From Co-operative Learning to the Virtual Class. An Experience in Composition Techniques.

University of Valladolid

Keywords: University of Valladolid, Faculty of Education, Composition Technique, writing, learner-centered, teaching, learning, collaborative learning, computer supported co-operative learning, group-work, new technology, computer application, pencacolas, web-based teaching, on-line teaching.

Article style and Source: Peer Reviewed. A version of this paper has been presented at the Conference in Bergen, The Future of Humanities in the Digital Age, September 25-28, 1998.

Contents

• Abstract
• Introduction
• Composition Techniques and CSCL
• Pencacolas in Composition Techniques
• Pencacolas
  • Set-up
  • Some difficulties, solutions to problems and future plans
• Conclusion
• References

Abstract

Introduction

The use of written language as a vehicle for knowledge and communication is extremely important in a world where information means power. The literacy rate of a country is an indicator of its progress. But today the word literacy means much more than the simple use of the alphabet. A person is competent in reading or writing when he or she fulfils efficiently the requirements for the development of these abilities in our society. Therefore, at this point, rates have little significance. Unfortunately many people with schooling have problems filling in a registration form or composing a letter. Such situations are not isolated. An increased number of students arrive at University with a limited experience in composing.

We do not analyse the origins of this problem here. However, it requires our attention as teachers, especially in the Humanities field, since writing plays a significant role in both the academic and future professional lives of our students.

In this paper we present PENCACOLAS, an application in the computer supported co-operative learning (CSCL) group, which was created to support collaborative composition in the Faculty of Education at the University of Valladolid. Our work may be conceptually divided into four parts. First we describe the course context and its educational requirements. Secondly we present the implementation of PENCACOLAS in a real classroom. Thirdly we discuss some of the technical and educational aspects of our research and finally we draw some conclusions and suggest projects for the future. top

Composition Techniques and CSCL

Writing is a very important element of education at the University of Valladolid because it is considered both a subject of study as well as a way to learn and demonstrate what is learned. Consequently, the Faculty of Education offers a writing course called Composition Techniques. This is not a basic or remedial course, but an opportunity for intensive writing practice and interaction among students. The course focuses on the writing processes in order to produce better final results and is learner-centered. So, Composition Techniques affords students the possibility of practising the different kinds of texts that they most often need at the University.

We investigate the teaching /learning processes from a socio-cultural perspective. (Vygotsky, 1934/1985; Witte-Cherry, 1994) in order to understand this process in the context of an interactive classroom, which surpasses the individual for more effective learning. From this point of view, collaborative writing defined as collaboration among people to generate a text is not only a way to compose, but also a way to learn. This approach is also valuable because collaboration plays an important role in the future of our students since group organization is one of the most common ways to work in today’s professional life.

Group collaboration has an important ally in the new technologies, especially in computer supported co-operative work (CSCW) applications. These kinds of systems have opened new roads for education, because they integrate the advantages of computer assisted work with group collaboration. This specific application from the CSCW is called computer supported co-operative learning (CSCL).

The use of CSCL systems is relatively widespread (Baecker, 1993). However, there is little evidence of the performance of such systems in a real classroom. In our research we present a pen computer aided collaborative system (PENCACOLAS). This is a working prototype in the CSCL group used for teaching/learning composition in our course at the University of Valladolid. In the following section we describe the context and functional requirements from an educational perspective while in section 4 we present the prototype, experimental findings and discuss its advantages and disadvantages. Finally, in section 5, we draw some conclusions and suggest future projects. top

Pencacolas in Composition Techniques

As mentioned above, Composition Techniques is a course created at the Faculty of Education to help the students with their writing problems. The course is designed on a constructivist model. From this point of view, we understand text composition as a typical problem-solving situation (Scardamalia - Bereiter 1992), which develops as a recursive process (Flower - Hayes, 1980) with different phases, such as planning, writing and revision.

Consequently, the class is structured in small groups and is mainly orientated to create texts in a collaborative process that includes: planning, draft designing, revision, co-evaluation and frequent dialogue with the teacher to obtain guidance or corrections.

The students assume different roles in each group, in order to organize the writing tasks. The roles are flexible and may change according to the decisions of the students, so that they are sometimes writers and other times reviewers. The writing processes are not linear. A major objective should be to support the model in such a way that every student is aware of the global process while participating in it.

PENCACOLAS allows the generation of a text by the simultaneous connection of a small group of students plus the teacher. The application is structured in the classical writing phases such as planning, composition and review, and permits users to follow them in real time. Besides the synchronous interaction, it is also possible to develop asynchronous activities, such as peer revision or exchange of suggestions in the form of short messages. The system uses a pen computer interface that allows it to convey metatextual information, typical for the work with classical pen and paper (Kimura, 1992).

The students and teacher need to look back through the different drafts and texts in order to review, correct and evaluate them. PENCACOLAS allows the users to access previous composition phases for all of these purposes.

In all of these operations it is important to preserve not only the final versions, but also all the intermediate writing-steps, which can provide valuable insights regarding the evolution of each piece of writing and the cognitive development of the students. To implement this approach we adopted a growth portfolio, a system (Arter et al., 1995; Cleland - Porter 1998) that we use for formative evaluation as well.

The main advantage of the PENCACOLAS portfolio system is the accuracy in the collection of writing events (Valencia, 1993). It therefore avoids some of the problems found when using a traditional paper portfolio such as a large volume of the generated files and the tendency of students to include in the portfolio only their best pieces and to forget or ignore many intermediate steps of their work.

To solve these problems PENCACOLAS records all writing events, during all phases so that they can be analysed later on. This is an important feature of the system because it allows us to investigate how the composition patterns are developed (Witte-Cherry, 1994), and how the writing skills used by the students can improve. Moreover, the opportunity to review the whole work helps the learners to be aware of their performance as writers. This aspect of the learning process is helpful for them and for CO-evaluation

From the point of view of the teacher, it is important to analyse what kind of learning processes happen in a whole session. In response to this requirement, PENCACOLAS automatically generates filenames that identify users, sessions, and phases, thus allowing easy searching and information retrieval. Therefore we obtain not only a final draft, but also a written interaction that allows us to follow the writing processes closely. The possibility of identifying every user in an exclusive file allows the teacher to analyse both co-operative and individual work (Figure 1). This opportunity is also helpful for the teacher as self-evaluation, since it allows him/her to see the interactions from each session (Blasco, 1998).

In a collaborative learning environment the teacher has to change. The role of "instructor" is not valid in a framework where students regulate their own learning patterns. For such a classroom the teacher becomes a moderator or weaver (Davie, 1989; Feenberg, 1989), who observes the transition between phases and analyses the whole teaching/learning process, with a view to improving it.

PENCACOLAS helps the teacher to follow each student during the sessions, as well as to participate in the common discussions and drafts without physical presence. With PENCACOLAS the teacher becomes a silent witness who appears only when necessary. This allows a more relaxed atmosphere in the classroom, because the students do not feel observed.

Nevertheless, the communication among all class members should be guaranteed. Students prefer common discussion spaces, but one to one interactions with the teacher or short talks among class members are frequent in collaborative work.

PENCACOLAS provides for both types of communication through the use of shared windows that are available during co-operative phases. A classical message interface is also implemented, so that every classroom member can use it for information, queries or advice. This kind of message is clearly shown on user screens and permits asynchronous communication. The short messages are especially useful for task co-ordination.

The transition from individual work to group work revealed some socio-cultural problems such as group and authoring awareness. Our experience with PENCACOLAS shows that students felt strange as anonymous workers and tended to identify themselves and the others quickly (Figure 2). This natural reaction related to the shared space reveals an ancient mechanism of human relationship, i.e. the need to know who is who in a group. To be aware of the presence of the other is important if one is working in a shared space. Thus, PENCACOLAS should provide answers to question such as how to recognize the activity of each member, or where the other members are.

To summarise the main needs of the Composition Techniques course, we could say that:

• Composition techniques need flexible and shared communication spaces for small groups
• The teacher is a co-ordinator and must have easy access to students' work
• Synchronous and asynchronous work and communication are necessary
• Group and authoring awareness must be facilitated.
• Awareness of every student's own writing processes must be enhanced
• The teacher needs an easy way to analyse the composition process after this has been concluded.
• Access to common and individual work must be facilitated

In the following section we describe how PENCACOLAS translated these functional and educational requirements into technical solutions. top

PENCACOLAS

An initial prototype of PENCACOLAS has resulted from the above requirements and has been used as experimental work by students of the Faculty of Education for two years. (Blasco et al., 1996; González et al. 1997). Currently, a systematic study is taking place in real courses, using the latest version, based on owner client/ server software and built in a TCP/IP network. In the rest of this subsection we discuss some of the most important aspects of the system, present its problems and advantages and propose future investigations.

Set-up

The set-up for the experimental sessions of up to three students and one teacher consisted of the following elements

• Computers: A Sun SPARCstation, IBM compatible desktop PC´s and NEC Versa-E notebooks
• Operating environments: Solaris 2.4 for the SPARCstation and Windows for Pen computing 1.1 installed under windows 3.11 for the PC´s and the notebook
• Platforms for Pen User Interface: a Calcomp 23120 opaque tablet, with a pen connected to serial port (only input device), a transparent digitizer Wacom PL-100 with a cableless pen and a Touchpen display with a cable pen for the notebook computer (input and output devices)

The set-up has recently been updated. The opaque Calcomp tablet has been replaced by a transparent digitizer and the operative environment has been installed under Windows 95.

Communication among all class members was supported in flexible ways that permitted a variety of group interactions. Students were able to draw individual brainstorming or scenarios, which appeared as white boards (read-only windows). However, they preferred the use of shared windows as a blackboard during co-operative phases. Read-only windows could also be shown in co-operative work in reduced size, to avoid the classical space problems of shared windows (Figure 3).

The use of the pen interface was helpful in the conveying of metatextual information such as drafts, annotations and gestures, and permitted a smooth transition from traditional pen and paper to computer. A message interface was also available for urgent information or queries.

The teacher was connected on line with the writing team (three students) and was able to follow individual work as well as the co-operative work. Teacher could also regulate the transition between phases. Likewise, the students were able to look at the teacher's window for information or help.

The use of shared spaces contributed to create group awareness. To facilitate identification among the writers, PENCACOLAS assigns a different cursor for each group member. The space distribution resulted from a simple social protocol. Each one assumes his space and tries not to invade the writing region of others. This approach prevented most of the concurrency problems.

Graphic markers -nicknamed "thermometers" by the students - indicated whether a writer used a certain page, while scrollbars permitted movement on each page.

Finally PENCACOLAS stored the collaborative work and the action of every person in separate files. The writing events were stored as they were generated (Figure 4). Therefore we obtained a dynamic image of the process instead of many static images. (pixmapfiles), with the following results:

• It helped the students to revise previous drafts and texts. This feedback made them aware of the recursive nature of writing.
• This kind of storage facilitated analysis of common and individual work to the teacher because PENCACOLAS makes it possible to see the different actions of the users as an overlapping slide
• For peer revision, students were able to select suggestions from each member separately (Figure 5).
• In the experimental work we observed that the storage requirements of the event files are about fifty times smaller than those of the pixmapfiles.
• Additionally, the files served as electronic growth-portfolios for the evaluation. top

Some difficulties, solutions to problems and future plans

The system developed provided valuable insights and findings throughout the period of study. However, the implementation of the system in a real environment posed some difficulties. First we had to solve certain problems of collaborative work such as the establishment of roles, the organization of the tasks among the students, the management of the discussion in order to write a text and the negotiation of the most interesting ideas. (For instance, many groups spent too much time discussing the topics and not enough time talking about the written structure).

Such problems are actually normal in collaborative writing situations with novice writers. They have to perform their thinking and learning styles in order to collaborate. The development of social abilities however, helps them in the task organization.

To solve these problems, the teacher at first assumed the greater part of the co-ordination. role, assigning limits for discussions and giving advice about the generation and collection of ideas or draft elaboration. Later the figure of the teacher gradually disappeared and the group worked autonomously.

Second, there were the computer and telematic problems, which emerged in the co-operative phases. In these phases the system had to implement mechanisms of concurrency control.

We proposed these solutions:

• To create cursors of different type which would identify the user and the
• action (write, erase, etc.),
• To allow any student or the teacher to observe the work of
• the others in any phase (either for orientation or help purposes),
• To solve the concurrency problem writers used a simple social protocol, dividing the space and assigning a region to each co-author during the co-operative phases.

Another problem was the excessive network traffic due to the size of the generated files. The traffic delayed the answer of the system to the actions requested by the users and therefore generated a sensation of frustration that sometimes misled the users to believe that the system did not work

The solutions adopted in this case were:

• To create a new logical telematic net among the computers (which permitted user
• access to the system)
• To redesign how events are captured through logical processes distribution techniques,
• To use mechanisms of data compression in order to facilitate file transfer
• in the network
• To design a new mechanism of the agent through the distribution of processing
• activities and concurrency control

Parallel to the use of the proprietary version of PENCACOLAS, we have designed and developed a Web-based alternative. Appropriate software tools for an Agent Description Language (ADL) and the code generator ADL2JAVA allow us to overcome the problems of software engineering in CSCW and rapidly obtain new prototypes, easily adaptable to various conditions.

Transition to a web environment was smooth because the PENCACOLAS architecture has a strong component of replication with a central agent that co-ordinates all processes. However, we need to reduce network traffic and dependence on a remote server that might be critical in distance education. top

In this paper we proposed an alternative approach for achieving the educational objectives mentioned. This approach is based on the design, development and experimental study of a system for teaching composition. Close collaboration with final users, as well as experimental work for the analysis of learning patterns and group dynamics, were shown to be crucial for the final success of this CSCL product. Finally, the pen-computer employed in our system permitted a smooth transition from conventional writing to the computer, and provided efficient graphic dialogues.

Besides studying written expression as a basis for collaborative learning, we were able to draw generic conclusions about the possibility of using it in inexpensive, low-band-width terminals and networks. Past experiences at Duke University (Stetten et al., 1995) and the University of California at Berkeley (Narayaswamy, 1996) showed that solutions of this type are technologically feasible. These findings suggest that integration of mobile computers with low bandwidth requirements is possible for both in-class and remote teaching, thus coming closer to a fully virtual class. top

References

ARTER, J. et al. (1995) Portfolios for assessment and instruction. Research Report. Available in: http://ericae2.educ.cua.edu/db/digs/ed388890.html

Baecker, R. M. (1993) Readings in Groupware and CSCW: Assisting Human Human Collaboration. San Francisco: Morgan Kaufmanns Publishers.

Blasco, M. T. et al. (1996) Estudio de escritura colaborativa asistida por ordenador. Aplicación basada en lápiz electrónico. Proceedings of the ECER´96 European conference on Educational Research. Sevilla.

Blasco, (1999) La evaluación de la composición escrita: El método de carpetas. Lenguaje y Textos, 13,151-164.

Cleland, J. & Porter, C. (1998) The Portfolio As A Learning Strategy. Available in: http://www.hcs.k12.sc.us/learnser/curr/stnd/langarts/strategy/portfol.html

Davie, L. (1989). Facilitation techniques for the on- line tutor. In Mason & Kaye (eds.). Mindweave. (pp.74-85). Oxford: Pergamon Press.

Feenberg, A. (1989). The written world. In Mason & Kaye (eds.) Mindweave.(pp.56-98) Oxford: Pergamon Press.

Flower, L. y HAYES, J. (1980) The Dynamic of Composing.Making plans and jugling constraints In Gregg, L.W. y Steimberg (eds.) Cognitive Processes in Writing (pp. 31-50). Hillsdale NJ: Lawrence Erlbaum.

González, O. M.. et al.(1997) Pen-computing and Groupware. A chance to improve collaborative writing. Proceeding's of HCI International 97. San Francisco.

Hayes. J, y Flower, L. (1980) Uncovering Cognitive processes in writing: an introduction to protocol Analysis. In Gregg & Steimberg (eds.) Cognitive Processes in Writing.(pp.55-70) Hillsdale, NJ: Lawrence Erlbaum.

Kimura, T. D. (1992).: A Comparison Study of The Pen and The Mouse in Editing Graphic Diagrams. Technical Report. Washington University, St. Louis: Laboratory for Pen- Based Silicon Paper Technology, (Pen Lab). Department of Computer Science.

Narayaswamy, S. et al. (1996) Application and Network Support for Infopad. IEEE Personal Communications, 3, 4-17.

Scardamalia, M. y Bereiter, C. (1992). Dos modelos explicativos de los procesos de composición escrita. Infancia y Aprendizaje, 58, 43-64.

Smagorinsky, P. (1994). Speaking about Writing. Reflections on Research Methodology. Thousand Oaks: Sage.

Stetten et al., (1995) Wireless Infrared Networking in The Duke Paperless Classroom. Technological fronts in Education, 23, 3.

Valencia, S. W. (1993). Método de carpeta para la evaluación de la lectura en clase: los porqué, los qué y los cómo. CL&E, 19-20, 69-75.

Vygotsky, L. S. (1985). Pensamientoy Lenguaje (trans. from original 1934). Buenos Aires. La Pléyade.

Witte, S.P. y Cherry, R. D. (1994). Think-Aloud Protocols, Protocol Analysis, and Research Design: An Exploration of the influence of Writing Tasks on Writing Processes. In Smagorinsky. (ed) Speaking about Writing. Reflections on Research Methodology. (pp.20-54) Thousand Oaks: Sage. top

If you want to print the screen capture please click here.

We wish to thank to Diane I. Garvey and Barbara C. Knowles for helping in the translation of this document

About the Authors:

M. T. Blasco, J. L. Barrio, Y. A. Dimitriadis, C. A. Osuna, O. M. González, M .J. Verdú, D.Terán

M. T. Blasco, J. L. Barrio,
Departamento de Didáctica de la Lengua y de la Literatura.
Facultad de Educación.

Y. A. Dimitriadis, C.A. Osuna, M.J. Verdú, D. Terán
Departamento de Teoría de la Señal y Comunicaciones e Ingeniería
Telemática.
E.T.S.I Telecomunicación.

O. M. González
Departamento de Organización y Gestión de Empresas.
E. U. de Estudios Empresariales.

All authors are members of the E.M.I.C. group (Education, Media, Informatics & Culture), an multidisciplinary team at the University of Valladolid which aims to promote dialogue and research in Telematics and education.