Designing collaborative learning contexts

AS THIS ISSUE ATTESTS, the social aspects of learning command considerable attention in contemporary discussions of schooling. Consistent with the emergence of social constructivist learning theories, concomitant with educational reform efforts that aspire to shape classrooms as learning communities (Brown & Campione, 1994), and spurred on by the design of new technologies that lend themselves to collaborations within and across contexts (Salomon, 1994; Scardamalia, Bereiter, McLean, Swallow, & Woodruff, 1989), there has been a growing interest in the role that interactions with others play in academic engagement and learning. Despite this enthusiasm, orchestrating productive peer learning remains a complex undertaking. In this article, we summarize lessons learned across two programs of research, each of which featured peer collaboration for the purpose of promoting advanced literacies, including text comprehension and scientific reasoning.

While peer learning refers to a host of learning experiences (O'Donnell & King, 1999)--including peer tutoring, cooperative learning, and peer response groups--our interest is peer learning that is designed to promote collaboration. The essence of collaboration is the construction of shared meanings for conversations, concepts, and experiences (Roschelle, 1992). Given this definition, there are certain conditions necessary for promoting collaboration. One such condition is that the thinking is distributed among the members of the group. All members of the group work on the same aspect of the problem at the same time, sharing cognitive responsibility for the task at hand. Furthermore, group members are encouraged to share their thinking as they work together (Brown & Palincsar, 1989). While there are certain forms of cooperative learning that can occur without collaboration, collaborative learning is generally assumed to include cooperation (Chan, Burtis, & Bereiter, 1997).

Given the complexities inherent in collaborative learning, there are many features of the instructional environment to which one could attend. We nominate three such features: the support of interactive patterns, the nature of the problem space, and the process of creating a shared social context. We begin by identifying lessons learned from reciprocal teaching (RT) research and the ways in which this research influenced the design of the second context, cognitive tools and intellectual roles (CTIR), which was designed to enhance student's ability to engage in scientific problem solving.

Two Contexts for Collaborative Learning

Reciprocal teaching (RT)

Reciprocal teaching (RT) was designed as an educational intervention for students who demonstrate significant disparities between their ability to decode and comprehend text. Typically, the participants in RT have fallen below the 35th percentile on standardized assessments of comprehension. During RT, students and teachers take turns leading discussions about shared text. These dialogues are structured using four strategies. Specifically, before reading the text, the group generates predictions regarding the upcoming content. Following the reading of the initial portion of the text, the discussion leader raises questions about the content of the text. The group discusses these questions, raises additional questions, and in the case of disagreement or misunderstanding, rereads the text. Whereas the questions are used to stimulate discussion, summarizing is used to identify the gist of what has been read and discussed and to prepare the group to proceed to the next segment of text. The next strategy, clarification, is used for the purpose of restoring meaning when a concept, word, or phrase is unfamiliar to someone or is a source of confusion to the group. Finally, the discussion leader proposes new predictions for the upcoming text, based on prior knowledge of the topic of the text, personal questions regarding the topic, or based on clues that are provided in the text itself (e.g., embedded questions). This leads to the selection of a new discussion leader and the reading of the next portion of text.

The value of RT as an intervention for poor comprehenders has been documented in many sources (Brown & Palincsar, 1989, Palincsar & Brown, 1984, 1989; Rosenshine & Meister, 1994). In the next section, we identify features of RT that are critical to its role of fostering collaboration by drawing upon a decade of RT research and literally hundreds of conversations with teachers engaged in RT dialogues.

The support of interactive patterns. Rogoff, Matusov, and White (1996) have argued that to promote collaboration there must be the development of an intersubjective attitude--a commitment to find a common ground on which to build shared understanding (Crook, 1994; Rommetveit, 1974). RT's dialogic nature and the explicit goal of making sense of the text provide the context for this interaction. Furthermore, the specific strategies used to scaffold the dialogues both invite alternative views of the text and encourage the participants to achieve consensus regarding the text. For example, the questions and predictions students raise in the course of RT dialogues may be relatively unconstrained by the text. They represent not only what the students have identified as information acquired in the course of reading the text, but also include issues that students perceive as related to the text, assorted recollections that have been triggered by their reading, or wonderings that may, in fact, never be addressed in the text. In contrast, when the students are summarizing or clarifying the text, they are asked to come to agreement regarding issues such as the "big ideas" in the portion of text under discussion and to use the text to support their interpretations.