole-class discussion presenting the problem phenomenon and top-level question (why a population of finches is dying and what enables the surviving finches to survive)
• The class broke up into groups and started working on their investigations at the computer.
• As students were working on their investigations, Ms. Patrick and two researchers continually circulated among the groups, engaging them in discussions about their investigations.
• As the enactment progressed, we found that these impromptu discussions with the individual groups were very helpful in sustaining and advancing students’ investigations. We concentrated our efforts on the investigation sessions and on these impromptu group-level discussions. Whole-class discussions amidst the investigation were minimal, consisting mostly of reminding students of the top-level question.
• In our retrospective analysis, found that students did not exhibit the level of sophistication concerning the “how” and “why” of strategies that we had hoped for.
• Therefore, decided that it was important to try and instantiate the whole-class structured discussions in our next cycle.
• Also found that the impromptu discussions with the groups, an element of the endogenous design, were critical to students’ productivity, however, and should be made a purposeful part of the exogenous design.
  
• Recognized that they should also be made a purposeful part of our empirical analysis and should be monitored more closely. We also recognized that we needed to incorporate more explicit support for the strategies in the software.
• 1997 study cycle.
• In contrast to 1996, in 1997, whole-class structured discussions were interleaved with group investigations.
  
• Similar to 1996, the investigation was launched with a whole-class discussion in which the top-level problem was introduced.
  
• In 1997, more emphasis was placed in this introductory discussion on brainstorming ideas about initial hypotheses and ways of exploring these hypotheses.
  
• Also in 1997, we used a revised version of the software, which included more explicit support for the inquiry strategies.
  
• The introductory launching discussion was followed by two investigation sessions (class periods of 45min),
  
• Followed by a whole-class discussion,
  
• Followed by an additional two investigation sessions and a culminating whole-class discussion.
  
• In the intermediate whole-class discussions, student groups reported on the course of their investigations
• The class discussed these processes and the group’s intermediate explanations.
  
• Ms. Patrick questioned students about the questions they raised
  
• The basis for their conclusions.
  
• The specific actions they took.

• In the impromptu discussions, Ms. Patrick would often sit and work with a group for a few minutes as part of the team.
  
• In our retrospective analysis, we examined transcripts of Ms. Patrick’s dialogues during the whole-class structured discussions and during the impromptu discussions.
• We found that the goal of helping students develop an understanding of the “how”—deciding which actions were most appropriate given this inquiry goal and orchestrating a sequence of inquiry actions—was achieved through the impromptu discussions.
  
  • In these discussions, Ms. Patrick would work with the group as a partner for a few minutes, and in doing so, she would voice her reasoning, making these strategies visible to students in relation to their own inquiry (Tabak & Baumgartner, 2004).

• We found that the goals of the “why”—understanding the utility and logic underlying the strategies—were not well met, however, as students did not gain as deep an understanding as we had hoped.
  

• We did find some compelling contributions afforded by the whole-class structured discussions. T
• The first main contribution was that they helped make private learning opportunities public (Tabak, 1999; Tabak & Reiser, 1997).
  
• The richness and significance of teacher–student interactions during group work was contingent on the students’ particular inquiry juncture at the point in time when the teacher approached the group.
  
• However, the whole-class discussions provided an opportunity to import some of these private learning opportunities from the individual group space into the public whole-class discussion space.
  
• The whole-class structured discussions also provided a venue in which the class could relate their findings to central disciplinary principles (Tabak, 1999; Tabak, Sandoval, Smith, Steinmuller, & Reiser, 1998).
  
• This is important in inquiry-based learning because students and teachers can become so immersed in the particular episode they investigate that they lose sight of the broader disciplinary principles that are an important part of the learning goals.
  

Discussion of Case Example (p. 230)

• In this case example, the exogenous design in the first study cycle was underspecified.
• It was expected that the teacher would circulate among the groups as they worked, however the design did not specify topics of discussion and instructional strategies that could be used in these teacher–student group discussions.
  
• It was expected these discussions to facilitate the students’ progress, but did not have clear expectations about what aspects of inquiry learning would be best addressed by small-group versus whole-class discussions.
  
• Elements of the exogenous design that were specified, such as structured discussions, that we intended to serve the “how” and “why” learning goals did not meet our expectations in the enactment.
  
• Initially, were not able to examine our conjecture concerning the efficacy of structured discussions for cultivating “how” and “why” understandings because unable to realize this part of the design in practice.
  
• Example of the type of “online” design decisions that need to be made as part of the microcycles of design and analysis.
  
• There were essentially two possible conclusions that we could draw.
  
• First, we could conclude that the structured discussions were not well suited to Ms. Patrick’s classroom.
  
• Second, we could conclude that, in the first study cycle, the overall intervention was not mature enough to allow for implementing the structured discussions.Given how difficult initial attempts at inquiry can be for both students and teachers (Blumenfeld,Krajcik,Marx, & Soloway, 1994; Krajcik, Blumenfeld, Marx, Bass, Fredericks, & Soloway, 1998), we concluded the latter.
  
• Although some of the learning goals—the “why” goals—had been suspended (Hammer, 1995), other goals—the “how” goals—had materialized through different venues than we originally projected.
  
• In the second study cycle, detailed analysis of the impromptu discussions helped recognize and document that the “how” goals were being met through the impromptu discussions. In a sense, the instructional support for these goals had migrated from their intended locus of the structured discussions to the impromptu discussions. In addition, discovered new roles for the structured discussion that were not part of the original intent, such as bridging between understanding of problem phenomenon and of general disciplinary principles.
  

• The fluidity between the exogenous and endogenous designs helped to realize that some of the learning goals were better met through an aspect of the endogenous design and that additional aspects of learning could be achieved through the exogenous design even though they were not part of the original specification.
  
• If we had maintained a tight focus on the exogenous design in the first study cycle and asserted the whole-class structured discussions, we may have curbed Ms. Patrick’s efforts at the impromptu discussions, keeping her from what proved to be critical to supporting student inquiry. Yet, if we had not persisted in incorporating elements of the exogenous design,we could have missed out on important forms of learning.
  
• For example, we may have forfeited an opportunity for students to relate the specific episode to disciplinary principles. This opportunity cost would have been high because the whole-class structured discussions were one of the few places in which we could identify an explicit attempt to bridge between episode examples and disciplinary principles.
  

IMPLICATIONS FOR THE RECONSTRUCTION OF CONTEXT

• Negotiating between exogenous and endogenous designs is difficult, and being partial to one or the other could result in overlooking important learning mechanisms.
  
• Traditional approaches to design in educational research foreground what I have termed the exogenous design and background the endogenous design.
  
• Although design researchers have justly been cautioned against never abandoning an unpromising design because the conditions of success are continually incorporated into the (exogenous) design (Dede, 2004), I suggest that there is also a danger of presenting distorted accounts of learning that over attribute success to the exogenous design and overlook the role of the endogenous design. This has been argued to occur in other areas of inquiry that took a figure-ground approach to context (Goodwin & Duranti, 1992) or that combined ethnography and intervention (Pors, Henriksen, Winthereik, & Berg, 2002a; Zuiderent, 2002). This figure-ground approach to context has been challenged by scholars in a number of disciplines (Burke, 2002;Goodwin & Duranti, 1992), including education (Lave, 1996).
• An alternative construction of context that is more aligned with its linguistic roots (“to weave”) sees it as intertwined and coconstituting elements that cannot be reduced to “the object” and “that which surrounds the object” (Burke, 2002; Cole, 1996, pp. 135–137), similar to what Cobb et al. (2003) referred to as an ecological approach. In relation to this discussion, an ecological approach or “weaving” metaphor of context views context as the amalgam of exogenous and endogenous designs, similar to what Suchman (1994) referred to as “artful integration.”
  
• The design-based research community has recently been challenged to articulate guidelines for minimizing bias (e.g., Kelly, 2004). As part of this process, we will need to articulate well-specified methods for embracing a weaving metaphor and for attending equally well to exogenous and endogenous design.
  
• In the confines of this article, I can only give this cursory treatment. I hope that drawing attention to the distinction between exogenous and endogenous designs, and the ways in which our conceptions of context can influence our ability to negotiate the tensions that can arise between them, provides an impetus for this future work.Clearly, there is much that we can learn from the existing knowledge base on general (LeCompte & Schensul, 1999) and critical (Carspecken, 1996) ethnographic methods, such as entering the field with multiple observers (designers and nondesigners) or peer-researcher interviews and debriefings that serve to uncover implicit assumptions of bias. We can also benefit from advances in the field of Computer-Supported Collaborative Work that also grapples with the tensions associated with combining ethnography and intervention (see, e.g., a recent special issue: Pors, Henriksen,Winthereik,&Berg, 2002b). In the design-based research literature, there are emerging calls to focus on constructing narratives of change as the products of our field work (Barab & Squire, 2004).
  
• In light of the issues that I have raised here, it will be important to consider who should be the “protagonist” of the narrative. Composing the story around the exogenous design might exacerbate the tensions that I pointed to in this article. A more profitable approach might attempt to construct multiple intersecting narratives with different local participants, such as students and teachers, as the main characters of the story. “Design success rests on the extent and efficacy of one’s analysis of specific environments of devices and working practices, finding a place for one’s own technology within them” (Suchman, 2000).
• If we move away from the notion of intervention as a “prepackaged” artifact that is imposed on the local participants toward a notion of intervention as a process of iterative coconstruction between “outsiders” and “insiders,” we may be in a better position to develop a rich understanding of the complex means that support novel forms of learning.