Download - Collaborative knowledge building: A case study

Collaborative Knowledge Building: A Case Study

Nancy J. GilbertMarcy P. Driscoll

What learning conditions are necessary toeffectively support the knowledge-buildingenterprise in a learning community? Toanswer this question, the case-study methodwas employed to investigate collaborativeknowledge building in a graduate level coursedesigned to incorporate specific constructivistlearning principles. These principles included(a) having a collective and authenticcommunity goal to facilitate collaboration andengagement in the community, (b) usingcooperative groups to provide for socialnegotiation and to promote multipleperspectives, (c) personally selecting of coursereadings and group decision-making toencourage ownership in learning, and (d)employing an integrated set of technologytools to archive the work of the community andfacilitate communication. Results indicatedthat the instructional strategies designed intothe course promoted collaborative knowledgebuilding and the acquisition of key conceptsthrough self-directed learning strategies.Results also pointed out the types oftechnological support required to implementan effective community of learners. A proposedset of guidelines is discussed to assistinstructional designers in the design anddevelopment of collaborative,knowledge-building environments.

In an information society, knowledge movesinto the forefront and becomes an object withwhich people work (Bowen, Bereiter, & Scar-damalia, 1992). This suggests that students mustbecome knowledge workers to function effec-tively in today’s digital age. A constructivistphilosophy of education may offer a means ofteaching students to become knowledgeworkers, because constructivist practices areconcerned with a learner’s ability to use andmanipulate information in authentic situationsthat have high relevance for the student (Lebow& Wager, 1994). Constructivist learning environ-ments can provide opportunities for students toaccess, use, manipulate, and evaluate informa-tion using authentic problem-solving activitiesand strategies that focus on generative teachingand learning.

Making a shift toward a constructivist educa-tional philosophy requires a change in focusfrom individual knowledge constructed singlyto public knowledge jointly constructed by stu-dents (Brown, 1994; Brown & Campione, 1990;Scardamalia & Bereiter, 1992; Wenger, 1998). Aconstructivist orientation represents a fun-damental change in education where the goal isto modify closed classrooms into knowledge-building communities that enable students tocontribute to each others’ learning through so-cial construction of communal knowledge(Lebow, 1995).

Collaborative Knowledge-BuildingCommunities

Knowledge-building communities arose fromthe idea that schools should be restructured ascommunities in which knowledge is constructed

ETR&D, Vol. 50, No. 1, 2002, pp. 59–79 ISSN 1042–1629 59

AAH GRAPHICS, INC. / (540) 933-6210 / FAX 933-6523 / 03-04-2002 / 17:27

as a collective goal (Scardamalia & Bereiter,1994). The main focus of activities within theseenvironments is on developing the collectiveknowledge base of the community and improv-ing the problem-solving expertise of thelearners. According to Lebow, Wager, Marks,and Gilbert (1996), a key principle guiding thedesign of knowledge-building communities isthat the artifacts or objects produced by thelearners are not simply passed in for a grade, butbecome public materials that support the goalsof the community and its individual members.The resulting database of information archivesthe materials produced by the learners and isevidence of the community’s advancingknowledge and expertise (Lebow et al., 1996).

This shift to a communal approach to learn-ing changes the focus of the teaching and learn-ing process. Bowen et al., (1992) characterizedknowledge-building communities with fourprimary traits:

1. A focus on knowledge and the advancementof knowledge rather than tasks and projects.

2. A focus on problem solving rather than per-formance of routines.

3. Dynamic adaptation in which advancesmade by members of the learning com-munity change the knowledge conditions re-quiring other members to readapt, resultingin continual progress.

4. Intellectual collaboration as members poolintellectual resources, making it possible forcommunities to solve larger problems thancan individuals or small groups.

In the last decade, increasing attention hasbeen paid to how computers can be used to sup-port and facilitate learners as they interact andsolve problems (Pea, 1994). This emerging inter-est is concerned with the design of tools thatcontribute to collective activity, and is charac-terized by authentic, collaborative workfacilitated through the use of networked com-puters. The use of networked computersprovides alternatives to traditional teaching andlearning as we move from a single-classroomconcept to the concept of a knowledge-buildingcommunity of learners.

The computer supported intentional learningenvironment (CSILE), developed by Scar-

damalia and Bereiter (1992), is an example of alearning environment that incorporates theprimary traits of knowledge-building com-munities and computers as tools. CSILE at-tempts to reform student achievement through avariety of methods, including collaborativelearning, and a focus on group projects ratherthan lectures. The groups are given the tools toresearch a particular topic, solve problems as-sociated with that topic, and present the infor-mation to their peers for review. Networkedcomputers are used as the primary vehicle topromote collaborative knowledge building be-cause they facilitate easy input and retrieval ofinformation as well as providing a variety ofcommunication features. Collaborativeknowledge building engages students with thetopic instead of leaving them on the outside aspassive observers. Learners do not rely onsimple memorization skills but must engage inhigher-order thinking and inquiry as they arriveat a variety of solutions to a particular problem(Scardamalia & Bereiter, 1994).

In CSILE, knowledge building, rather thanknowledge replication or retrieval, is the focusof student activity (Scardamalia & Bereiter,1994). Knowledge in this environment isdynamic, and is changed and reconstructed overtime. That is, students jointly build a body ofknowledge represented by the communitydatabase that is an aggregate of the communityrather than the individual.

More than 10 years of research on CSILEshow that this approach can have a powerfulimpact on learning (Bereiter & Scardamalia,1996; Scardamalia & Bereiter, 1996). Results ofstudents in grades 5–6 using CSILE in all contentareas suggested that students demonstrate ahigh level of knowledge building compared tonon-CSILE students. Knowledge building in thiscontext is the construction of new informationstored in the communal database associatedwith the formulation of questions to help stu-dents notice what information to add to thedatabase. Results also showed that a largeproportion of the questions generated by CSILEstudents are rated as “challenging and con-sidered worthy of research” by teachers andother adults (Bowen et al., 1992, p. 92).

Additional research with CSILE demon-


60 ETR&D, Vol. 50, No. 1

strates that students “greatly surpass students inordinary classrooms on measures of depth oflearning and reflection (awareness of what theyhave learned or need to learn), and under-standing of learning itself” (Bowen et al., 1992, p.282). Individual achievement, as measured bystandardized tests, reflected increases in reading,language, and vocabulary for CSILE studentscompared to their non-CSILE counterparts (Scar-damalia & Bereiter, 1992). Finally, Bowen et al.,(1992) suggested that CSILE students appear tofunction beyond their years in solving problemsand constructing knowledge at levels “one doesnot find in ordinary schools, regardless of thecaliber of students they enroll” (p. 282).

Supporting Knowledge-BuildingCommunities with Technology

How do computers function to facilitateknowledge building in a community oflearners? Technology is typically discussed interms of hardware and software. However,Jonassen (1995) posed the notion that technol-ogy is more than hardware and software and in-cludes instructional strategies and teacher-student relations. Technology, in a constructivistlearning environment, engages students in theenterprise of constructing knowledge and com-prises tools that enable learners to build per-sonal interpretations (Jonassen, 1995) byconstructing and manipulating information(Perkins, 1991). In other words, technology com-prises tools that extend cognitive functioning(Jonassen, 1995); that support, guide, and extendthe thinking process (Derry & LaJoie, 1993; Han-nafin, Land, & Oliver, 1999), and that engage thelearner in critical thinking (Jonassen, 1995). Han-nafin et al. presented a taxonomy of tool typesaccording to the specific cognitive functions theysupport, and described integration tools as thosethat “help learners to link new with existingknowledge” (p. 129). Such tools also supportknowledge representation, annotation links, andelaboration functions that help learners in “or-ganizing ideas from a variety of perspectivesand integrating them with personal knowledge”(Hannafin et al., p. 129). Because of these func-tions, integration tools should be particularly ef-

fective in supporting knowledge-building learn-ing communities.

Jonassen (1999) argued further that toolsused to support knowledge-building learningcommunities must also promote discourseamong community members and open access toshared information. “Textualizing discourseamong students makes their ideas appear to beas important as each other’s and the instructor’scomments” (Slatin, 1992, cited in Jonassen, 1999,p. 229).

Testing Design Features in aTechnology-Supported,Knowledge-Building Environment

Our desire in this study was to integrate con-structivist principles with the concept of learn-ing communities in such a way as to yielddesign features that could be implemented andtested empirically. CSILE is unquestionably themost prominent example of a technology-sup-ported learning community, but it has beenemployed almost entirely with children work-ing on knowledge building over year-long spansof time. Indeed, virtually all of the research onlearning communities that we are aware of hasbeen conducted with school-age children (e.g.,the collaboratory notebook—Edelson, Pea, &Gomez, 1996; the knowledge integration en-vironment—Bell, Davis, & Linn, 1995). In addi-tion, students using CSILE work on general,knowledge-building goals, creating theories andbuilding explanations. Finally, until the adventof Web CSILE, the knowledge-building com-munity was not networked to resources outsidethe classroom.

Construe©, Web software designed specifi-cally for the purpose of supporting aknowledge-building learning community, wasdeveloped in 1996 for an application in highereducation (Lebow et al., 1996), but it was not em-pirically validated. It therefore offered an oppor-tunity to extend the findings on CSILE into thehigher education arena with adult students. Asschool children must become knowledgeworkers to function in today’s society, so muststudents in higher education, since most of themwere educated under more traditional models of


COLLABORATIVE KNOWLEDGE BUILDING 61

schooling. Differing from CSILE, Construe isdesigned to manage a one-semester course, in-corporates a database of articles that learnerscan easily access from the course Web site, andoffers a means for students and the instructor tobuild links within the environment to resourcesoutside of it. As an integration tool (Hannafin etal., 1999), it supports the ability of students toreact to information, integrate new knowledgewith personal knowledge, and view multipleperspectives. Construe also provides the openaccess to shared information that Jonassen(1999) indicated is critical in constructivist learn-ing environments.

While Construe constituted the softwareaspect of a technology-supported learning com-munity, we wanted to explore several construc-tivist instructional strategies as learningconditions to facilitate a community of learners.For instance, Wenger (1998) stated that “a sourceof community coherence is the negotiation of ajoint enterprise” (p. 77). Combining that notionwith the emphasis in constructivist learning en-vironments on authentic tasks, we hypothesizedthat incorporating a collective and authenticcommunity goal could be an effective instruc-tional condition to facilitate the development ofa learning community. A community projectwould be the joint enterprise within which bothgroups and individuals could negotiate morespecific learning goals and tasks. The coopera-tive groups were also expected to promote mul-tiple perspectives along with social negotiationof meaning. Finally, we wanted to encourageownership in learning as a learning conditionthat may facilitate self-regulation in learners(Duffy & Cunningham, 1996; Lebow, 1995). Toaccomplish this, we incorporated an instruction-al strategy enabling learners to select coursereadings for personal interest and to meet thegoals of their learning teams.

Purpose of the Study

The purpose of this study was therefore to inves-tigate two research questions, one primary andone secondary. The primary question was: Howeffectively did the instructional conditions im-plemented within a graduate course dealing

with alternate views of teaching and learning(known as Altviews) contribute to a collabora-tive knowledge-building enterprise? That is,what effects, if any, did a collective goal,cooperative groups, ownership in learning, andtechnology tools have on student acquisitionand use of key concepts, self-regulation, andhigher-order thinking? The secondary researchquestion was: What instructional managementissues arose at an individual, group, and com-munity level and how were these issuesresolved?

METHOD

The design of this research was a single casestudy focused on identifying and understandingthe learning conditions that promote knowledgebuilding in collaborative learning communities.According to Yin (1989), the case study as a re-search strategy is preferred for investigating howand why questions regarding a contemporaryphenomenon occurring in a real-life context (p.13). Once the case study’s questions have beenidentified, the case is defined and sources ofevidence from the case are determined that bearon the questions being asked.

The case was the Altviews course composedof 20 graduate students at Florida State Univer-sity. The subunits of the case were individuals,groups, and the community as a whole. Datawere collected over the span of an entire 15-week semester from students enrolled in thecourse, the course instructor, the course teachingassistant (who was also the primary researcher),and a student informant (identified at the begin-ning of the term and asked to take on the role ofparticipant observer).

Learning Environment

The purposes of the Altviews course includedintroducing learners to the goals, methods, tech-niques, and theory base of a variety of recentand emerging models of learning, instruction,and instructional design. Because the course in-structor wanted learners to experience, as muchas possible, the theories and models they were


62 ETR&D, Vol. 50, No. 1

learning about, the course offered an ideal con-text for this research. Therefore, four instruction-al conditions were designed into the course,each with a specific function as indicated by theliterature on learning communities:

1. Having a collective and authentic communitygoal to support collaboration and engagement in thecommunity. Throughout the semester, the classwas engaged as a learning community in thedesign of a charter school whose guiding prin-ciples were based on the knowledge base (ar-ticles) of the Altviews course and outsideresources added by class members. This projectwas designed by the researcher and instructor asa vehicle to promote collaboration andknowledge building. It was authentic for the stu-dents enrolled in the Altviews course in thesense that charter schools have evidenced agreat deal of interest in Florida education,generally, and the local area, specifically.Debates are common in the students’ academicprograms regarding how instructional designprinciples might inform public education, andthe project offered a concrete opportunity to putthese ideas into practice.

2. Using cooperative groups to provide for socialnegotiation and to promote multiple perspectives.

Six learning groups were formed (four of thegroups were composed of three students, andtwo—Planning, and Apprenticeship—had fourstudents) to discuss course readings, cooperateon developing the community project, and cre-ate a concept map. To accomplish the courseproject, class members negotiated tasks relatedto (a) Planning, (b) Learning Objectives and Cur-ricula A and B, (c) Learning Environment, (d)Community Integration, and (e) Apprenticeshipand Extra Curricular Activities. These becamethe basis for the six learning groups (two groupstackled Learning Objectives and Curricula). Par-ticipants indicated their preferences for workingon a particular set of tasks, and the instructorused this information to assign them to learninggroups. These were the chosen roles of the learn-ing groups to achieve the collective goal ofdeveloping a charter school. Additional groupresponsibilities included the development of aconcept map using Inspiration™ software to pic-

torially represent the relations among diverseconcepts taken from the readings.

3. Personally selecting course readings and groupdecision-making to encourage ownership and respon-sibility in learning. Based on the tasksnegotiated within each learning group, in-dividual members of the group selected read-ings from a database of articles and soughtoutside resources of interest to them and theirgroup. The articles included recent publicationson topics such as constructivist theory, coopera-tive learning, problem-based learning, technol-ogy as tools, assessment from a constructivistperspective, and learning communities, amongothers. The articles also represented both em-pirical research and conceptual or theoreticaldiscussions. Students wrote reactions to the ar-ticles they read, selected and defined key con-cepts, and abstracted excerpts from the articlesto illustrate defined concepts. Individuals alsowrote confidential reports to the instructor twiceduring the semester reporting on the quality oftheir learning experiences. Finally, individualssubmitted final course reflections evaluating theeffectiveness of the instructional strategies incor-porated into the course.

4. Employing technology tools to archive the work ofthe community and facilitate communication. T h eprimary tool used to promote the collaborativeeffort of the community, Construe, was original-ly designed and developed during the Springsemester, 1996, for a distance learning class. It issoftware developed collaboratively by FloridaState University and N.C.R. Corporation for thepurpose of facilitating the development andmanagement of learning environments over theWorld Wide Web (WWW). A key principle inthe design of Construe is that the artifacts or ob-jects produced by the community have value.The knowledge objects that members create andappropriate from the outside world becomepublic materials supporting the goals of thecommunity and its individual members. The ac-cumulating database produced by the par-ticipants is evidence of the community’sadvancing knowledge (Lebow et al., 1996).Thus, individuals and groups “published” all oftheir work on the course’s Construe Web site,



which also contained the article database. In ad-dition, from within the site, class members couldsend e-mail messages to another individual, aselected group, or the entire class.

Collectively, these instructional conditionswere expected to contribute to the effectivenessof the learning community through enhancingacquisition and use of key concepts, promotingself-regulation, and facilitating knowledgebuilding.

Data Collection

Ten different data sources were used to collectevidence for this investigation. Some of the datasources were expected to bear on more than oneresearch question, as shown in Table 1. In thissection, the data sources are described. Eachdata source is presented along with the means

by which data were collected from that sourceand the information each source was expected toyield.

Figure 1 displays what happened in theAltviews course over the 15-week semester. Theinstructional conditions that were the focus ofthe study had been incorporated into the designof the course prior to the beginning of thesemester. Thus, during the semester, the courseproceeded according to its normal calendar, theonly addition being the background survey ad-ministered in the first week. Artifacts of thecourse served as the primary sources of data.

Background survey

Questions on participants’ educational back-ground, familiarity with the technology used inthe course, and professional goals and experien-ces comprised the Background Survey, which

Table 1 Relevant data sources for the two research questions

Partici- FinalBack- Confi- Concept Teaching Instruc- pant Course

Data ground dential Reading Definition Concept Assistant tor Observer Reflec- FinalSources Survey Reports Reactions & Excerpt Maps Journal Journal Journal tions Project

How effective were the instructional conditions in promoting knowledge building?

Anticipated ResultsAcquisition × × × × × × & use of key conceptsStudent × × × × × self-regulationHigher- × order thinking

Instructional ConditionsCollective × × × × × × purposeCooperative × × × × × × × groupsOwner- × × × × × × ship in learningTools × × × × ×

What management issues arose as the instructional conditions were implemented and how were they resolved?

Manage- × × × × × ment issues


64 ETR&D, Vol. 50, No. 1

included the voluntary agreement of class mem-bers to participate in the study. The backgroundsurvey also provided information on theparticipant’s name, degree program or major,level of computer expertise, and experience orinterest in educational reform. It was expectedthat initial differences among participants mightinfluence their reaction to the design features in-corporated into the course.

Reading reactions

Reading reactions consisted of essays writtenand posted to the course Web site by students inresponse to the articles they selected and read.Students were instructed to react not only to areading but also to any reactions already postedby other participants. On a Web form underheadings for each, students entered their “gut

reactions,” “big ideas,” “implications for teach-ing and learning,” and “nagging questions.”Class members posted reading reactions at theirown rate, but they were encouraged to write andpost at least one each week.

The purpose of the reading reactions was topromote generation of knowledge rather thanreplication of knowledge. The quantity andquality of the participants’ reading choices andassociated reactions reflected their contributionsto the knowledge base and the knowledge-building effort of the community. The readingreactions were downloaded from the Web siteand expected to include appropriate use of theconcepts associated with the individual’s role inthe collective goal of the community, personalattitudes and values concerning learning and in-struction, and critical evaluation of models andtheories studied throughout the semester.

Week 1

Participants complete background survey

. . . . Weeks 2–14 › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › ›

Participants:—Select readings—Write and post reading reactions on course Web site—Select & define concepts on course Web site

. . . . Weeks 2–4 › › › › › › › › ›

Class defines parameters of charter school project and learning teams are assigned.

. . . . Weeks 4–14 › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › › ›

Learning teams:—Work on their part of class project—Construct concept maps

. . . . Week 4 . . . . Week 8 . . . . Week 13

Participants submit confidential reports to the instructor via e-mail.

. . . . Week 15

Teams submit—Final projects—Final concept mapsParticipants submit—Final course reflections

Figure 1 Course activities and data collection in Altviews over the 15-week semester



Confidential reports

Twice during the semester, students submittedconfidential reports, in which they reflected ontheir learning and course process. Students e-mailed their confidential reports directly to theinstructor, who shared them with the class onlywith approval of the student submitting thereport. The purpose of the confidential reportswas to provide information on the overall qualityof the learning experience, how well each learn-ing group was functioning, perceptions about theprocess of the community project, and any sug-gestions for change or improvement. This datasource provided information on individual,group, and community issues, and evidence forhow well the individual was adapting to the in-structional conditions related to the learningcommunity. A file was created for each learninggroup, with the confidential reports andinstructor’s comments in response to the reportsarranged in sequence by date.

Concept definitions and excerpts

The purpose of this data source was to provideinformation on how well the participants under-stood the articles and the associated conceptsderived from the articles. The Web site waspopulated with 20 key concepts prior to thebeginning of the course. These initial 20 conceptsformed the foundation of the knowledge base.Concepts were added by the participantsthroughout the semester based on the readingsthey chose and their role in accomplishing thecollective goal of the community. Participantsgenerated concept definitions and linked ex-cerpts from the readings based on their under-standing of the article; they were not to quote adefinition directly from an article or repeat apreviously entered definition. For example, onestudent added the concept abduction to theknowledge base. She described abduction as“the process of making connections from onecontext to another.” Her excerpt included thisexplanation:

While biases and preferences often impeded under-standing, these mental idiosyncrasies may also serveas powerful, if quirky and personal, guides to solu-tions. C.S. Peirce coined a word to express the imagina-

tive mode of reasoning involved in such mental leap-ing.

All the concepts were archived and downloadedfrom the Web site by learning group.

Concept maps

In their learning groups, participants generatedconcept maps using Inspiration™ software. Themap’s organization and the concepts selected bythe participants were directly associated withtheir chosen role in achieving the collective goal.Learning groups submitted concept maps inhard copy and in electronic form, either on diskor posted to the Web site, at the end of thesemester. The purpose of concept maps was toprovide information on each group’s under-standing of relations among the diverse con-cepts. The maps were dynamic, and changedover the semester relative to each group’s role inaccomplishing the collective purpose. The finalversions of the concept maps were printed forexamination and comparison.

Final project reports

Learning teams wrote final project reports at theend of the semester describing their team’s con-tribution to the collective goal (development of acharter school), which they published on thecourse Web site. All records were archived inand downloaded from the Web site. The pur-pose of this data source was to provide informa-tion on individual and group effort forachieving the collective purpose of the com-munity. The reports were expected to show adirect connection to the concept maps and thecollective purpose of the community.

Journals of the teaching assistant-researcher,instructor, and student informant

Throughout the semester, the teaching assistant-researcher, instructor, and student informantkept journals of their reflections on class ac-tivities, administrative issues, and the coursestructure in general. These journals alsoprovided information on the minutes of meet-ings from the planning stages of the course to


66 ETR&D, Vol. 50, No. 1

the end of the semester. These meetings amongthe instructor, teaching assistant-researcher, andparticipant observer were held on a weeklybasis. The notes summarized the wide range ofissues that arose from week to week and werekept in a loose-leaf notebook and then trans-ferred to a word processing file.

Participants’ final course reflections

Participants were asked to respond to a set ofquestions asking for specific feedback on aspectsof the course. Participant reflections weresolicited via a set of questions e-mailed to themat the end of the semester. The purpose of thisdata source was for the student to provide acritical evaluation of the theoretical and practicalconsiderations of learning, instruction, and in-structional design models designed into theAltviews course. Participants reflected on courseexperiences and evaluated the effectiveness ofthe teaching and learning models used. Specificquestions focused on the nature and degree ofindividual and group collaboration relative tothe instructional conditions of the course. Thisdata source provided information onparticipants’ personal attitudes and values con-cerning learning and instruction, and criticalevaluation of models and theories studiedthroughout the semester. The final course reflec-tions were retrieved in electronic form from theinstructor.

Data Analysis

The first step in data analysis was to transfer toword processing files all of the data archived onthe course Web site or sent electronically to theinstructor. Electronic files were also created forthe data contained on the background survey;the journals kept by the instructor, teaching as-sistant-researcher, and student informant; con-fidential reports; and the final course reflections.Once electronic files existed for every datasource, the data relating to each source had to beaggregated, summarized, and in some cases,coded.

For instance, information from the back-ground surveys was summarized by individual,

resulting in a profile for each participant.Responses to the confidential reports and finalcourse reflection questions were also sum-marized by individual. Concept definitions andexcerpts were summarized by learning group,with the researcher recording how many con-cepts defined by the group came from theoriginal 20 included in the Construe database,how many came from articles in the databasebut not included in the original concept list, andhow many came from articles and resources out-side the course Web site altogether. Conceptdefinitions were also examined for accuracy.Concept maps were considered in relation to theconcepts identified and defined by each group.Individual links between concepts and the over-all structure of the map were examined to iden-tify patterns or gaps in connections amongconcepts and their connection to the finalproject.

Then reading reactions, confidential reports,and journals were coded following the samebasic procedure. First, each document was readand codes relevant to the research questionswere identified and marked on a printed copy.For instance, the first question concerned howeffectively the instructional conditions designedinto Altviews contributed to the knowledge-building enterprise of the community. The read-ing reactions were designed to serve as ascaffold to promote knowledge building andhigher-order thinking associated with the theorybase supporting alternate views of teaching andlearning. It was also expected that participantswould read and refer to the reading reactions ofothers as they constructed their own responsesto a particular article. Thus, codes emerged fromthe reading reactions data pertaining tostudents’ reflection on what they read, their refer-encing other resources, and evidence of higher-order thinking (e.g., combining ideas from severalreadings to offer a new perspective). Withineach of these three categories, codes were fur-ther refined based on what the data revealed.Table 2 displays the final set of codes and codedefinitions that resulted from the analysis ofreading reactions.

To determine the reliability of coding, the re-searcher asked the instructor to use the codingscheme that emerged in the analysis to code a



reading reaction she had coded previously.There was almost complete agreement betweenthe two; the few areas of disagreement were dis-cussed and the codes modified until both partieswere satisfied with the result.Thereafter, the re-searcher asked the instructor to review codesand examples from the data that illustratedthose codes every time a new code became ap-parent.

When all the individual data sources hadbeen coded and recorded, the sheer amount ofdata precluded an in-depth examination of allclass members. Instead, the results of six par-ticipants were investigated in depth through the

use of a time-ordered metamatrix (Miles &Huberman, 1994) that displayed the sources ofdata over a 15-week semester for these sixpeople. The six were chosen to reflect the diver-sity of students in the course with respect tocompletion of assignments and satisfaction withcourse goals and structure. That is, two com-pleted all assignments and expressed high satis-faction with the course, two completed most ofthe assignments and expressed moderate satis-faction with the course, and two completed onlya few of the assignments and expressed lowsatisfaction with the course.

From the time-ordered metamatrix, patternsin the data were sought to answer the primaryand secondary research questions. Thus, datasources were examined first for evidence of theeffects the instructional conditions may havehad on acquisition and use of key concepts, self-regulation, and higher-order thinking. Then, is-sues related to course management wereidentified and evidence sought in the data sour-ces concerning their resolution.

For example, with regard to the primaryquestion, an issue related to the cooperativegroups arose in several of the data sources in-cluding confidential reports and final coursereflections. Specifically, comments referred tohow a group’s functioning may have affectedthe students’ knowledge building (individualand community-wide), self-regulation, and theachievement of the collective goal. One studentwas not completing his work on time and felt hewas not supporting either his group or the com-munity. Failing to do the reading reactions in atimely way affected his group’s ability to com-plete work on the concept maps and thecommunity’s ability to use what he was learningto advance its goals . The recurrence of thesetypes of comments indicated the development ofa pattern.

RESULTS AND DISCUSSION

The purpose of this study was to implement andevaluate the effectiveness of specific instruction-al conditions designed to facilitate knowledgebuilding in a community of learners. This sec-tion begins with a brief profile of the par-

Table 2 Reading reaction categories,codes, and definitions

Reflection Category

Surface Reflection (SR)—ties into personalexperience (relevancy) without offering any newperspectiveDeep Reflection (DR)—ties newly acquiredinformation from the readings into previousexperience and restructures based on individualperspective

Referencing Category

Reference Participant (RP)—references anotherparticipant (recognizes value of the work of thecommunity)Reference Participant Plus (RP+)—referencesanother participant and adds another perspectiveReference Author (RA)—references anotherauthor (recognizes the value of the readings)Reference Author Plus (RA+)—references anotherauthor and adds another perspectiveReference Course (RC)—references course goalsor structure (recognizes value and relevancy ofcourse goals or structure)Reference Course Plus (RC+)—references coursegoals or structure and adds another perspective

Higher-Order–Thinking Category

Knowledge Synthesis (KS)—combines ideastogether from the readings and restructures newinformation to provide a different perspectiveCritical Analysis (CA)—critically analyzes author,article, or participant; agrees or disagrees andprovides rationale, identifies knowledge gap,offers suggestion, opinion, or new approach tothe problem


68 ETR&D, Vol. 50, No. 1

ticipants from whom data were investigated indetail and then presents the effects of the in-structional conditions that were incorporatedinto Altviews. The effectiveness of the instruc-tional conditions and how they supported col-laborative knowledge building as well asclassroom management issues are discussed interms of the key constructivist principles imple-mented in the Altviews learning environment.We chose this organizational structure to discussthe results because our focus is on how the condi-tions worked, not just whether they succeededin having an effect. Although we use these prin-ciples as a guiding framework for presentingand discussing specific results, our findingsreflect synergistic effects that are hard to at-tribute to one or another learning conditionalone. This recalls Bell and Winn’s (2000) argu-ment that successful classroom interventions in-volving technology tools result in outcomes thatare “systemic effects of the package, not of par-ticular components in the intervention” (p. 141).Finally, the section closes with a brief discussionof implications for instructional design and sug-gestions for design guidelines.

Participants

The class comprised 20 master’s and doctoralstudents. Half of the class was in the Instruction-al Design program. The remaining students in-cluded majors in Art and Social StudiesEducation, and Human Sciences. The computerexpertise varied, but all students had basic com-puter skills. This was important as we did notwant the lack of computer skills to interfere withthe quality of learning. Six participants werechosen for in-depth investigation of their data.All names are pseudonyms.

Warren and Mary completed all assignmentson time and expressed a high level of satisfac-tion with the course in their final course reflec-tions. Warren is a graduate student incommunication research and a high-level com-puter user, based on his familiarity with IBMand Mac platforms, e-mail, WWW, and a varietyof other software programs. Warren hopes todesign Web-based education programs upongraduation. Mary is a doctoral student inelementary education. She is a medium com-

puter user and hopes to be involved withteacher training at the university level.

Two students, Chris and Carol, completedmost of their assignments on time and expressedboth positive and negative comments about thecourse. Chris is a graduate student in instruc-tional systems and a high-level computer user.He hopes to apply instructional design conceptsand theories to performance problems in highereducation. Carol is a master’s student in instruc-tional systems and a medium-level computeruser. She has some familiarity with PC plat-forms and limited experience with e-mail andthe WWW.

Finally, Matt and Julie were two studentswho did not complete all their assignments andhad mostly negative comments about their satis-faction with the course. Matt is a doctoral stu-dent in instructional systems and a high-levelcomputer user. His professional goals includeworking in a corporate environment. Julie is amaster’s student in instructional systems and ahigh-level computer user. Julie hopes to developmultimedia programs for corporate environ-ments upon graduation.

Space considerations preclude the inclusionof comments from all six participants for each ofthe constructivist principles discussed in the fol-lowing sections. Comments were selectedprimarily from Warren, Mary, Matt, and Juliebecause they provided the most meaningful in-sights into the Altviews learning environment.

Collective and Authentic Community Goal

According to Wenger (1998), a knowledge com-munity must have three interacting dimensions:(a) mutual engagement, (b) a joint enterprise,and (c) a shared repertoire. In this study, a col-lective and authentic community goal was usedas a means of implementing the three dimen-sions. The problem, creation of a charter school,was expected to provide the joint enterprise thatwould mutually engage participants. Theshared repertoire was the Construe, Web site,which contained some resources and providedthe structure for participants to add informationthat would support community efforts.



The final project reports constituted thecommunity’s attempt toward achievement ofthe collective goal, that is, development of thecharter school (which the class named SynergyCenter). Producing the reports was the last ac-tivity accomplished in Altviews, and it broughttogether all the readings, concepts, group work,and class discussions. Class members, togetherwith the instructor and teaching assistant-re-searcher, negotiated six teams based on func-tional roles the class identified as necessary forachieving the goal. The teams thus became (a)Planning, (b) and (c) Learning Objectives Cur-ricula A & B (two teams), (d) Learning Environ-ment, (e) Community Integration, and (f)Apprenticeships. Once the teams were in place,individuals began selecting readings based ontheir team’s role, and each group designed aconcept map that was used to support its role inaccomplishing the collective goal.

For instance, the Planning team went outsidethe articles database to incorporate Kaufman’sorganizational elements model (Kaufman &Grise, 1995) as a needs assessment tool to iden-tify gaps between what is and what should beconcerning Synergy Center. Likewise, the Com-munity Integration team decided to emphasizethe life-long learner as the focus of its concepts.In the final project report, that team wrote, “aSynergy Center learner will develop into anadult learner, who in turn is a life-long learner.”The primary concepts identified to support thisfocus were theories of teaching and learning,problem-based learning, authentic achievement,and knowledge-building community. Thisteam’s concept map also included detailed keyideas, such as cooperative learning, novice to ex-pert transition, modeling, authentic assessment,and mentoring. For each of the concepts on themap, the team provided a definition, related theconcept to prominent aspects of the communitygoal as well as its own goal of producing a life-long learner, suggested an evaluation strategyregarding the concept, and included links to re-lated Web sites.

Evidence from the team concept maps indi-cated that all groups used key concepts that theyappropriated from sources outside their teams.In addition, it was apparent from Construe’stracking of who posted concept definitions that

all participants, rather than only a few in-dividuals in the community, contributed to thiscollaborative effort.

Although community members had a jointenterprise in achieving the collective goal, and ashared repertoire in the course Web site, it is notso apparent that the collective goal encouragedmutual engagement (Wenger, 1998). Thedevelopment of the charter school required in-tense interaction across groups to share ideasand strategies, and ensure a cohesive approachto this problem-solving exercise. The inter-change among groups was weak, however, withalmost no exchange of ideas outside of class.One possible reason for this was Construe’scommunicative function, which enabled classmembers to send e-mail messages only to thecommunity, a team, or an individual. Warrensaid, for example, that he rarely contacted anyother participants even though he wanted to be-cause Construe did not support quick inter-change of ideas in the same way a bulletin boardwould.

Participants repeatedly recognized, in theirconfidential reports and in class, that the classwas not functioning so much like a communityas a group of teams. Julie offered this suggestionto improve the quality of cross-group inter-change: “Some mechanism [should] be put inplace to allow more exchange of information be-tween functional groups and tasks.” Late in thesemester, a “jigsaw” routine was implementedduring class in an attempt to create more mutualengagement among members of the community.Teams met outside of class to determine whatprogress they wanted to communicate to otherteams. Then, during class, teams regrouped injigsaw fashion to share their reports with mem-bers of other teams. Unfortunately, this occurredtoo late in the semester to solve the problem ofmutual engagement and was more of an after-the-fact, “Band-Aid” solution. Communityinterchange is critical for collaborative knowl-edge-building communities; provisions shouldhave been made from the beginning to ensurethis effort was supported. As useful as Construewas for the individuals within a team, it fellshort as a community communication tool. Ac-cess to a listserve, chat room, or bulletin boardfor posting issues related to course assignments,


70 ETR&D, Vol. 50, No. 1

ideas, technology problems, and other items ofinterest might have promoted greater interac-tion across groups.

Finally, it appears as if the charter schooldevelopment was not as authentic a task orproblem as we originally thought. That is, it didnot bring the community together as a coherentwhole for all groups but instead appealed tosome groups and individuals more than others.Julie stated that the collective goal “. . . was notas effective as we hoped. Once we split into ourrespective groups, the collective purpose waslost because most were only concerned aboutthe issues directly affecting their own functionalarea.” Another student reported in a confiden-tial report that she had been excited at first bythe idea of developing a charter school. But themore she and her team got into it, the less inter-ested she became, primarily because she did notview her future work as an instructional desig-ner to be related in any way to schools. Thus, al-though we had a collective goal, we did not havea shared community vision in which all groupsworked together in a collective way. For some,this whole-class problem-solving activity turnedinto busy work, and just another project to com-plete before the semester ended.

Cooperative Groups and Collaborative Activity

Discourse is the heart of collaborative knowl-edge-building communities (Scardamalia &Bereiter, 1994). The use of cooperative groups inAltviews was designed to support knowledge-building discourse and facilitate collaborativeactivity. The students worked at an individuallevel reading and reacting to articles and iden-tifying key concepts. Then, individuals broughttheir personal understandings of the readingsand concepts to the team where the conceptmapping activity prompted them to confrontand resolve multiple perspectives and negotiatea common understanding with respect to theteam’s role in accomplishing the collective goal.

Participant responses to the final coursereflections provided insight into their percep-tions of collaboration. In several questions, stu-dents were asked to comment on the extent to

which they used other people’s work, either inwriting their own reading reactions or bringingideas to the team. Warren commented that col-laborative effort “depended on the time ele-ment.” Reading reactions were time intensive,because students were instructed to read all thereactions and consider them before posting theirown. Some students reported that they foundthe reading reactions beneficial and read all pre-vious reactions before posting their own. Thiswas supported by the analysis of reading reac-tions, which revealed a substantial number ofreferences to the work of others. However, someof the students felt that “reacting to the reac-tions” resulted in repetition of effort. After threeor four responses to an article, the main ideasand concepts are usually identified. We couldhave restricted the number of responses, but thiswould have limited the reading selections forthe community and the level of self-direction wewere trying to encourage, a point which will berevisited in the next section.

Warren completed all the reading reactionseven though he had a large gap of inactivity inthe middle of the semester. His first confidentialreport indicated satisfaction with the learningexperience. “I find the interactions excellent. Theideas and concepts are first-rate . . . I get a sensewe are forming a cohesive group of like-mindedindividuals embarking on a shared goal.” Mid-way through the semester, however, Warren ex-perienced a personal crisis that prevented himfrom completing course assignments onschedule, which in turn affected the collabora-tion in his team and the community. He ex-pressed concern about his inability to completeclass assignments on time and reaffirmed hisresponsibility in an e-mail message to the entireclass.

When I agreed to be part of this community oflearners, I accepted the responsibility to contribute tothe knowledge base. I have felt for sometime that I wasnot and am not contributing my fair share. I want to letall of you know that I take this opportunity seriouslyand am reaffirming my commitment.

After this note to the community, Warren’sfinal report indicated a lack of leadership on histeam, “. . . no leader has emerged . . . causing adrift ... we haven’t fused into a working unit.”



His final course reflection indicated the coursestructure supported “isolationist activities” bet-ter than it did collaborative activity:

I was able to access the readings very easily and thereactions of my fellow peers. I even liked (although itscared me) the listings of my accomplishments . . . Theweb-structure allowed me to access the course evenwhen I wasn’t in town . . . How liberating . . . it waswonderful.

But, again in Warren’s words, “As well asConstrue supported me working alone, it didnot support me working in a group.”

Like Warren, Mary had a positive view of herlearning experience early in the semester. In herfirst confidential report, she expressed positivecomments for her overall learning experience. “Iam thoroughly enjoying this course . . . I amlearning and exploring new material and ideas. . . and exposed to a new model of teaching. Iam very excited about both of these aspects ofthe course.” By the end of the semester, how-ever, Mary’s experience was deteriorating: “. . .my individual experience had declined . . . dueto the fact that most of my effort in the last fewweeks has been expended on the group project.”These comments are indicative of some of theproblems Mary’s group experienced thatdegraded her learning experience.

Mary wrote a letter to the instructor express-ing her concerns about her group. One of hergroup members had a life-threatening illnessand was missing group meetings. “I’m con-cerned because of his illness . . . because he ishaving surgery.” Another group membermissed a class and a group meeting, and Marystated, “I’m concerned that our group is notworking well together . . . I’m disappointed inour group’s effort.”

Mary’s group had worked out a plan to in-clude a peer review of each other’s work, put allthe pieces together, and then do a final reviewon the whole product. However, this was not ac-complished in a timely manner, and Mary’sreaction was that “I feel the sections are not theresult of group collaborative effort . . . ratherthey are individual products.”

Warren and Mary appeared to experience thefeeling of being part of a larger group, and theyunderstood the quality of their learning ex-

perience was dependent on the work of others.They also accepted their own responsibilities tothe community. Other students seemed to stayat the group level and were more concernedwith their individual learning than that of thecommunity. Those students who failed to com-plete their reactions in a timely manner anddidn’t feel empowered to regulate their ownlearning were somewhat lost in a self-directedlearning environment. It appears as if the levelof self-regulation was limited to only those stu-dents who were engaged at a community leveland felt part of a larger group.

Matt, for instance, described Construe as aconstraint and stated in his second confidentialreport, “The internet-based aspect of the classmade me feel alienated from other class mem-bers. Writing does not come natural like speak-ing . . . posting articles cannot be a substitute fororal communication.” Matt also wanted a dif-ferent teaching strategy: “I would rather havelectures by you” (the instructor). Matt failed tocomplete his reading reactions and preferred to“look at mostly my group’s work” or “discussverbally because talking is natural.”

Julie also failed to complete her reading reac-tions and to engage at the community level. Sheconsidered others’ work “very little.” She alsofelt that “the different groups are working instove pipes . . . besides the last few meetingsthere really is not valuable exchange of informa-tion . . . there should be more interaction acrossgroups.”

It appears our learners might have benefitedfrom a more traditional approach first, followedby more gradual introduction of the construc-tivist strategies or additional scaffolding to sup-port these strategies.

Carol, for instance, explicitly requested morefeedback.

I think this class was unique in its combination of in-dividual and group tasks . . . this has a tendency to in-crease anxiety about grades, as things are beyondone’s control . . . this was compounded by the lowlevel of feedback received to the work posted.

And it is clear from Matt’s comments that hewould have preferred more direct instruction, “Iwould have rather had lectures” and “neededmore instructional support.”


72 ETR&D, Vol. 50, No. 1

Course Readings and Ownership inLearning

Brown and Campione (1990) wrote that under-standing is more likely to occur when studentsare required to “explain, elaborate, or defend”their position to others. The participants inAltviews were asked to read theory-based ar-ticles in the field of alternate views of teachingand learning and to complete a series of ques-tions for each of the readings. By answeringthese questions, participants contributed theirown explanations to the articles and reacted tothe ideas of others previously posted in responseto the same articles. The assignment requiredlearners to engage in a type of self-evaluation oftheir own personal position relative to that ofothers, to reevaluate their own position, and toelaborate the information in new ways. Thereading reactions were designed to serve as ascaffold to promote knowledge building andhigher-order thinking relative to the theory basesupporting alternate views of teaching andlearning. The substantive quality of the readingreactions showed evidence that new levels ofunderstanding were reached by the participantsin the Altviews class for those students whocompleted the reading reaction assignment.

Specifically, analysis of the reading reactionsresulted in three categories (reflection, refer-encing, and higher-order thinking). The reflec-tion category was further elaborated in terms ofsurface and deep reflections. A surface reflectionwas one that related a personal experience in thestudent’s response to an article showing somelevel of relevancy in the reading. A deep reflec-tion related the student’s personal experienceand included some restructuring of informationbased on personal feelings and beliefs. Referback to Table 2 for the categories, codes, anddefinitions that emerged from the analysis Anexample of a deep reflection appears in this reac-tion to the Slavin (1991) article from one of theinternational students:

In the Fall of 1993, I had the privilege of taking a re-search survey course which incorporated the use of thecooperative learning approach, which was new to meat that time. It was one of the most effective experien-ces in my life as a student. The cooperative learningprocess really impressed me, and I was convinced that

I should use it as an art educator in the near future. Ieven asked myself why such an approach was notused in Kuwait? . . . Our ultimate aim was to get thejob done together, and we did. I am still proud of thatwork, and still in touch with the members of mygroup. Cooperative learning is a method for improv-ing the achievement of learning and other educationaloutcomes, and implementing such an approachshould be a necessity in educational settings. This canbe accomplished if we can succeed in creating a senseof commonality, and overcome the lack of communica-tion among these groups.

The higher-order–thinking category includedreading reactions that demonstrated knowledgesynthesis and critical analysis of the articles. Anexample of a critical analysis to the Drucker(1994) article included the following response:

My initial reaction to this article is that it is extremelybiased towards business interests and rather cavalierlyoverlooks some tremendous social problems causedby the industrial revolution. It seeks to present thatrevolution as one that proceeded with a minimum offriction and upheavals without delving into the his-torical reasons behind such a revolution, i.e. the warmachines and imperialist machinations it typicallysupported. The author also dismissed the violence ofthis century re: Hitler, Stalin, Mao as signifying noth-ing. This guy needs to go back to school and take somereal history courses.

The transformation of the social structure by the in-dustrial revolution will be dwarfed by the transforma-tion experienced by the information revolution. Iwould have to agree in some respects, but again, thisauthor ignores the fact that this “information revolu-tion” is just another form of imperialism—whoeverholds the keys to the information is “the winner.” Hetalks in terms of winning and losing—competitivenessmay be everything in terms of business, but the idea oftranslating this to educational theory or ideas is reallynegative. What about sharing and cooperation?

Knowledge building is the creation of ar-tifacts that capture the student perspective in theform of explanations, theories, and solutions(Bereiter, 1995). The quality of the reading reac-tions showed evidence of knowledge buildingfor all of the participants in Altviews. That is, allstudents had instances of coding in at least oneof the three categories with most participants’reactions falling into all three categories. Thehigh quality of the reading reactions producedby individuals added new perspectives, ex-planations, and solutions to the accumulating



knowledge base showing evidence of the exper-tise of the community.

The reading reaction questions and iden-tification of concepts or key ideas from the read-ings were intentionally designed as scaffolds forpromoting higher-order thinking and, ultimate-ly, knowledge building. The extent to whichthese scaffolds were effective directly con-tributed to the collaborative knowledge-build-ing efforts of the community.

A strategy related to the course readings wasalso designed to promote ownership and self-direction in learning, and that was the oppor-tunity for students to select the articles theywished to read throughout the semester. Wehoped that students would be empowered totake personal responsibility and accountabilityfor their own learning by selecting articles of in-terest and identifying associated key ideas, aswell as assessing their own learning processthrough the confidential reports, fulfilling theirgroup’s role in accomplishing the collectivegoal, and participating in the democratic opera-tion of the classroom environment. In hersecond confidential report, Mary expressedsatisfaction with this aspect of the course. “I par-ticularly enjoy the freedom and responsibilitythat allowing students to choose their own read-ings affords.”

Article selection was based on tasks,negotiated in each learning group, associatedwith the group’s role in developing the charterschool. In addition to the readings, outsideresources were added to the communitydatabase to assist with the overall effort of thecommunity in accomplishing the collective goal.Group decision making was necessary in theselection and definition of the concepts and as-sociated excerpts. This agreement was necessaryto create the concept maps and select the read-ings to support the accomplishment of the col-lective goal.

Even though Altviews provided an in-frastructure designed to enable students to takepersonal accountability and responsibility fortheir own learning, the provision of appropriatetools and resources to support students inknowledge-building activities does not assurethat students will take advantage of the oppor-tunity to learn. Those students functioned effec-

tively in the Altviews environment who gainedcontrol of their learning process early in thesemester, demonstrated a high level of valuetoward the community and their individual ef-forts, and had a high level of satisfaction fromtheir experience in the course. Conversely, stu-dents who failed to work steadily throughoutthe semester, functioned primarily at an in-dividual level, and did not see the relevance ofcourse assignments, experienced a low qualityof learning.

Lack of motivation leads to low levels ofproductivity (Keller, 1987). Keller has suggestedthat it is important to analyze the motivationallevels of the audience and establish motivationalstrategies in order to measure the effect thatmotivation plays in the learning process. Toomuch stress or overload is destructive and doesnot stimulate growth, suggesting a need forbalance between motivation and the feelings ofoverload. Perhaps an instrument could havebeen used that measured student motivationallevel toward the course goals and structure toidentify those students who might need addi-tional assistance. A simple survey administeredat the beginning of the course asking each stu-dent to formulate personal course goals and toestimate how much time they expected todevote to the course might have been bothrevealing and useful.

Technology Tools

Leinhardt (1992) offered several core assump-tions about teaching and learning that support aconstructivist approach to designing collabora-tive knowledge-building environments. First,learning is an active process of constructingknowledge. Second, knowledge is a “cultural ar-tifact”; we create it, share it, and transform theknowledge as individuals and groups. Third,knowledge that is distributed among groupmembers or communities is an aggregate ofknowledge that is greater than the knowledge ofany individual within the community.

The overall effort of the Altviews communityappeared to be supported by the Construesoftware. This tool was used to store informa-tion in the form of reading reactions, concept


74 ETR&D, Vol. 50, No. 1

definitions and excerpts, related Web sites, andother sources of outside information, all ofwhich were used in the development of thecharter school. However, as effective as Con-strue was for storing and integrating informa-tion, it may have been deficient as a tool topromote communication at a group and com-munity level.

According to Warren, for example, the Con-strue system “should be the reinforcingmechanism” for class participation. He sug-gested that we add a chat room component toenable synchronous group discussion ratherthan asynchronous e-mail. Warren wanted im-mediate response to his issues, rather thanhaving to wait for e-mail responses and thenhaving to forward and resend to other groupmembers. “It is very labor intensive to send e-mails, then re-send, forward, then re-forwardjust to discuss an idea. The more people that areadded, the harder it becomes to circulate ideas.”

Chris requested more traditional, face-to-faceinteraction. “Construe was an asynchronousand alone environment of ideas, information,and knowledge transfer that must be offset bythe intimate, oral, and aural interchange be-tween students . . .”

Other students had similar responses. Marysaw the effectiveness of the tool, but understoodthat the level of effectiveness was directly re-lated to the quality of the input by all par-ticipants in the community. That is, the integrityof the output of a system is only as good as itsinput. Computer science has a term for this: gar-bage in, garbage out (GIGO). Because some of thestudents did not complete the reading reactionson time as Mary did, she believed the quality ofher learning experience was somewhatdiminished. “I am disappointed that not morestudents have been keeping up with theirresponses. I am somewhat angry with them forlessening the quality of my learning ex-perience.”

Most of the students thought that Construestrongly supported their learning efforts at anindividual level, but was less effective at thegroup and community levels. This may be dueto the lack of a vehicle for quick interchange ofideas and the need for all students to completetheir reading reactions in a timely manner so

that the information could be used by other par-ticipants.

Inspiration™ was the software tool used todevelop the team concept maps. This diagram-ming tool allows the creation of nodes and linksso that concepts can be spatially arranged andconnected. The intent of this tool was to promotecollaboration, discussion of multiple perspec-tives, and negotiation at the team level. Therewere no operational problems in using thesoftware; however, the extent to which the toolwas perceived to be useful varied among theparticipants. For example, Warren thought thatconcept mapping was “very effective in layingout core ideas.” He stated that he could “get asense and view of the key concepts and relation-ships to each other.” Warren also thought thatconcept mapping was very effective in achiev-ing the goals of the course and would use thisstrategy in the future.

Mary’s view of the tool was somewhat dif-ferent:

Although I see this activity fits within the 2nd coursepurpose (generating connections among diverse con-cepts), I feel that I did not really get much out of the ac-tivity. Doing it at a group level was tedious at best. Iwould have preferred to work this out individually, ornot at all.

There were several other comments similar toMary’s, all coming from participants in groupsthat experienced problems of one sort or anotherduring the semester.

It appears, then, that the tools functionedwell technologically in that students had no dif-ficulty using them for individual and group as-signments and, indeed, felt that their use of thetools enhanced the quality of their learning ex-perience. Yet use of the tools did not achieve allwe had hoped for. Certainly, the communicativefunctions of Construe bear improvement, butthe issue probably goes beyond the tools them-selves. That is, for a knowledge-building com-munity to be truly effective, there seems to be aneed for a feeling of connectedness. It takes acertain amount of time for the community todevelop that feeling and come together in a col-laborative fashion. The researcher noted severaltimes in her journal that the students asked formore time for their final projects. Most felt that



another semester could have been beneficial inachieving the goals of the community. It takestime for teams to rise above a group level and toaccomplish a shared vision of the community.The quality of the final reports indicated thatcollaborative knowledge building occurred at agroup level; however collaboration at the com-munity level was minimal and incidental ratherthan systematic. According to Senge, Kleiner,Roberts, Ross, and Smith (1994), this feeling ofconnectedness in learning organizations takesseveral years to accomplish in corporate set-tings. We may have needed more time inAltviews to develop a feeling of connectednessfor the community as a whole.

Conclusion and Implications

It appears that how much learning takes place ina learning environment based on a communityof learners approach depends to a great extenton the beliefs, expectations, and attitudes of thecommunity members. Ingrained beliefs and anexisting paradigm structure based on traditional

instructional models cause tension and result ina continual struggle on the part of the student.The approach used in Altviews did not result incommitment from all students and failed tosome extent at promoting collaborativeknowledge-building discourse at the com-munity level. Even though the class was strong-ly supported through a variety of scaffolds, andconducted in a democratic fashion, several of thestudents did not make the shift to self-directedlearning. So, what can we learn from this studythat may be useful to others interested in design-ing learning environments of this nature? Dis-played in Table 3 are design guidelines thatcomprise the lessons learned from the currentinvestigation, integrated with results of pre-vious research.

Use scaffolds

Scaffolds can be used to promote higher-orderlearning and should be developed to promoteexplanations, elaborations, and self-evaluationaround interrelated themes. The scaffolds used

Table 3 Design Guidelines for Collaborative Knowledge-Building Environments

Design Feature Guideline

Use scaffolds Use scaffolds at several levels to accommodate freedom and learning orientation; and that promote higher-order thinking and guide rather than drive the learningprocess to support students in self-regulated learning

Track the learning Use self-report techniques to measure the quality of the individual learningprocess experience and use the information to guide intervention as appropriate

Balance tension Promote productive discomfort leading to growth without overloading the student

Promote relevance Analyze motivational levels and establish strategies to measure relevance and and motivation motivation relative to class activities by evaluating confidential reports and class

discussion as well as using other instruments that might be useful for gathering this type of information

Promote a shared vision Use instructional strategies and technological tools that promote a buy-in from thestudents, across group collaboration, and quick interchange of ideas at the group and community levels

Promote the Teacher and students identify key concepts or ideas in the content domain and acquisition of use generative learning strategies and discourse to promote their acquisitionknowledge


76 ETR&D, Vol. 50, No. 1

in the Altviews environment included readingreactions, concept identification, and conceptmapping. Scaffolds should be developed atseveral levels to accommodate the freedom andlearning orientation of the students in a self-directed learning environment. It appears as ifthe more freedom provided in self-directedlearning environments, the more scaffolding re-quired to support students not oriented towardself-regulated learning. However, this supportshould guide rather than drive the learningprocess.

Track the learning process

One approach to guiding rather than driving thelearning process could be to closely track thequality of the student learning experiencethrough self-report measures, and intervenewhen necessary. Tracking the learning processthrough self-report measures helps identify is-sues before they become problems that interferewith learning. Some of these issues can be an-ticipated and provided for in advance, andothers occur during the implementation phase.A vehicle could be provided to identify andsolve these issues.

Balance tension

The expectations, values, and beliefs that stu-dents bring into the learning environmentdirectly affect the quality of their learning ex-perience. Altviews used a nontraditional ap-proach to teaching and learning that was new tothe students. This shift from a teacher to a learn-ing directed environment caused some level ofdiscomfort that needs to be balanced with theworkload. That is, students should not ex-perience a level of discomfort that interfereswith the quality of their learning. Perhaps an in-strument that measures students’ epistemologycould help identify which students may needadditional support to make the shift without thefeeling of overload and discomfort that inter-feres with a quality learning experience. Toomuch stress and overload is destructive anddoes not stimulate learning. Also, otherdemands, affecting students outside the class-room, should be considered.

Analyze motivational levels

Relevance and motivational elements are criticalto the learning process. Lack of motivationdiminishes participation and ultimately learn-ing. It is important to analyze the motivationallevel of the audience and establish motivationalstrategies to measure the effect motivation playson learning. Self-report measures, such as con-fidential reports and instruments specificallydesigned to measure motivational levels relativeto a particular approach to learning, could beused. Additionally, tasks and activities per-ceived as busy work negatively affect the col-laborative discourse of the community.Community activities should have an apparentpurpose in achieving the goals of the com-munity and should be authentic relative to thecontent domain. A whole-class problem-solvingactivity holds promise to promote collaborativeknowledge building if the activity serves as ashared vision rather than a collective goal.

Promote a shared vision

The shared vision of the community requires so-cial support at the group and community levels.The instructional strategies and technologicaltools should promote collaboration acrossgroups and group work space enabling the com-munity to engage in quick interchanges of ideasand knowledge-building discourse. This dis-course should provide feedback to the systemfrom all the participants in the learning com-munity, including students, teachers, and out-side experts.

Promote the acquisition of knowledge

The acquisition of key concepts is critical inproviding a strong foundation for a knowledge-building enterprise. Key concepts could be in-tegrated into class activities supported bygenerative learning strategies and discourse topromote their acquisition. Key concepts are fact-based information that provides the foundationnecessary to enable students to engage inhigher-order thinking and promote knowledge-building discourse beyond the group to the com-munity level.



CONCLUSION

The importance of tracking the learning processto guide intervention by the instructor, the needto balance tension by promoting productive dis-comfort leading to growth, and the promotion ofa shared vision and tools that support com-munity collaboration are all new ideas thatemerged from this study. The use of scaffolds(Rosenshine & Meister, 1992), the promotion ofmotivation in learning environments (Keller,1987), and the acquisition of knowledge topromote discourse (Scardamalia & Bereiter,1994) have been examined previously and reaf-firmed by this study.

Educators have recognized the need to cul-tivate higher-order thinking skills and alternateviews of teaching and learning, even thoughthey have disagreed about how to achieve theseends (Resnick, 1989). One potential solution isthe design and implementation of collaborativeknowledge-building communities that incor-porate constructivist principles. A collaborative,knowledge-building approach offers an alterna-tive model for both the design of classroom en-vironments and distance courses. There is a newtrend in education to use Internet-based tools,and to use distance education to reach more stu-dents over a larger geographical area; however,alternative models are needed to support this ef-fort.

There is a growing body of knowledge aboutthe contribution of constructivist principles tothe design of collaborative learning environ-ments based on a community-of-learners ap-proach, and how emerging technologies can beused to support these ideas. This investigationrepresents a step toward providing empiricallybased guidelines to assist in the design anddevelopment of collaborative, knowledge-build-ing communities.

Nancy Gilbert is Assistant Professor and Director ofthe Instructional Technology Program in theGraduate Education Department at MarywoodUniversity in Scranton, PA. Marcy Driscoll is the Department Chair ofEducational Psychology and Learning Systems atFlorida State University in Tallahassee. This study was conducted by the first author as adissertation at Florida State University. She would

like to thank Marcy Driscoll for her patience andguidance, Walt Wager for his mentoring, and BobReiser and George Weaver for their support.

REFERENCES

Bell, P., Davis, E.A., & Linn, M.C. (1995). Theknowledge integration environment: Theory anddesign. In J.L. Schnase & E.L. Cunnius (Eds.),Proceedings of CSCL ’95: The first international con-ference on computer support for collaborative learning.Hillsdale, NJ: Erlbaum.

Bell, P., & Winn, W. (2000). Distributed cognitions, bynature and by design. In D.H. Jonassen & S.M. Land(Eds.), Theoretical foundations of learning environments(pp. 123–146). Mahweh, NJ: Lawrence Erlbaum.

Bereiter, C. (1995). Constructivism, socioculturalism,and Popper’s world 3. Educational Researcher, 23(7),pp. 21–23.

Bereiter, C., & Scardamalia, M. (1996). Rethinkinglearning. In D. Olson & N. Torrance (Eds.), Handbookof education and human development: New models oflearning, teaching, and schooling. Cambridge, MA:Basil Blackwell.

Bowen, B., Bereiter, C., & Scardamalia, M., (1992).Computer-supported Intentional Learning Environ-ments. In F.Y. Phillips (Ed.) Thinkwork: Working,learning, and managing in a computer interactivesociety. Praeger: Westport, Conn.

Brown, A.L. (1994). The advancement of learning.Educational Researcher, 23(8), 4–12.

Brown, A.L., & Campione, J.C. (1990). Communities oflearning and thinking, or a context by any othername. In D. Kuhn (Ed.), Developmental Perspectives onTeaching and Learning Thinking Skills. Contributionsto human development, vol, 21, 108–126.

Derry, S.J., & LaJoie, S.P., (1993). A middle camp for(un)intelligent instructional computing: An intro-duction. In S.P. LaJoie and S.J. Derry (Eds.), Com-puters as Cognitive Tools. Hillsdale, NJ: LawrenceErlbaum Associates.

Drucker, P.F. (1994, November). The age of socialtransformation. Atlantic Monthly, 274(5).

Duffy, T.M., & Cunningham, D.J. (1996). Construc-tivism: Implications for the design and delivery ofinstruction. In D.H. Jonassen (Ed.), Handbook of re-search for educational communications and technology.New York: Macmillan.

Edelson, D.C., Pea, R.D., & Gomez, L. (1996). Con-structivism in the collaboratory. In B.G. Wilson(Ed.), Constructivist learning environments: Casestudies in instructional design. Englewood Cliffs, NJ:Educational Technology Publications.

Hannafin, M., Land, S., & Oliver, K. (1999). Openlearning environments: Foundations, methods, andmodels. In C.M. Reigeluth (Ed.), Instructional-designtheories and models (Vol. II). Mahwah, NJ: Erlbaum.

Jonassen, D.H. (1995). Supporting communities oflearners with technology: A vision for integrating


78 ETR&D, Vol. 50, No. 1

technology with learning in schools. EducationalTechnology, 35(4) 60–63.

Jonassen, D.H. (1999). Designing constructivist learn-ing environments. In C.M. Reigeluth (Ed.), Instruc-tional-design theories and models (Vol. II). Mahwah,NJ: Lawrence Erlbaum Associates.

Kaufman, R., & Grise, P. (1995). Auditing your educa-tional strategic plan: Making good things better.Thousand Oaks, CA. Corwin Press, Inc.

Keller, J.M. (1987). The systematic process of motiva-tional design. Performance and Instruction Quarterly,26(9), 1–8.

Lebow, D. (1995). Constructivist values for instructionaldesign: A case study of a graduate-level learning environ-ment. An unpublished doctoral dissertation, FloridaState University, Tallahassee.

Lebow, D., & Wager, W. (1994). Authentic activity as amodel for appropriate learning activity: Implica-tions for emerging instructional technologies.Canadian Journal of Educational Communication, 23(3),231–244.

Lebow, D., Wager, W., Marks, P., & Gilbert, N. (1996,June). Construe: Software for collaborative learning overthe World Wide Web. Paper presented at theAECT/FSU Conference on Distance Learning, Tal-lahassee, FL.

Leinhardt, G. (1992). What research on learning tells usabout teaching. Educational Leadership, 4, 20–25.

Miles, M.B., & Huberman, A.M. (1994). An ExpandedSourcebook: Qualitative Data Analysis (second edition).Sage Publications. Thousand Oaks, CA.

Pea, R.D. (1994). Seeing what we build together: Dis-tributed multimedia learning environments fortransformative communications. The Journal of theLearning Sciences, 3(3), 285–299.

Perkins, D.N. (1991). Technology meets construc-tivism: Do they make a marriage? Educational Tech-nology, 31(5), 18–23.

Resnick, L. B, (1989). Introduction. In L.B. Resnick(Ed.), Knowing, learning and instruction. (pp. 1–24).Hillsdale, NJ: Lawrence Erlbaum.

Rosenshine, B., & Meister, C., (1992). The use of scaf-folds for teaching higher-level cognitive strategies.Educational Leadership, 4, 26–33.

Scardamalia, M., & Bereiter, C. (1992). An architecturefor collaborative knowledge-building. In E. DeCorte, M. Linn, H. Mandl, & L. Verschaffel (Eds.),Computer-based learning environments and problemsolving. [NATO-ASI Series F: Computer and Sys-tems Science] (pp. 41–46). Berlin: Springer-Verlag.

Scardamalia, M., & Bereiter, C. (1994). Computer sup-port for knowledge-building communities. The Jour-nal of the Learning Sciences, 3(3) 265–283.

Scardamalia, M., & Bereiter, C. (1996). Adaptation andunderstanding: A case for new cultures of schooling.In S. Vosniadou, E. deCorte, R. Glaser, & H. Mandl(Eds.), International perspectives on the design of tech-nology-supported learning environments. Mahweh, NJ:Erlbaum.

Senge, P., Kleiner, A., Roberts, C., Ross, R.B., & Smith,B.J., (1994). The Fifth Discipline Fieldbook: Strategiesand Tools for Building Learning Organizations. NewYork: Doubleday.

Slatin, J.M. (1992). Is there a class in this Text? Creatingknowledge in the electronic classroom. In E. Barrett(Ed.), Sociomedia: Multimedia, hypemedia, and thesocial construction of knowledge. Cambridge, MA:MIT Press.

Slavin, R.E. (1991). Synthesis of research on coopera-tive learning. Educational Leadership, 2, 71–77.

Wenger, E. (1998). Communities of practice: Learning,meaning, and identity. New York: Cambridge Univer-sity Press.

Yin, R.W. (1989). Case study research: Design andmethods. (rev. ed.). Newbury Park, CA: Sage.

Circle Your Calendar Now

The AECT Annual ConferenceNovember 12–16, 2002

Wyndham Anatole HotelDallas, Texas

Back in the Saddle Again!



Download - Collaborative knowledge building: A case study

Top Related