This study is about how people say they learn in a context that enables communication only by computer text messaging. The process of how people learn - cognitive psychology - is not new, but the context - computer-mediated communication (CMC) - is relatively new. The additional perspective - the students' perceptions of how they went about learning - is definitely new for research into CMC in higher education.

The study was prompted by several questions and guided by several bodies of literature. This chapter gives those questions, lists the four research questions for this study and links them to several literature bases. Those bases are explored in chapter 2. Definitions of the key terms are needed first to help set the context.

Wittrock's definition stresses the relationship aspects of learning:

...student use of cognitive procedures for generating relations across bodies of information and between new information and memory (1992, p. 702).

One of the key experts in this field from whom Tessmer & Jonassen drew guidance is Weinstein who defines learning strategies as

Any behaviours or thoughts that facilitate encoding in such a way that knowledge integration and retrieval are enhanced. More specifically, these thoughts and behaviours constitute organized plans of action designed to achieve a goal...Examples...include actively rehearsing, summarizing, paraphrasing, imaging, elaborating, and outlining (Weinstein, 1988, p. 291).

Learning strategies are to be distinguished from learning styles. The concept of style refers to "stable ways of approaching tasks that are characteristic of individuals...styles are focused on the person, strategies on the task" (Biggs, 1988, p. 185). In other words, style refers to the consistent choice and use of preferred behaviours that are idiosyncratic to the person and usually consistent over time and task.

Learning styles are characteristic cognitive, affective and physiologic traits that serve as relatively stable indicators of how learners perceive, interact with, and respond to the (external) learning environment (Keefe, 1987, p. 16).

The term cognitive style is also used in the literature about learning, and sometimes with confusing effects. After discussing the confusion with a senior colleague, it was decided to regard cognitive style as a term referring to more internal-to-the-student processes of handling information. In the words of my colleague, cognitive style "describes ways of processing information... (and is) more closely associated with physiological factors such as cognitive functions, brain organization and internal energy patterns" (MacKeracher, 1988). The term learning style is used as an inclusive term to refer to how the student behaves consistently in the external and interpersonal contexts of learning, as well as the internal cognitive processing preferences.

The second key term for the study is computer-mediated communication (CMC). It is defined in terms of three functions:

Computer-mediated communication (CMC) is the set of possibilities which exist when computers and telecommunications networks are used as tools in the communications process: to compose, store, deliver and process communication. Such systems rely on a basic configuration of a mainframe computer with appropriate software, connected via telephone and data networks to users with terminals or micro-computers.

CMC covers a range of different facilities:

electronic mail, which allows messages to be sent to electronic letter-boxes for named individuals, which can be accessed when the named user logs on;

computer-conferencing, which allows messaging to be shared `openly' by all `members' of a conference;

access to remote data-bases, such as bibliographic sources, sets of abstracts, or details of courses available (Mason, 1990, p. 22).

This study focusses on the second function, that of computer-conferencing.

The applications of CMC in higher education are increasing rapidly. It has been stated that CMC can

...provide students with opportunities for convenient course-related or social interaction with peers... enable collaborative group work by distance students ... facilitate interaction with an instructor ... decrease turnaround time for instructor feedback ... allow students access to on-line resources, e.g., databases, library catalogues, and course registration ... and enable students to upload and download assignments and take on-line quizzes and tests ... (Wells, 1992, p. 2).

Would the introduction of CMC into higher education drastically reconstitute the conditions of learning? Or would it be less ecological in effect and more an extension of functions we already carry out? Draper uses the analogy of "gloves for the mind" to describe the latter type of impact. Gloves fit our hands with comfort and allow us to handle a wider number of materials and substances than we could with our bare hands (Draper, 1988). He predicts that the likely scenario for computers in education will be

...not that they teach anything that could not be learned otherwise, but by bringing material into easier and more convenient accessibility, they may transform what can in practice be learned in a given time...Their essential advantage... is analogous to tools: to extend the range that can be brought within the scope of personal interaction...(Draper, 1988, p. 170).

Ironically, the very same Riesman quote was used later for opposite reasons by another educator whose critical concerns about the societal effects of technology are well known. Postman argues that computer technology

...has not yet come close to the printing press in its power to generate radical and substantive social, political and religious thought. If the press was as David Riesman called it, "the gunpowder of the mind", the computer, in its capacity to smooth over unsatisfactory institutions and ideas, is the talcum powder of the mind (Postman, 1992, p. 116).

In any situation of educational technology innovation, I have learned from experience that lasting changes are often not as extensive as early enthusiasms would have indicated. The chief reason seems to be that people get caught up in the positive aspects of the change before recalling that there are less-than-positive effects for some of the users. Technology has a double effect - "[it] giveth and ...taketh away" (Postman, 1992, p. 5).

What then of CMC for higher education: will it be talc, gunpowder, or atomic fusion? While Levinson's suggestion may be dismissed as early-adopter hyperbole, there have been other but less dramatic suggestions and claims for the power of CMC. In the first book that claims attention to theory building for CMC in higher education, Harasim (1990b, p. 43) points to five characteristics of CMC that justify the elevation of CMC into a new domain of learning. Those characteristics are many-to-many communication, place independence, time independence, text-based, and computer-mediated interaction. These characteristics appear to be a reasonable selection if we accept that they are an educator's view of what is unique about CMC and that some of them are not exclusive to CMC. Harasim also claims that one of the themes running through her book is that "on-line education is a new environment, with new attributes, and requires new approaches to understand, design, and implement it" (Harasim, 1990, p. xviii). I interpret the Harasim view of the impact of CMC as more an ecological one than additive. Harasim argues that the

...fundamental practical issue for on-line education are two questions: "What are the effective uses of these new CMC media? Can we just transfer our existing set of conventional instructional practices, or do we need to develop a new set of practices better suited to the new tools (Harasim, 1990a, p. xviii)?

In introducing this landmark book, Turoff also appears to take an ecological view: "The approaches that teachers take in the face-to-face classroom are not appropriate" (Turoff, 1990, p. xii).

What to me was missing from discussions of such claims was the possibility that the alleged features of CMC might act to produce antinomies, i.e., situations having opposing effects or conditions. An example of this issue relates to the features of time availability - from asynchronous messaging, and the richness of data - from many-to-many communication: could it be possible that time availability for some students might become time compression for others as they almost sink under the deluge of many messages left at will (Grint, 1992)?

Those mental skills have been discussed at length (Anderson, 1990), especially in terms of how students acquire and understand knowledge (Wittrock, 1992). The skills needed to organize and store information in various forms have guided the algorithms used in educational technology to help the educator design learning activities and the student use learning strategies (e.g., Gagne, Briggs & Wager, 1992). The strategies have been categorized into at least two taxonomies. The taxonomy selected as a guiding framework for this study is from Tessmer & Jonassen (1988). They divide their taxonomy into "primary" and "support" strategies that, as a collective, support four basic functions of learning - internal information processing, interaction with external learning materials, establishing a productive and comfortable state of mind, and monitoring internal cognitive processes. The Primary Strategies refer to the first and second functions: the direct information processors - what it takes to organize, elaborate, recall and integrate information; and the active study strategies - what it takes to deal with actual learning materials. The Support Strategies refer to the third and fourth functions: they are designed to "improve the general cognitive functioning of the individual by relaxation, anxiety reduction and time management". These support strategies facilitate the operation of the primary strategies. Two sub-categories exist for the support group - meta-learning and preparation/execution strategies. Appendix B shows the full details.

There appear to be no overt links yet being made between the literature of learning strategies and the principles for teaching and learning in CMC contexts (Eastmond, 1992b; Wells, 1992). This situation may be only a function of the newness of CMC, but it does indicate an opportunity for research into how students learn in CMC. If such research is linked with the summaries of research into media in learning (Clark, 1992; 1983), then the claims of superiority and innovation laid for CMC in higher education may get closer examination. Hooper and Hannafin add their cautionary reminders to those of Clark:

... unique causal relationships between technology and learning have not been established conclusively. Indeed, it has been suggested that learning from any media has little to do with the inherent capabilities of the medium per se. Improvements in learning are more likely due to mediating factors, such as increased emphasis on lesson design, than to inherent differences among technologies ... Moreover, the effectiveness of IT [instructional technology] is more likely contingent upon the informed application of time-tested learning strategies than the technological capabilities of a medium (Hooper & Hanafin, 1991, p. 69-70).

Such a cautionary approach prompts questions additional to those posed earlier in section 1.3.

1. What features in CMC hinder and help the cognitive processing of information into long term memory and later retrieval for application?

2. How does a teacher recognize and promote cognitive processing in a CMC discussion?

3. Does the asynchronous mode of CMC actively inhibit cognitive organization of information?

4. Is information load an issue for learning? If so, what are its origins? To what extent is it a function of the inherent characteristics of CMC?

5. How do students transcend the sometimes inhibiting "published" nature of text messaging?

6. Does CMC - in and of itself - promote anything unique in learning? To what extent and under what conditions may CMC act as "gloves for the mind"?

Distance educators have recognized the potential of CMC for the current developmental wave in distance education (Bates, 1990; Kaye, 1992; Mason, 1988). Distance education itself is now an established and expanding arena of education. In Canada, for example, close to 2,500 courses in 38 universities were listed in the national directory for the 1992/93 academic year (Canadian University Distance Education Directory/Répertoire de l'enseignement à distance dans les universités canadiennes, 1992). Contrary to popular belief about distance education being used for the geographically isolated, an analysis of student enrolments in many university distance mode courses reveals that 60-65% of their students live in urban or semi-urban areas (Lumsden, 1990). The secondary school level shows equally strong activity. In Ontario, for example, approximately 90,000 students (most of whom were adults) were enrolled for the 1991/92 year in distance mode courses run by the Ministry of Education to help adults gain a high school diploma (MacKinnon, 1992). Distance educators have long recognized that many adults either cannot or will not attend time and place-bound education because their life schedules or their personal situations call for flexibility in what and how they study. Effort is put into the design of quality distance courses and the supply of materials and tutorial help so that adults can blend work, family and study. The success of The Open University in the U.K. (with approximately 95,000 students) and its continuing emulation around the world is evidence that this form of education has successfully met student needs. The major form of communication between tutor and student is often still paper-based in the form of assignments and examinations (the same as for on-campus students), especially in developing countries. Telephone consultations between student and tutor, and residential schools or weekends are provided in many countries. The advent in the early 1980's of audio-conferencing for distance education enabled group interactions of varying degrees of frequency and the technology became established in many areas of distance and continuing education in Canada and elsewhere. The linking now of simultaneous audio and computer communications and compressed video for voice and visual conferencing will help close the distances, psychological and otherwise. CMC, in its present text-only mode, has to be regarded as a useful interim technology that offers, to those who can afford it, significant advantages in terms of communication speed and ease of operation. Distance educators are exploring the strengths and weaknesses of CMC, especially its time convenience, freedom from scheduled travel, faster feedback, on-going dialogue and group orientation (Kaye, 1992; Lauzon, 1992; Mason, in press; Mason & Kaye, 1989).

The rationale underlying the use of CMC as a component of distance teaching systems can be summarized in terms of :

the convenience of an asynchronous communication mode, which liberates users from both time and space constraints;

its value as a medium of written communication, within a system in which students are graded essentially on the quality of their written work;

the enhanced levels of interactivity between and amongst students, tutors, course developers, and other members of a widely dispersed learning community;

the reduction of the isolation felt by many distance learners, and the potential of CMC for collaborative learning (Kaye, 1990, p. 228).

Kaye's summary leads into the issue of adults learning in CMC contexts. Here again I searched with little success for in-depth explorations that link adult learning and its facilitation with CMC. Mason & Kaye (1990) raise important issues and opportunities for enhanced learning but they do so with an apparent lack of experience with audioconferencing; they seem to imply that only CMC can be used for "cooperative projects" and peer sharing of experience. They also appear to regard the evolution of "truly independent learners" as the ultimate goal of distance education. While I respect their experience and their values for education and I agree with their conclusion that CMC be seen as "one of a range of interactive media", one of my educational goals is to help the evolution of interdependence in students through the provision of highly interactive conditions for peer learning. For me, there are some key issues to be explored in the adoption of CMC in distance education. Two questions immediately come to mind: how are students actually using the new technologies? and to what extent is that use a function of the medium itself, as distinct from skilled course design or facilitation?

This study therefore works toward an understanding of how CMC is perceived in two higher education contexts and how those perceptions relate to established knowledge of learning strategies.

We understand very little about the new phenomenon of learning in an electronic space. There is as yet very little data describing or analyzing teaching and learning within this asynchronous text-based (screen) environment (Harasim, 1987b, p. 119).

It is worth noting that any analysis of CMC literature reveals no published research that links learning strategies as developed in the cognitive psychology literature to CMC applications, let alone a further link to students' perceptions of their use of learning strategies. Descriptions of institutional applications, reports of research into instructor and moderator activity and message analysis, and discussions of its potential abound - all with thoughtful insights, but also mostly from the educator's viewpoint.

For this study, I also looked for material that discussed adult learning strategies as they related to the two basic drives or motivators of adult action - achievement and affiliation. How are the drives manifested through learning strategies and how could both motivators be represented in CMC - based learning? At the time of my data collection two key books on adult learning - Merriam & Caffarella (1991) and Smith & Associates (1990) - had not appeared but they will be used in this study to help interpret the findings.

The second reason is its prompt for researchers of CMC in higher education to assess the relevance of learning strategies research for areas already partially discussed such as self-direction (Mason, 1988), facilitation techniques for tutors and moderators (Davie, 1988), and the provision of "effective, active learning" (Hiltz, 1986).

Third is the attempted interdisciplinary approach to the study. The findings should help relate some dimensions of adult learning to the cognitive psychology literature, and link the new learning context of CMC with the literature of adult learning to learn strategies.

The literature relevant to this study comes from four bases. They are learning strategies from the cognitive psychology literature, applications of CMC in higher education, the learning to learn strategies discussed in adult education, and distance learning. The adult learning to learn strategy taxonomies provide some contrast to the learning strategies taxonomy used for this study. The use of the four bases should help to focus attention on students in CMC and on their own perceptions, to relate some dimensions of adult learning to the learning strategies literature, and to introduce a new learning context, i.e., the reduced cues and asynchronous communications of CMC, into the literature of adult learning to learn strategies.

Weinstein used more detail:

Learning strategies are considered to be any behaviours or thoughts that facilitate encoding in such a way that knowledge integration and retrieval are enhanced. More specifically, these thoughts and behaviours constitute organized plans of action designed to achieve a goal... Examples of learning strategies include actively rehearsing, summarizing, paraphrasing, imaging, elaborating, and outlining (Weinstein, 1988, p. 291).

Various descriptions of strategies are given (Dansereau, 1985, 1978; Derry & Murphy, 1986; Garner, 1990; Holley & Dansereau, 1984; Palmer & Goetz, 1988; Tessmer & Jonassen, 1988; Weinstein, 1988; Weinstein & Mayer, 1988; Weinstein & Meyer, 1991). All share the same basic cognitive goals - the selection of information, the building of internal connections between ideas, and the building of connections with life experience in the external world. Five broad categories organize the strategies: rehearsal (at basic and complex levels), elaboration (at basic and complex levels), organization of information, comprehension monitoring, and affect (Weinstein & Mayer, 1986, p. 316). Appendix A shows the Weinstein (1988) taxonomy as an example of categorization.

I selected Tessmer & Jonassen's taxonomy because it is relevant to my own learning processes and because its internal division of primary and support strategies seemed to be appropriate and comprehensive. Appendix B shows the full detail. The taxonomy rests on four key concepts in learning - internal-to-self information processing, interacting with external learning materials, establishing a productive state of mind, and monitoring learning.

The Primary Strategies fall into two groups. The information processing strategies cover what it takes to recall, integrate, organize, and elaborate information. The active study strategies cover what it takes to work in sequence through print learning resources.

First, I discuss the information processing strategies. Straight recall of information may involve behaviours such as chunking, organization, practice, and the use of mnemonic devices. Integration, the transformation of incoming information into a "more memorable form", involves the use of paraphrasing, examples (and non-examples) of the new information and relating it to the student's prior knowledge. Organization strategies help the student either develop a new structure or modify an existing one by grouping ideas into broad categories or displaying ideas and their relationships in visual or spatial maps. Elaboration strategies help the student add new information to old information; examples are creating analogies or images, looking for applications, or developing propositions.

Active study strategies focus on how the student treats learning resources - text, film, and audio material. These have been widely used as tool kits for students, but with some criticism that one person's tool kit is not another's. The SQ3R method and the MURDER method are well known study strategies. The steps in MURDER are setting Mood, reading for Understanding, Recalling information, storing (information) to Digest it, Expanding knowledge by self-inquiry, and Reviewing mistakes. The SQ3R method is based on the sequence of Survey, Question, Read, Recite and Review (Robinson, 1961).

The Support Strategies, the second major group, are designed to "improve the general cognitive functioning of the individual by relaxation, anxiety reduction and time management" (Tessmer & Jonassen, 1988, p. 34). They facilitate the primary strategies. Two categories - meta-learning and preparation/execution- are used to organize them. Meta-learning strategies help the student to monitor cognitive processes (as distinct from producing learning outcomes), and provide an executive-level control process for the choice of suitable activities for specific learning tasks. The planning and execution strategies help create optimal psychological states to maximize cognitive processing.

In terms of research into strategies, there are various directions and recognitions. Tessmer & Jonassen, for example, recognize the need for further searching:

While research has to some degree demonstrated that these strategies can be effectively learned and utilized, further research is needed to determine what strategies exist, which should be learned by a given learner population, and how they can be taught (p. 45).

Other writers have explored the impact of context on the use of learning strategies (Perkins & Salomon, 1989; Ramsden, 1988; Rogoff & Lave, 1984). Garner has indicated the lack of researchers' attention to context - "a footnote or worse, a nuisance" (Garner, 1990, p. 523). She has argued for the study of strategy use in and across various contexts so that strategic behaviour is better understood. The context problem is related to another research direction, that of failure to use strategies when it would have been appropriate to do so. In some contexts, strategies may not be necessary - "when learners have a great deal of well-organized and accessible conceptual knowledge, they have little use of general strategic routines for acquiring concepts" (Garner, 1990, p. 517). But when strategies are necessary, some students may fail to use them for any one of five key reasons: (i) failure to see that learning is not happening; (ii) poor study habits that merely "get the job done" at a surface level but which do not prompt higher order thinking; (iii) inadequate knowledge schemata that prevent a student from proceeding; (iv) low self-esteem or dysfunctional attributions of success that prompt doubts about personal success; and (v) failure to transfer strategies learned in an earlier context to a different present context (Garner, 1990).

When affect-related learning strategies were discussed in the literature, they focused on ways to eliminate internal and external distractions, on student beliefs about self-efficacy and attributions of strategies, and on ways to organize in order to meet stated learning goals (Dansereau, 1978;McCombs, 1988; Palmer & Goetz, 1988; Weinstein & Underwood, 1985). Palmer & Goetz (1988) recognize that while adult use of strategies is under-researched, there is little information available to indicate that "background knowledge, reading ability, metacognitive knowledge, and affect are associated with their use of strategies" (p. 52). Good readers, aware of text structures and the dynamics of reading and comprehension, consciously use a variety of strategies in flexible ways (Fischer & Mandl, 1984). Palmer & Goetz propose a model of learning strategy use that depends on three key student perceptions: the self as student, the attributes of the learning task, and the attributes of various learning strategies (p. 44).

Apart from the notes of Palmer & Goetz, I found no references to studies by the cognitive psychologists on strategy use by part-time and mature adult students in higher education. In addition, there were no mentions of related issues such as interdependent behaviours with peers, nor of the various psychological and cognitive stages that adult students may experience over an extended period of formal learning (e.g., Taylor, 1986).

The use of printed texts as resources for learning is a focus of the writing on learning strategies. Topics include the structuring of texts in order to improve comprehension (Hartley, 1992, 1985; Mayer, 1988; Weinstein & Mayer, 1986); strategies for reading to learn (Brown, Campione & Day, 1981); on the differences between surface level and deep level text processing (Biggs, 1988; Marton, Hounsell & Entwistle, 1984; Van Rossum & Schenk, 1984); and on the role of writing in learning (Applebee, 1984; Biggs, 1988). Writing to learn happens when that writing "is being seen as the tool par excellence for clarifying and extending thought in most content areas" (Biggs, 1988). Applebee (1984) explained that writing acts to clarify information processing because it "externalizes thought" and it demands that the writer be "particularly clear and explicit" and pay proper attention to her/his audience.

In summary, the research into learning strategies and the research into student approaches to study presents epistemological limits for this CMC study. It has been carried out under relatively controlled experimental conditions; the subjects are usually young and full time college students or children; it has emphasized the teacher's manipulations of learning environments and processes; it appears to see the student in isolation from peer influence or benefit; and, Richardson says, has "virtually ignored...individual differences and qualitative change among the students themselves" (Richardson, 1987, p. 4).

The key characteristics in this image are task and time management, variety, self competence, connectedness, self-awareness and evaluation, and success. These characteristics are echoed through much adult education literature in discussions about learning and its facilitation (Brookfield, 1986; Smith & Associates, 1990). This section of the literature review therefore outlines topics from those discussions that are most salient for the study. It also presents two taxonomies which show how the topics translate into some broad guidelines for practice. The salient topics are goals and conceptions of learning, societal contexts of learning, self-concept and self-esteem, and facilitation.

The generic goals of learning for adults may be summarized as learning to live - to solve problems, to complete tasks or manage various life roles (Cross, 1981; Knowles, 1984, 1990; Merriam & Caffarella, 1991; Tough, 1982). Some of this learning may be procedural in the sense that students need to know specific knowledge and/or skills for a particular situation. Some of it may be generative in that it has to be transferable to other settings. Learning may have the goal of reproduction of information or critical reconstruction (Brookfield, 1987; Friere, 1970; Mezirow, 1990). Gibbons argues that there are two broad aims in adult learning:

...effective performance is often the raison' d'etre of learning, and achieving healthy emotional states is often the primary goal (1990, p. 81).

These aims relate to two basic drives of human behaviour - the drives for mastery of experience and belonging with others (Brundage & MacKeracher, 1980; Jones, 1968), or "agency and communion" (Bakan, 1966), or as Ryle put it, "an innate and powerful drive to relate to others, and a continuing attempt to make sense of their experiences." (Ryle, 1975, p. 1). These motivators of achievement and affiliation are concerned with feeling and being competent and being connected with other people - in Gibbons's words, being "confident, competent and compelled" (1990, p. 88).

Various conceptions of adult learning have been proposed. Boud describes four conceptions of adult learning: the training for specific competencies approach (often work-based), the self-directed, andragogical approach, the humanistic, student-centred approach and the critical pedagogy, social action approach (Boud, 1987, p. 223-227). Cheren says that the process has to be "efficient and effective, more active and reflective" (Cheren, 1990, p. 267). Candy refers to the "more heuristic and unpredictable nature of learning in the `real world' [that] tends to emphasize the search for meaning" (Candy, 1990, p. 37). Griffin, in a detailed exploration of learning processes (1987, p. 216-218), stresses the advantages of students naming their own learning processes as a function itself of learning, and lists forty uncategorized learning processes (which I see as strategies) from her graduate students. My analysis of those processes produced seven key categories: intrapersonal and interpersonal competencies, looking for both security and risk, seeing connections in new learnings, using multiple ways of thinking, unlearning, creating own meanings, and becoming self-aware. Brookfield, in a recent summary of students' perceptions of facilitation, identifies similar connections, challenges and skills involved in adult learning:

...the learners involved reported that the educational activities most meaningful for them were those in which they could make a direct connection to their past experiences or current concerns....learning episodes...of greatest personal significance were those in which they had to confront and work through some kind of challenge...they spoke proudly of how they had come to resolve contradictions, to appreciate ambiguities, or to develop problematic skills, overcoming their fears and poor self-images as learners in the process (Brookfield, 1991, p. 206).

Any discussion of adult learning should be placed in a context of relevant societal contexts. Today much of how society operates is dependent upon the generation, selection, organization, storage, and retrieval of information in a large array of formats. Such formats give their users unprecedented speeds of access to and manipulation of data and capacities for data base access and personal networking (Scientific American, 1991). Such trends also bring the issue of managing that information (Postman, 1992; Roszak, 1986). Gooler proposes four management strategies: finding and accessing information resources, conceptually organizing information as it is discovered in a variety of essentially different sources, self-diagnosis of one's own learning needs and progress, and strategies for collaborative learning (Gooler, 1990, p. 322-324). The first three strategies are not an unusual choice, given the characteristics of variety, speed, and quantity of information sources and the adult students' needs for efficient and productive learning situations for them. In my observations of busy adult students, and of myself for that matter, pragmatism often has to win over perfection; any steps necessary will be taken to maintain some semblance of order and containment without loss of integrity. The fourth strategy - collaborative learning - is somewhat of a surprise. I (like many other women and some men) already have been socialized to work in collaboration, not in competition. As an adult educator who has worked in audio and visual classrooms and who has been additionally socialized in the andragogical values, I take collaboration in learning as a given; albeit a complex and sometimes difficult one to implement especially when its social demands conflict with my task-related anxieties about academic rigour.

The areas of self-concept and self-esteem and learning are crucial for adult learning. An adult's life history, present stress load, quality of self-concept and level of self-esteem may contribute in a learning situation to feelings of anxiety and undue dependence on the teacher. The ending of Laing's poem "Knots" captures well this essence of dependency :

You may know what I don't know,
but not that I don't know it,
and I can't tell you,
so you will have to tell me everything (Laing, 1970, p. 56).

As adult students make the shift from teacher-dependency to self-responsibility, they feel confused and may engage in dysfunctional behaviour as they struggle to let go of old ways and trust themselves to be competent, interdependent students (Brundage & MacKeracher, 1980). In some recent reflections on our helping adult students to change their image of themselves as students (Burge & Haughey, 1993), the words "confusion, resentment and resistance" (Knowles, 1990, p. 123) captured well the feelings we saw in our students. Self-deprecation and an inability to recognize their own competencies and achievements has been documented by Tough:

People simply are not in touch with the variety, competence, and success of their self-planned learning projects or their learning with friends and family, or with how thoughtful, active, and responsible they are during such learning (Tough, 1990, p. 294).

Adult learning strategies have often been discussed in relation to the duties of the facilitator (Brookfield, 1986, 1987; Brundage & MacKeracher, 1980; Cranton, 1989; Galbraith, 1991; Hiemstra, 1991; Hiemstra & Sisco, 1990; Knowles, 1984). The key aspects here regarding the use of learning strategies are ergonomic and psychological conditions and those related to adulthood, such as life experiences that may help and hinder learning, the capacity to critically reflect upon experience, the benefits from effective group work, knowledge of personal learning style preferences, and problem-solving approaches to learning. Facilitation may be seen in terms of helping the development of strategies for application in such contexts as experiential learning (Cunningham, 1983: Kolb, 1984), holistic learning (Griffin, 1988), perspective transformation and critical thinking (Brookfield, 1987; Mezirow, 1990, 1991), reflective learning (Boud, Keough & Walker, 1985; Boyd & Fales, 1983), group-based learning (Johnson & Johnson, 1991; Tiberius, 1991) and problem-based learning (Barrows, 1985).

The most comprehensive discussion of strategies I have seen to date are those in Smith and Associates (1990). Most relevant are the Gibbons taxonomy (Gibbons, 1990) and the Alberta taxonomy (Collett, 1990); details are in Appendices C and D. These taxonomies were published after I had collected the data for this study.

The taxonomies are important because they show strategies that complement the information processing ones found in cognitive psychology. Specifically, the additional strategies are ones for management of self-concept and self-esteem, management of stress, and building relationships with relevant people. Three themes emerge here: the "presence" of the student as a strong figure in the learning context, the emphasis on activity that takes re-constructive, not reproductive approaches toward knowledge building, and collaborative or peer group learning. These themes are not evident to me in the Tessmer & Jonassen taxonomy.

The Alberta taxonomy best integrates these themes and my own umbrella categories of cognition, self-management and relationships. The `self' of the student is emphasized here in the identification of skills for understanding one's learning styles and needs and accepting responsibility for learning, and skills for managing one's mind, body and time commitments. A second group of strategies relate to understanding the learning environment - how to use its resources and learn on one's own and in groups. The third group of learning strategies in the taxonomy refers to the application of problem-solving techniques to make "rational and reasonable decisions" (Collett, 1990, p. 261). The full taxonomy (Appendix D) emphasises independent and interdependent activity, self- management and self-competence, variety, and relevance to personal interests. The information processing and metacognition emphases of Tessmer & Jonassen are represented here too, but the Collett taxonomy appears to be more inclusive for adult learning contexts.

Having shown how theory building has focused on factors such as autonomy (Wedemeyer, 1981) and individualization and dialogue (Moore, 1980) without much concern for the student's psychosocial environment, Gibson calls for more research into autonomy from the perspective of the students and their context. She refers to three other writers in distance education who focus on the need for two-way communication between student and tutor, and cites the well-known concept of guided didactic conversation (Holmberg, 1983) as an example of communication. Garrison & Baynton (1987) proposed a theoretical model of control that needs a balance of independence, power and support; Gibson refers to this linking in order to emphasize the need to take the environmental factors of support into account when thinking about student learning (Gibson, 1990, p. 131). But she does not discuss learning strategies per se.

If research into distance learning is narrowed to qualitative studies, then the area is comparatively very small. Researchers have called for the increased use of qualitative methodologies and the naturalistic paradigm (e.g., Grace, 1990; Minnis, 1985; Morgan, 1990, 1991) but as Morgan summarizes it, these methodologies and values are outweighed by other factors in distance education:

There is a diverse collection of projects clearly located in the experimental and correlational tradition. For example, experimental studies comparing the efficiency of learning through televised instruction, correlations of course features to test scores, aptitude - treatment - interaction studies, to mention a few. It seems that the culture of positivism and the research and evaluation originating within this culture is well represented. There are a few studies in the qualitative - illuminative tradition, as well as the use of survey within the context of lobbying for underprivileged groups, but hardly on the scale to suggest a revolution is occurring (Morgan, 1990, p. 19).

The topics of study include students' perceptions of the factors that contributed to either course attrition or completion (Brindley, 1987; Gatz, 1985) the events and processes considered significant for learning (Fales, 1984), drop-out (Parlett & Woodley, 1985), the teaching and learning conditions that affected the development of learning independence (Inglis, 1989), the transitions of neophytes developing new work roles and self concepts as a result of studies (Herrmann, 1988) and the effects of social integration (Hotchkis, 1992). A multidimensional, longitudinal Australian study is investigating the whole experience of doing an MBA by distance and its impact on the students' ongoing professional life and development (Holt, Petzall & Viljoen, 1990).

Most of these qualitative studies focus on approaches to studying, with some mention of strategies. The most recent example is Olgren's doctoral research (1991) which integrates learning goals, strategies and outcomes in order to focus on approaches to learning. Her results enabled her to conceptualize three approaches - reproducing, comprehension and application. While students in the first two approaches were oriented to extrinsic issues around grades and learning, the students in the application approach group were driven to a higher quality of educational experience:

They balanced intrinsic goals with assessment demands, had good organization and integration skills, and they reflected about the meaning of the material in relation to their lives. They conducted, as Holmberg (1986) recommends, a conversational dialogue with the text that contributed to outcomes of understanding and application (Olgren, 1991, p. 83).

The most extensive group of published reports has, since 1982, come from the OU (UK) and uses the phenomenological work of Marton and Svensson to provide qualitative descriptions of student experiences as they illuminate orientations to study. Morgan (1984) originally reported on this pioneering work of the Study Methods Group at Open University. He later reported the results of a similar enquiry into general student perceptions in relation to a course on computing (Morgan, 1989), and co-authored a summary of longitudinal studies of learning (Morgan & Farnes, 1990). In one of their earliest public reports, Group members (Morgan, Taylor & Gibbs, 1982) used their own and others' research to suggest that learning is influenced by students' orientation to study and development as students, their approaches to texts (i.e., deep or surface processing) and the demands of the learning materials.

While the use of qualitative methodologies is being encouraged, the results so far do not directly assist this present study. It is fair to point out though that CMC is so relatively new; perhaps it is just too early yet to expect some qualitative studies. In addition, this present study includes the context of peer or interdependent learning. Most earlier studies on two-way communication and interaction in distance learning have of necessity focused on the student-tutor dyad because the chief formal communication vehicle was the written assignment. This study, if it does nothing else, will add to a small data base some results that may generate studies to assess transferability of the findings. It may also go some way toward meeting Gibson's demand for an increased understanding (or conceptual dry cleaning) of that much used word - autonomy:

[What is needed] is a look at autonomy from the learner and learner environment perspective (Gibson, 1990, p. 30).

This study will also break new ground in its examination of learning strategies that are not focused on published texts, as was the case in Small's small-scale study of learning strategies (1985). My interviewees are "publishing" their own texts.

However, much of the writing on CMC to date is of little relevance to this study. Here is an example. Some CMC proponents have argued that a significant advantage of CMC is its augmentation - the many-to-many communication capability of the software leading to enriched resources for learning (Harasim, 1990b). But there is little discussion about the practical issues of such extensive communication, especially as it may produce too many messages for a busy student to handle easily and productively. Augmentation as a concept has many dictionary meanings - magnification, intensification, accumulation, addition, reproduction or repetition, but there is not yet enough empirical research evidence to help identify its meanings in practice.

There are other reasons why the current literature on CMC in higher education is not helpful for this study. It is often so institution specific or even lacking in contextual detail that the reader cannot draw out general ideas. The repetition of case studies of the "this is what we did" variety, without any analytical referencing to theory testing or theory generation is even less useful, especially when references to adult learning principles are missing. An example for me was Hiltz's opinion that "CMC supports collaborative assignments and in-depth discussions, whereas the traditional classroom does not" (Hiltz, 1990 p. 164; Hiltz & Meinke, 1989), or that active and group learning are equivalent to "curricula reform" (Hiltz, 1990, p. 137). That kind of learning is not reformist to many adult educators, particularly those concerned with helping students develop self-responsibility and multiple dimensions of adulthood (Boud & Griffin, 1987).

It is fair to estimate that because of the early adoption phase of application of CMC in higher education, almost all the recent literature is concerned with three types of issues - implementation ones such as costs, access, and comparisons with face-to-face classes; teaching ones and technical ones.

The functions of CMC in higher education are now many:

...provide students with opportunities for convenient course-related or social interaction with peers... enable collaborative group work by distance students ... facilitate interaction with an instructor ... decrease turnaround time for instructor feedback ... allow students access to on-line resources, e.g., databases, library catalogues, and course registration ... and enable students to upload and download assignments and take on-line quizzes and tests ...(Wells, 1992, p. 2).

CMC is being used across many disciplines - engineering, management, computer science, education, humanities; in short, in the arts and in the applied, social and pure sciences (Romiszowski, 1992; Wells, 1992), and at undergraduate and graduate levels. CMC may be applied as an optional, adjunct or exclusive format. It may be used for various student-empowering strategies (Davie & Wells, 1991) and for group-based activity (Eastmond, 1992b; Harasim, 1989; Hartman, et al, 1991), including group or paired assignments, sophisticated and extended role plays (Davie & Inskip, 1993) and collaborative discussions (Feenberg, 1987; Harasim, 1990b; Higgins, 1992; Hiltz & Meinke, 1989; Paulsen, 1992).

Research studies report (i) analyses of instructional messages (Levin, Kim & Riel, 1990) and student messages (Mason, 1991; Winkelmans, 1988); (ii) applications of facilitation or moderation techniques and discussions of issues (Brochet, 1985; Davie, 1988; Davie & Inskip, 1993; Davie & Wells, 1991; Hiltz & Meinke, 1989; Kerr, 1986; Mason, 1991; Rohfeld et al, 1991); (iii) analyses of turnaround time for responding to student contributions (Hahn et al, 1990; Meurs & Bouhuijs, 1989); (iv) changes to appropriate teaching styles (Kaye, 1989; Turoff, 1990); (v) defining an appropriate personal and professional presence on-line (Feenberg, 1987; Harasim, 1987a); (vi) specific skills training (Phillips & Santoro, 1989); and (vii) the quality of class discussion (Harasim, 1987b; Marantz & England, 1992; Smith, 1988).

Reports on student use of CMC appear to focus on rate of participation (Wells 1992) and time - how students use synchronicity (Higgins, 1992), experience synchronous messaging (Bump, 1990; Siegel et al, 1986; Wilkins, 1991) and use asynchronicity (Davie, 1988; Davie & Wells, 1991; Gledhill & Dudley, 1988; Harasim, 1987a; Hiltz, 1990).

The research methodologies have varied from experimental designs to surveys, message analysis, small scale interviews and evaluation instruments (Eastmond, 1992a), but overall, more often fit into the positivist than naturalistic paradigm (Eastmond, 1992a; Wells, 1992). For example, Hiltz's complex studies on the outcomes of "Virtual Classrooms" and their associated variables (Hiltz, 1988a & b, 1990) used mostly quantitative data of student-based outcomes and some qualitative data from instructors. Her use of pre and post-course questionnaires for measuring student expectations and characteristics and for evaluating course effectiveness and student comparisons of CMC with traditional classrooms was based on her assumption that "we still think it is necessary to prove that it is just as good as a traditional classroom for mastery of facts and information" (Hiltz, 1990, p. 154).

Various claims have been made for CMC, and they tend to focus on elements of power and inclusion. Perhaps the most graphic image of the potential of CMC was given by Levinson as he situated CMC within the evolution of media:

If, as David Riesman said, print was the "gunpowder of the human mind"...we may be in for an unleashing of mental energies on an atomic-fusion level as a result of electronic creation and dissemination of text (Levinson, 1990, p. 10).

In educational settings, various claims have been made for CMC. Mason (1988), for example, has argued that CMC ought to facilitate self-directed learning as well as cooperative learning activities and a sense of synergy or flow in the activities. Others have promoted the transcript as a valuable record of learning, an antidote to memory loss, and a challenge for all to be accountable for their contributions (Davie & Wells, 1991). An educational change function for CMC has figured in some claims. Turoff, one of the pioneers of CMC, was convinced that "CMC...promises to bring about important changes in the educational process" (1990, p. ix). CMC would open up new educational domains and "unprecedented options for teaching, learning and knowledge building" (Harasim, 1990, p. xvii). It is a new domain of learning for Harasim because of five attributes: independence of place and time, many-to-many and text-based communications and the information and time control/organizational capabilities that only a computer supports (Harasim, 1990). Its applications in education may be undreamed of by CMC's original designers (Feldman, 1986, p. 74). CMC is regarded by some as more effective than the face-to-face classroom (Hiltz, 1986).

The focus of these claims are in part reflected in centripetal referencing, i.e., referencing inwardly to each others' ideas and case studies, rather than in centrifugal referencing, i.e., reaching out to a wide range of related disciplines such as cognitive and social psychology or adult learning, or referring to technologies related to CMC, e.g., audio or video conferencing. Such claims also reflect the enthusiastic early adoption stages and promotion of CMC's capabilities, without an extensive grounding in research of either positive or naturalistic paradigms.

Appropriately, however, the early hyperbole and claims are giving way to more cautious and balanced reflections. One example is the Higgins (1992) study of synchronicity within a comparison of asynchronous and synchronous messages as they affected cognitive activity, subjective reactions, learning outcomes and cooperative activity. Higgins's findings should draw greater attention to synchronous behaviour in CMC (as distinct from asynchronicity). Another example is the report of a study of gender and power relationships in a CMC discussion. Selfe & Meyer (1991) list the academic advantages of CMC as chiefly inclusionary - it may enable access or minimize the impact of cues about status, age, gender, etc that inhibit communication or encourage broader and deeper discussion. But their findings of a study of gender-based differences in message frequency and style showed that the men contributed twice as many messages as did women and were more verbally assertive - a situation often observed in visual classrooms. A further example is Harasim's developing focus on the key challenges to educators - how to promote the active learning and knowledge-construction potential of CMC. She defines active learning as active participation and construction of meaning, but then proceeds to use the additional concept of knowledge-building which she defines as the generation, linking and structuring of ideas (1990b, p. 55). She explains that while the CMC technology enables a rich generation of ideas in terms of quantity and methods, e.g., brainstorming, analogies, CMC is not yet able to promote the important converging processes of linking and structuring ideas because it lacks the software to organize messages into orderly sequences and conceptual hierarchies (pp. 57-8). She also indicates a further 'load' for the students: "The text-based environment is such a narrow bandwidth of information that, to compensate, clear and explicit articulation is essential for effective group interaction" (Harasim, 1990b, p. 49). Add to this the other limiting factors of lack of software to help organize the message into conceptually meaningful categories and threads, and what Harasim calls the "vulnerability" of the text - the public record capacity of the software - and the medium does indeed present challenges to students and educators alike. Harasim points to the "double-edged" quality of some CMC processes:

For example, increased participation and interaction can lead to information overload, both in terms of the volume of input and discontinuity of discussion threads. Related to this is the need to manage and focus the discussions, particularly within the seminar activities, to avoid 'on-line brainstorming' - a situation in which comments do not relate to and build upon one another (Harasim, 1987a, p. 133).

Harasim's colleagues are also becoming more cautious. Davie listed no less than nine "design constraints" for CMC: accessibility to computers and modems, the small-window (screen), keeping track of discussions and handling disjointed transactions, metaphors for the branch structure that confine strategic behaviour, negotiation and decision making, socio-emotional issues, fear of writing in order to leave a message, and poor keyboarding skills (Davie, 1989). Mason drew a balanced picture in her current conclusions about CMC in higher education:

The appreciative comments from distance learners about the benefits of easy access to colleagues, of discussion of course issues and of overcoming the isolation of learning at a distance, are certainly rewarding for those who work on implementing conferencing applications. However, some of the present enthusiasm for computer conferencing may be due to its novelty value and even to its exploitation of the current vogue for participatory education. Time will tell whether the benefits triumph over the difficulties and allow computer conferencing to make a significant contribution to distance education (Mason, 1990, p. 226).

Given the increasingly balanced reactions, what is known about students' actual behaviour and reactions in asynchronous CMC? Partial answers are available but they are based in analyses of types of messages sent and in lists of strengths and weaknesses of CMC. One report of a US implementation records that social relationships between students helped break isolation, that there was some discomfort in posting messages, and that only three people contributed the most messages (Gunawardena et al, 1991). Mason's study dealt with OU(UK) students who used CMC only as a very small proportion of their overall course activities. She consolidated the reasons for some but not extensive student disappointment with CMC as technical difficulties, irrelevant messages, lack of motivation to contribute, lack of a sense of spontaneity in discussion activity, the existence of peripheral comments that were broad rather than deep, and messages of poor quality (Mason, 1989, p. 14). Mason's later summary of the whole experience is cautious:

[CMC] is useful for broadening and enlivening the learning environment, but it is less successful at focused, in-depth analysis and delivery of core material for mastery... as part of a multi-media approach to distance education, computer conferencing in this application found its "unique niche - for communication amongst a wider community for increased feedback, for support and guidance, and for social contact (Mason, 1990, p. 237).

Other researchers found that graduate students were more active and sent more messages to each other than was the case with undergraduate students (McCreary & Van Duren, 1987). In Grint's data (1989), students had some difficulties with asynchronicity and lurkers (those who read but do not contribute). Haile's (1986) investigation into how CMC could support distance students showed that their messaging was about clarification, applications, extended thinking about course content, motivation, technical assistance and class management. Harasim (1987a) had graduate students who were very lively: they contributed 90% of course messages, stayed longer on-line than she had anticipated, and felt that CMC was an effective medium for study. Hiltz's reports on the " Virtual Classroom" project have also stated some satisfactory experiences; in particular, students perceived the overall quality to have been better in some ways than for the visual classroom, and experienced increased participation and access to the professor; (Hiltz, 1990, 1988a, 1988b). Davie & Inskip (1993) report great student satisfaction, despite some technical difficulties, with the process of a sophisticated role play they designed for a graduate course.

Studies at Bath University in England (McConnell, 1988, 1990) showed mixed reactions from students about CMC. In addition to the expected enthusiasms for CMC's convenience, some students indicated that they had wasted time, experienced undue formality in messaging, did not feel close to other students, and generally experienced less student-to-student interaction than in the face-to-face classroom. Rohfeld et al (1991) also found a time problem experienced by students: they spent much time downloading and printing the screen messages so that they could more easily read the messages. Brochet (1985) showed that students also gave mixed reactions.

Other student-related difficulties and issues have been recorded. Perceptions of information overload are a focus (e.g., Davie & Inskip, 1993; Harasim, 1987a; Marantz & England, 1992; Mason, 1990). Participation problems similar to those listed by Harasim are listed: no body language cues, difficulties in following multiple threads and in writing for the screen, and delayed feedback (Mason, 1990). Bump (1990) applied the term technostress as an umbrella for three problems for users of networked computers - the stress of having to learn at the same time the course content and the CMC software; the forest from the trees problem in information selection and organization; and handling the well known response of `flaming' when a message is felt as confrontational, sarcastic, or otherwise dysfunctional.

Some student perceptions of the strengths of CMC have been listed. Harasim provided a comprehensive list of strengths identified by students she studied (1987a) and advantages and disadvantages cited by students (1986). The strengths and advantages were the removal of time and distance constraints, round-the-clock access, the equalizing effect of reduced cues, on-line democracy, choice about type of response, great resources because of peer contributions, affective support from peers, the print record, and diversity of information exchange. The weaknesses and disadvantages were the loss of visual cues, group decision making, some technical problems, potential health problems and the high cost for going on-line in some geographical areas.

Several writers want to see research that will improve our understanding of CMC as it relates to cognition because, in Harasim's words, "we understand very little about the new phenomenon of learning in an electronic space" (Harasim, 1987, p. 119). Higgins (1992), Hiltz (1986), Marantz & England (1992), Newman (1990) and Winklemans (1990) have all referred specifically to the need to investigate cognitive tasks, interactions between students and teachers and the interpersonal elements in learning. However, none of these writers refer overtly to learning strategies as they are understood in taxonomies from cognitive psychology or adult learning. The next chapter details how this study responds to the general need in CMC in higher education for understanding about learning per se.

Elspeth Huxley's evocative analogy was borrowed by Worthen & Sanders to characterize a naturalistic approach to evaluation (Worthen & Sanders, 1987, p. 138), but it made a great impact upon me, and not just because it contrasts so well with Eisner's graphic image of researchers who charge into schools to "conduct commando raids" on unsuspecting students and teachers (Eisner, 1988, p. 19). The impact lies more in what the image suggests for a researcher's stance. Huxley (1959) in effect and Wolcott (1992) by directive argue for respectful and unpretentious strategizing in a stance of being alert and receptive. Wolcott, however, indicates that the security generated by alertness may go too far - to the point where the respectful and by now anxious researcher wants to capture everything that moves, so to speak:

No matter how tentatively you go about it, you must position yourself adequately to have a sense of purpose that includes some hunch about those data that may prove of greatest use... the real secret in descriptive work is not to gather as much data as possible but, rather, to get rid of as much data as possible, as soon as possible... (Wolcott, 1992, p. 44).

Between them, Huxley and Wolcott (Wolcott 1992, p. 43-44) form the boundaries of my experience with the methodology of this study. This chapter therefore details (i) my research approach - its goals, rationale and general characteristics, (ii) the actual procedures used to collect and analyze the data, and (iii) the contextual issues that influenced my decision-making and my experience of trying to "sit still", be "very patient" and rid the study's report of data to which I had developed unproductive attachments.

Criteria for a sound qualitative study have been discussed from two key perspectives: an adoption of the four conventional Campbell and Stanley experimental criteria of internal and external validity, reliability and objectivity (Goetz & LeCompte, 1984); or a new stance altogether that denies, on philosophical and procedural levels, the applicability of those conventional criteria (Lincoln & Guba, 1985) and argues for a new set of four criteria. Researchers in the latter stance use new terms for criteria. To meet the truth value issue in the internal validity criterium, they use the term credibility; to meet the applicability issue in the external validity criterium, they use transferability; to meet the consistency issue in the reliability criterium, they use dependability; and to meet the neutrality issue contained in the term objectivity, they use confirmability (Eisenhart & Rowe, 1992; Marshall & Rossman, 1989). Lincoln & Guba (1985) list the four canons of all research which establish the trustworthiness of any results: the extent of truthfulness and applicability to other settings, the assurances that the results do reflect the participants' perceptions (not the researcher's biases), and the results would be found again if the study were replicated (Marshall & Rossman, 1989, pp. 144-5). So specific criteria would include participant-checked, rich, thick and accurate descriptions of the phenomenon and its context (to meet the credibility criteria); triangulation of data by the use of multiple sources of data or different methods of analysis, and demonstration of theoretical frameworks used to guide data gathering and analysis (for transferability); explanations of the conditions of the phenomenon and how they change, and explanations of the evolving research design and any changes made to it as the research proceeded (for dependability); and evidence that the researcher gained acceptance by the participants, controlled for personal and conceptual bias in data collection and analysis, and generally provided an audit trail of procedures (for confirmability). These basic criteria are discussed in more practical detail in the different models of qualitative research (e.g., Fetterman, 1989; Glaser & Strauss, 1967; Lincoln & Guba, 1985; Marshall & Rossman, 1989; Miles & Huberman, 1984; Smith, 1987; Strauss, 1987; Strauss & Corbin, 1990). In his exploration of the relevance of qualitative research to the improvement of educational practice, Eisner uses only three criteria: coherence, insight, and instrumental utility (Eisner, 1991, pp. 53-60). Part of coherence is structural corroboration, or what others call triangulation, through the use of multiple data sources. Another form of corroboration is multiplicative (Eisner, 1991, p. 56) whereby the research results find agreement or consensus with other research results. The use of a study lies in its comprehensiveness, in how "things fall into place" from the complexities of the research setting. Usefulness lies also in the provision of descriptive, representative "maps" of the territory, and guides to what may be expected in future travels into that same or similar territory. The underlying question that summarized for me all the above criteria is one of authenticity: have I represented accurately the perceptions of the participants, taken account of my own biases, and produced descriptions and interpretations that are relevant to those perceptions? The following sections of this chapter detail how I addressed that question.

There was another issue in the interviewing process which initially surprised me but on reflection should not have. The issue was the small amount of information given about internal learning strategies that were not of the technological-procedural kind such as downloading files into a personal computer. Why were some interviewees apparently unable to talk in specific terms about what went on in their heads? The answer may lie not so much in terms of memory loss or memory distortion as in the expert student status of the interviewees. Some of their strategies must have become so routinized and automatic during undergraduate studies that they were not then available for verbalization in this present context. Garner points to this problem:

Perhaps the most basic concern is the accessibility of cognitive and metacognitive processes for introspective analysis...We all have had the experience of generating relatively vague, inarticulate descriptions of processing when called upon to produce them unexpectedly (Garner, 1988, p. 67).

I dealt with this problem by keeping the time delay between the end of the course and the second round of interviews as short as possible, using prompts during interviews, and asking for or referring to specific examples or situations in their learning. These tactics appeared to work.

The task of first level coding of the transcripts from the first round of interviews had to proceed under the time pressures of the three month courses and my need to have a second round of follow-on questions ready by late May for administration by mid June. The procedures used were largely guided by Strauss's list of basic work processes in qualitative analysis (1987, p. 17-20) and were similar to the practical advice on coding given by Kirby & McKenna (1989). These processes focus on (i) developing "generative questions" to help in proposing relationships and comparisons between ideas; (ii) coding that is not so elaborate that it bears little relevance to the practical world, but that is "conceptually dense" enough to promote complex linkages toward theory development; (iii) checking the developing theory during the subsequent processes of data collection and analysis; (iv) integrating the "dimensions, distinctions, categories, linkages" so that it depends on sound core categories; (v) writing theoretical memos to track developing ideas and increasing linkages; (vi) maintaining the non-linear movement between data collection, coding and memo noting for facilitating easy and repetitive access to old as well as new data, as needed; and (vii) writing the results which may demand the collection of new data or the filling in of gaps. Strauss (1987) recommends a four-element coding framework (p. 27-28) to help sort sub-categories and develop strategic linkages among categories. This framework helps avoid coding an event or process just as indicated in the transcript when that level of coding would be inadequate. The four elements - conditions, interactions, strategies and tactics, and consequences - I found initially useful as "thinking units" for organizing the in-vivo codes and later for the constructed codes. I tried other dimensional and procedural terms to help sort the draft codes until one set seemed to best accommodate the data (Lofland & Lofland, 1984, p. 71-92).

The first task for coding, apart from photocopying every page of my handwritten transcripts, was to split up the data so that all interviewees' responses to each question were in a separate folder for that question. Then a first scan occurred for every folder before a careful read-through to get a sense of the whole, to begin jotting down reactions and ideas for broad code categories, and to list my own emerging questions. In the third read, several days later, I generated in-vivo categories and noted them in the left margins of each response. I then deliberately left the coding, not just because I felt some doubts about how the large number of in-vivo codes could be converted into constructed codes of some theoretical density but also because I needed to get some distance from the mass of detail. That detail fermented (or so it felt!) in my head while I continued to jot down memos and questions about possible linkages and theoretical issues. I was quite amazed that during this kind of idea agitation, I generated 170 questions (brainstormed would be more accurate) which required their own procedures of coding and constant comparison (Groves, 1988) to develop dense enough categories to work up into questions for the second round. As with the earlier draft coding of the transcripts, I went looking for features in the landscape and "regularities - things that happen frequently with groups of people" (Goetz & Le Compte, 1984, p. 191). After playing around with matrices and Venn diagrams and recognizing that some questions had to be eliminated because they were repetitive, unanswerable, biased, or evaluative, I condensed the draft codes into 12 categories for exploration in the second round of interviews. These topics centred on asynchronicity, information management, reflective thinking, peer interactions, strengths and weaknesses of CMC, and comparisons with face-to-face classroom experience. The second set of questions, after the final approval, are listed in Appendix G.

After the second round of interviews (again transcribed by hand, which proved to be an excellent way to become immersed in the data), I photocopied the transcripts and again divided the data so that responses to each question were collated into their own folder. I assigned to each interviewee an alphabetic letter as a code name: some of the interviewees I found easier to talk with than others and I wanted to avoid bias in my interpretations of their responses. The contents of each folder were scanned and then read carefully before being coded later in detail with in-vivo codes. Then, after every folder had been dealt with, I again retreated and let things "cook" in my head for a while. Then the codes were re-read and collapsed into codes of my own construction. This process took much longer than I had expected and again resulted in a deluge of codes. However, little reduction actually occurred at this stage - my basic goal was to get onto file cards all the verbatim transcript extracts relevant to each constructed code category. Eventually, in fact, most of the data in each transcript was recorded on these cards (somewhat to my chagrin, but to be expected, given my learning style for fine discriminations in data).

I then divided the cards into 11 broad categories to represent the responses of 58 of the 66 interview questions. This division was not easy and I tried all sorts of thinking units for the division before deciding that I should keep to the relatively pragmatic and down-to-earth ones that originally guided the data collection for the second round of interviews. The other eight interview questions were not analyzed: either because the results were inadequate or the question was not asked (because it would have been repetitive) or it proved to be poorly worded.

The cards in each of the 11 categories were later re-read anew, but this time much more critically. The extracts were re-coded using fewer constructed codes and constant comparisons among the transcript quotes to get greater conceptual density. In general my approach here involved looking for key abstract and concrete concepts and making distinctions between them and among their subconcepts. Strauss (1987, p. 28-33) uses the more sophisticated terms of open, axial and selective codes to describe this analytical process.

Finally, in order to facilitate the organization and reporting of the findings, I collapsed several categories to create a final set of seven:

1. General Learning Strategies
2. Learning in a CMC context
3. The Uniqueness of CMC
4. Strengths and Weaknesses of CMC
5. Affective Reactions and Behaviours
6. Comparisons Between CMC and TC
7. Summary Comments and Images

Representative quotations from the cards on each category were then selected in preparation for the writing up of the findings. I later edited approximately half the quotations: in hindsight some quotes were repetitive and others added nothing substantive.

It is no exaggeration to conclude this section by saying that the data reduction and analysis took much longer and gave me more headaches around categorization than I had expected. Wolcott's admonition to "get rid of as much data as possible, as soon as possible" was also a difficult task. Indeed it proved to be cognitively almost impossible and affectively most unnerving, at least until I had completed several iterations of code construction. The findings are presented in the next chapter. In those findings will be found some of the evidence needed for meeting the criteria of soundness, or, in Eisner's term, coherence. In many instances, the same data were collected from multiple interviewees. Further corroboration of the findings is of the multiplicative kind, i.e., in the interpretive links between my findings and extant theory.

The other instructor worked differently, spending an hour on-line almost each day of the week.

That includes my reading their comments, thinking about it, perhaps referring back to the literature, responding when I needed to, organizing the course, starting it up, checking on things. I work typically on-line - unless there's a big debate discussion. If things are more in depth or really require some reflection, some concentrated effort for me, I'll print it out and work off of [sic] paper - at least once a week - as soon as there's a need to pull together some themes or to really study them, or something's controversial. It's easier on my eyes and I can draw on it or manipulate it (V:V13).

The instructors handled the parallel discussions in different ways. One took ideas that overlapped from one conference to another and linked them in order to show some integration. The other believed that dividing the course content discussions into thirteen simultaneous conferences not only legitimized the social kinds of talk (in, for example, a "Coffee House" conference), but also allowed the academic content to emerge in uncluttered and focused formats.

One instructor judged the level of academic discourse to be mixed because the students were very mixed in their interests and ability to handle theoretical studies. The other instructor judged the level to be generally as good as in the face-to-face equivalents (people came prepared), but felt there was something affective missing:

The excitement and spark... when folks are all of a sudden in-tune on the issue (W:W28).

The final grades given by both instructors were similar to grades given in face-to-face sections of the same course or in other courses taught by these instructors. Students in each course had to complete three assignments for their grade, with one assignment being a major integrative and reflective piece. A certain portion of the course materials fee was set aside to cover the costs of the Datapac transmission of messages from beyond-Metro Toronto students into OISE's central VAX computer. However, as both instructors noted with some satisfaction, the students diligently downloaded their messages and worked off-line to avoid extra charges.

The instructors explained that their interventions in class discussions were for discussion management and contributions of content. One would intervene only to answer direct questions, coach for greater participation, help people link ideas and give summaries. By the end of the course under study, this instructor was contributing just content because the students were judged to be managing the discussion effectively. The other instructor played a reduced role throughout the course because of personal values about teaching, and having students who were all experienced CMC users doing most of the conference moderation (chairing).

The second way was the incorporation of new learning into an existing framework:

Oh dear...[pause] I guess when something, when an idea makes sense to me, and I start relating it to other things that I already know and I start putting pieces together, putting it into framework, or that is helping me to establish a framework for a larger body of knowledge (LN:I.1.4).

What I think about it whether it's correct. "Ah yes, that fits with what I've learned before or that fits with my experience" (J:I.1.4).

Or, third, the new learning could be applied:

When I can apply it in a practical setting, when I can use it (G:I.1.4).

Most interviewees did not plan any specific cognitive learning goals for their courses. The goals most often cited were to react to others' messages, to complete the assignment and to read the notes on the screen. Other goals were reading the texts, planning study time, prioritizing reading and logging on. Several pointed out that they did not have any learning goals, while others had pragmatic goals:

Find out what is said and react to it, or read some of the readings and comment on them. There's no strategy (EN:I.2.2).

Since part of the context was graduate level learning, people were invited to define it. Regarding their definitions of graduate level study, seven people explained that it was the taking of personal responsibility for learning:

Taking control of my own learning. I'm doing it for myself, make me a better person (J:I.1.3b).

Think at a higher level - to be able to assimilate ideas - the "So what do we do now and what effect is it going to have?" (S:I.1.3b).

Other definitions focused on using their own experience, having higher levels of motivation than in undergraduate courses, articulation of ideas, and personal growth.

Reactions to a question about preparatory activity for learning indicated that the concept was not very meaningful for most interviewees, so I did not probe for details. Some did not do any preparation at all, but others would do one of several things: decide on a contribution, review past notes, or gain peace and quiet in their households.

The metacognitive issue of monitoring learning also did not meet with widespread responses of recognition. Almost half the interviewees said that they did not monitor their learning, and indeed the question appeared to surprise and puzzle almost everyone. Some strategies were offered though - the use of external rewards such as grades, occasional reviewing of their draft messages, keeping a learning journal, noticing feelings, setting own goals and checking accomplishments. I did not probe extensively when given clear cues by many interviewees that the question was of little or no importance to them.

I sift [new information] around with other ideas I've got before I finally make up my mind what I think the person is after;...let 'em float around in your mind, compare 'em... (K:I.2.3).

The expression of new learnings involved naming and labelling ideas, making them more concrete for verbalizing, and claiming the new learning, sometimes in the form of a published message to another person - peer or instructor. Insights were construed as one form of inner expression:

Once I understand it, I've got it - the `aha', things start falling into place (EN:I.2.3).

Other strategies for making sense that were mentioned less often than choice were to underline printed-off notes, keep up with the flow, be focused, and bring personal experience to bear on the analysis of new information.

I underline if I agree - make little notes in the margin - how this relates to how I've learned or things that are important (S:I.2.3).

These strategies were confirmed later when interviewees discussed a particular learning experience. Mentioned most often were the strategies of getting an insight, examining the relevance of ideas to their own personal experience, managing the incoming messages, and having their learnings confirmed by others.

In summary, interviewees made sense of incoming, on-screen information in several ways. They chose relevant messages because they appeared to be new, applicable, relevant to prior knowledge, or because the student had to set limits on the quantity of messages. They organized the ideas of others by filtering, sorting, and linking. They expressed their own thinking-in-progress and thinking-that-was-concluded by naming and concretizing ideas, presenting them to a peer, instructor or self, or experiencing an insight.

A number of skills were mentioned and I categorized them as operations, information processing and management of stress and frustration. The operational skills were listed as operation of the CMC software, reading, writing, decision making, and group interactions. The information processing skills seemed to echo an affective need to stay in control. The skills were choosing a focus of attention for dealing with messages, handling the parallel nature of conference branch discussions (several topics being discussed at the same time), and developing a personal system of managing all the messages (including finding the common threads in discussion). A common strategy was the general approach of processing information quickly so as to keep up with the flow of incoming messages. One unsolicited metaphor captured the essence of managing the complex information environment:

It's a difference between a single rope and a loom in a sense... you have got to be able to sort of switch from one line to the other and then keep those lines from being tangled (G:I.2.4).

Other skills mentioned, but less often, were the filtering and synthesizing of ideas, deciding which incoming messages would receive attention, and working successfully in a small group.

The responses required of peers would involve giving constructive feedback, answering questions, not being repetitive, being responsible generally in small group work, complimenting peers, and engaging in the content of the messages. Giving timely and helpful feedback was mentioned specifically. The giving of affective support by peers would involve use of a person's name, helping people belong, being patient, complimenting others, and providing a climate that is "sustaining and confirming":

I feel [in learning to use the CMC software] the same as a six year old learning to tie laces... so you couldn't ever be too encouraging (DN:I.2.6).

Acknowledge what the others have said before you take off on a new thought...I guess along the same lines, it's support: you know [that] you're still part of the functioning group (F:I.2.6).

The production by peers of short, focused messages drew comments of some feeling:

Not [to] feel [that] you have to say absolutely everything that comes into your head to comment on..(J:I.2.6.).

...don't sit there and jabber (F:I.2.6).

In CMC... you can ignore dominant people that you consider to be useless (QN:II.18).

These behaviours required from peers in any CMC context - participation, response, creation of conditions for affective support, and the production of short, focused messages - were echoed later in their reasons for not participating in CMC discussions. The majority of those who could give answers (only nine did) referred to the inhibitors of irrelevant or repetitive discussions, or messages that were either long winded or unfocused, or were seen to have been written more for the benefit of the writer than for any recipient. Other interviewees named unplanned or unavoidable external-to-course events or feeling excluded, bored or frustrated, and generally getting behind or even out of the flow of the class discussion.

Dysfunctionally divergent peers, poor message threading and a focus on one's project to the exclusion of anything else were also mentioned as factors in feeling out of synchronization.

Various strategies for keeping in sync were reported by seven interviewees. They chiefly focused on deliberate selection of messages for more detailed attention, or on keeping up with incoming messages:

Just stay on top of the notes (U:II.5).

I picked and chose a little more carefully (G:II.5).

There were two perceived advantages of asynchronicity - the opportunity to choose messages for closer attention, and a feeling of coping with the load of information. Choice, by far the most often cited, operated in several dimensions: the timing of logging on to the computer, location of their work station vis-a-vis job or home location, content area to be studied, organization of messages, and freedom to reflect before composing and responding.

I prefer to work... after I've had some time with my family and after I've got all my other things out of the way I don't have to worry about. I was able to work at my time and leisure when I wanted to do it (F:II.6).

Interviewees could "take the dialogue away" while they gathered content-relevant resources; they could join or ignore a discussion - to "go wherever my head is at a certain point"; and they could use the transcript facility for a delayed response, or just take the time to think through a response and organize thoughts: "[take] considerable time if I needed" (C:II.6). Five used choice to do reflective thinking.

The second advantage of asynchronicity, mentioned much less often, was affective in nature: asynchronicity helped in coping with and getting space for managing the information:

It gave me a chance to juggle several conversations at once: it stretched out time for me (TN:II.6).

...allowed me to get the most out of processing what was going on in the course (QN:II.6).

These advantages, however, were not listed unconditionally by all the respondents. Several people referred in passing to the trap of getting behind with the class flow, but one interviewee was particularly articulate. He did not see his behaviour as asynchronous because he believed that the apparent freedom of asynchronicity was in effect balanced by the limit of discussion topic relevance; there were times of opportunity to contribute, but these did not last for long.

I said [to classmates] a number of times - it's asynchronous, but it's not atemporal; you can't be out of step with [class discussion]... even though the medium is advertised as being asynchronous. It doesn't mean that time is not important. There is a time factor, a window within which you have to be involved (P:II:6).

The disadvantages of asynchronicity were felt to be of several types: the pressure to log on regularly and often to keep up with class discussions, the feeling of information overload, and the sometimes self-imposed exclusion from discussions as a measure to keep sane.

Overload just kills the whole thing because you're no longer in the conversation -the conversation has already taken three turns since you last spoke (R:II.6).

Everytime you log in, your expectations decrease a bit - if it takes six or seven times before you get a response, by that time you kind of forgot the comments that you made (NN:II.6).

Certain peer behaviours, which affected almost half of the interviewees at some time or another in various dysfunctional ways, were seen as another resultant disadvantage of asynchronicity. What I have named the `loss of contributive energy' was felt by some interviewees as a disappointment in their peers and a disinclination to send messages to them because they had failed to send messages that were either timely or relevant.

Other cognition-related disadvantages of asynchronicity were listed as difficulty in synthesizing ideas, repetition of ideas - "while I was making up my message, other people had written the same thing" (S:II.7), and the fear of missing some idea of value:

There's a certain discomfort with realizing that there are all these different conferences and people are plugging into them at different rates - there's a bit of a terror there that, oops, I'm going to miss out on something. So I think it takes a fair while to be comfortable with that (LN:II.6).

G's assessment of asynchronicity hinted at his need for what I term cognitive synchronicity and the potential for time freedoms to become time tyrannies:

If the pattern within the asynchronicity that you can set for yourself does not match the group pattern, then you feel left out and you have to fight to get on as often as they do, or within certain limits - in order to get back into the flow (G:II.7).

In summary, a majority of interviewees, 14 out of 21, said they felt out of sync often or sometimes, and that this feeling was caused by infrequent logging on, the intervention of unexpected, external-to-course events or the experiencing of negative responses because of off-topic messages or unhelpful peers. Additional reasons were irrelevant topics, technological difficulties and confusions over the content. Ten out of 21 interviewees assessed asynchronicity as in some way not contributing to their learning. Mostly they experienced problems in getting feedback, but they also experienced difficulties in synthesizing information in the messages, coping with repetition of ideas, and feeling time tyranny and some anxiety that good contributions might be missed. However, two chief advantages of synchronicity were acknowledged: opportunity for choices of various kinds and positive feelings of coping with the messages.

Several notes of caution regarding reflective thinking came from three people. DN thought that "I feel like I did a pretty superficial job...it was pretty scattered reflection - a function of the many possibilities" (DN:II.17). TN wondered if he could have produced more reasoned thinking:

I'd like to think [my responses] were more reasoned but when I look back over my notes they're not substantially different although I feel like they should be... It seems to suggest that the answer's in me, and whether it takes a long time or a short time or whether I fancy it up...it's still the same message (TN:II.17).

Most interviewees, however, rarely re-read old files.

The final features mentioned were linked: the general excitement and novelty of using such a different medium and "the writing aspect" of CMC.

Typical of the explanatory comments on the strength of reflection was this one:

The chance to look back and reflect, question, link back to previous learned material, create new concepts, new ideas based on two to three past experiences (O:I.2.13).

The operational strengths were seen as CMC matching the individual's preferred learning styles, the availability of choices for and independence of action, the transcript, and finally the opportunity to locate the computer in a physically convenient and comfortable place.

I do tend to read and have had a loner style of learning. [CMC] plays into that very much (QN:I.2.13).

I enjoy working at the computer. I get a kick out of manipulating things with the computer (C:I.2.13).

The strengths of peer interactions came from the giving of help or from thorough and critical feedback, the human contact, and the different perspectives because of people's range of experience:

Being able to interact with other people electronically; there's something quite special about that. You're able to put an idea down and get a comment later. It's like receiving personal mail (IN:2.13).

And you get better feedback - there's the opportunity for more than one person to give you feedback, so you get different perspectives on things (P:II.26).

Strengths less often mentioned were software and learning skill development, fascination with the technology per se, the opportunity to be "very organized" (S:II.26), feeling "obliged to try to take on information and assimilate [it], focusing on different senses" (TN:II.26), and the succinct nature of the messaging - "not as much garbage in the way of what they're trying to say" (F:II.26).

As a matter of interest at this point, I asked for estimates of the extent of discussion fragmentation. Discussions were seen by half the interviewees (10 out of 21) as "somewhat" fragmented, and by five as "a little". The linking of individual messages also was not the norm: ten students talked about no linking at all.

Two key coping strategies for these information handling problems were suggested. The first focused on choices around filtering out unwanted information - "selective neglect" (QN:II.14c) and keeping what appears to be useful information - judicious selection (my term). Twelve interviewees made deliberate choices such as scanning through peripheral conference branches, but there were some feelings of anxiety that some useful information might be lost, or, that after downloading and printing off a batch of notes to scan at the end of a week's activity, the opportunity for a timely response would have been lost:

At first... I didn't want to miss that possible gem that would be slipped in that mound of dirt. Now I've proved [that] there's no gem coming [out of] there (O:I.2.1)!

The second coping strategy was to download the information on screen into their own files and then browse through and organize it at their own convenience.

I'd just go in there and download everything and come back out and then start cutting and pasting (in the computer) and shoving the stuff in number order into different conferences. I found that hard copy is easier to ingest than screen copy - no idea why (H:II.14c).

Discussion fragmentation was seen as a weakness by two thirds (14) of the interviewees. This sentence sums it up:

They're just disjointed - it's not like reading ...an expository paragraph; you need to know what you're looking for and then piece it together (IN:II.12a).

The coping strategies focused on keeping up to date with the discussion, downloading messages into personal files, keywording ideas in messages, producing a paper transcript, and scanning the messages on-screen in one concentrated period of attention.

The Principle of Recency (O:II.12).

Be on-line a lot and be able to assimilate... I was dealing with relatively small amounts of information and plugging it into my own thinking and cognitive structures (P:II.12).

Time problems, the final, most often cited weakness of CMC, were seen to affect information processing and management. The problems were shown in the effects of delayed responses to others' messages, in the demands for processing information speedily, and in the desire of some interviewees for real-time interaction with peers and with their instructor.

The time delays - it's a strength in so far it gives me time to reflect and think about things; on the other hand... a weakness because things pass me by almost too late to jump in and react (EN:II.27).

Nine comments referred to software deficiencies that hindered the organization of information and possible solutions:.

A hypertext system...[would] help you organize your material, open [some] windows and you'd be able to move information around and perhaps run searches in different windows (H:II.25).

All the elements and conditions in the two CMC courses which emerged as features, strengths and weaknesses are summarized in Table 4-4. It is interesting to note that time and peer interactions are perceived as both strengths and weaknesses.

One common thread in comments was the need to keep up with the incoming messages. One interviewee used a graphic image of a weaving loom to represent the skill of managing all the branches in the computer conference:

It's a difference between a single rope and a loom in a sense... you have got to be able to sort of switch from one line to the other and then keep those lines from being tangled (G:I.2.4).

Some experiences of stress about the information load generated some graphic metaphors. Several metaphors referred to the volume and speed of information:

Every time I logged on it was like `Here comes the wave!' ...I could see myself trying to build the castle before the water comes (R:II.14c).

Another metaphor referred to a self defence mechanism that one interviewee felt was a useful way to reduce the inhibitions around contributing to discussions:

You have to develop an electronic skin [in order to] just jump into the activity and not reflect too much on the writing (IN:I.2.4).

An engaging metaphor captured the feelings of one interviewee who had to learn the PARTI Software and rely on peers to speed that learning:

I feel... the same as a six year old learning to tie laces... so you couldn't ever be too encouraging (DN:I.2.6).

The final metaphor referred to managing a feeling of potential loss of information:

I didn't want to miss the possible gem that would be slipped in that mound of dirt. Now I've proved [that] there's no gem coming [out of] there! (O:I.2.1).

The most dramatic metaphor of all did not emerge as a response to my direct question about metaphors. It reflected one person's feelings of continuing pressure felt from the speed of the incoming messages over each 24-hour day of each of the 13 weeks of the course:

...a huge river flowing... it's like riding the rapids, you've got to get in a boat and you've gotta keep going - you have to keep that boat moving along the river of information because if you ever stop you're going to go down (G:I.5.1).

A.2 Active Study

• Work through instructor materials X
• Underline X
• Take notes X
• Summarize X

B. Support Strategies

B.1 Metacognitive

• Plan
• Attend X
• Encode
• Review
• Evaluate own progress X

B.2 Preparation/Execution

• Relax X
• Reduce anxiety X
• Manage concentration X
• Manage time X
• Set progress goals X

Three points relevant to Tessmer & Jonassen's taxonomy are worth noting here because if I have assumed correctly they may help to explain the interviewees' focus on information management. First, Tessmer & Jonassen assume that the print-based learning materials provided to students by teachers and librarians show certain levels of conceptual organization and clarity of expression because they have been edited and prepared as learning tools. In the two CMC courses studied, the instructors do provide a structural, logistical, macro-level organization to the messaging, but micro-level conceptual organization and clarity is a function of the actual written interactions as they occur. Each instructor is only one writer in a multi-writer, asynchronous context in which differences in attendance at the computer screen, in attention span, in focus of interest, in software skills and in learning styles may aggregate to produce an almost anarchic collection of messages. The results indicate that interviewees experience the fragmentation of discussion with some feeling. Second, Tessmer & Jonassen are implicitly referring to face-to-face classrooms where the expression and regulation of speech is facilitated by all the paralinguistic cues of vocalizations and body languages and the speech conventions of turn-taking and opening and closing phrases. In the CMC context of this study, the paralinguistic cues are lost and students have to communicate silently and with some uncertainty about getting feedback on ideas that nest in a body of often fast-moving and unthreaded text. Third, there is no reference in Tessmer & Jonassen to the time sequences and scales associated with human activity (Hesse, Werner & Altman, 1988; McGrath & Kelly, 1986). This omission is not unusual; McGrath & Kelly explain how most social psychologists have not seen time as anything much other than something to be "filled" or a part of the context that is stripped away, i.e., it is not a real variable to be studied in and of itself. So the taxonomy of Tessmer & Jonassen is "timeless...in the worst literal meaning" (McGrath & Kelly, 1986, p.5).

Two further observations are appropriate. The first is the concern shown by interviewees about their conditions for learning which I categorize as relational, affective, logistical and cognitive (listed in Table 4-6). The conditions were discussed earlier in relation to inclusion and control (Schutz, 1966) and to the basic drives to action of achievement and affiliation. Tessmer & Jonassen do not appear to take into account the contextual conditions I categorize; their focus appears to be on the individual student with no references to strategies for peer learning in group contexts or in contexts where teacher control is less overt. Merriam and Caffarella (1991, p.178) stress the importance of context for adult learning in terms of the social and cultural influences on adult cognition: the life tasks and problems that require cognition to fine-tune and re-structure existing schemata. Such fine tuning and restructuring is not done well by adults under the same conditions as would apply for young students who are still building their own schema. Dansereau's work (1988) may provide some help in understanding how the mix of contextual conditions helps in cognition. He cites research (albeit with young students) in which peer behaviour was seen to help in two goals of learning: acquisition of information and its transfer into other settings. Acquisition was helped by what Dansereau defines as the metacognitive activity of questioning and by general discussion among peers that helped the student correct errors in misunderstanding, select key ideas, and detect gaps in knowledge. Transfer into other settings was helped by peer use of elaboration strategies - metaphor, imagery and applications to other contexts. So questions that may arise for adult students in CMC contexts include how peer behaviours might help in the person's fine-tuning and restructuring of schemata, and how those behaviours might affect the use of an existing repertoire of learning strategies and promote the development of new ones. The results of this study allow me to suggest that peers are very important for several of the Tessmer & Jonassen learning strategies. Peers provide different perspectives on knowledge; their contributions could be construed therefore as an elaboration strategy. Peers also give feedback; this process could function at the knowledge integration level if peers give examples or paraphrase, or it could function at the levels of knowledge organization and elaboration if peers generate implications or inferences, provide cognitive `hooks', or trigger transformations of public knowledge into personally meaningful schemata or conscious insights. Only a transcript analysis and analyses of student protocols would provide valid answers to those questions. The influence of peers was certainly relevant to the interviewees' use of the preparation and execution strategies of Tessmer & Jonassen that are designed to "create optimal psychological conditions".

I would add therefore to the Tessmer & Jonassen taxonomy a new group of strategies named "III. Meta-Context Management". The meta-context is the set of interpersonal dynamics and the individual student's sense of purpose and presence in those dynamics. Key strategies for handling those dynamics have to be based on knowing when to be fully involved so that cognition is enhanced and when to withdraw gracefully but temporarily so that guilt or isolation is reduced while the student re-groups to consolidate control or explore a tangential topic. Strategies for involvement would include community building, negotiating personal and group goals, establishing behavioural norms and standards for message threading, claiming attention, exercising choice, giving support, resolving conflicts, and expressing thinking-in-progress, insights and conclusions. Strategies for temporary but graceful withdrawals from class discussion would include giving notice of that intent, letting go of messages or transferring them to an archive, doing intensive reflective thinking. Strategies for re-entry into the class discussion would include accurately analyzing the state of cognitive synchronicity of that discussion and deciding about a focused contribution.

The second observation is related to the conditions for learning: it is the emphasis on management of information. Tessmer & Jonassen refer to preparation and execution strategies but the findings of this study indicate that information management is concerned also with complex strategies for maintaining feelings of control, inclusion and self-competence. Indeed, the "self", in terms of self-esteem has a stronger presence in the interviewees' strategies and conditions than appears to be the case in the Tessmer & Jonassen taxonomy. This omission is not so surprising given that most of the work on cognition and learning strategies concerns children whose learning motivations are of an entirely different quality than those of adults (Merriam & Caffarella, 1991, p. 159).

Now we realize much more that whether students are learning is a function of what they want to learn; not just the value they place on the learning experience but also their expectations about the chance of success if they devote effort to learning (McKeachie, 1988, p.10).

Interviewees' self-competence extends beyond being a skilled manager of time; it has more self-responsible connotations than Tessmer & Jonassen's younger students would be able to show. For the learning in this study, the self of the student is engaged in meaning making (as the cognitive psychologists suggest) but that process is concomitant with strong affect and efficiency considerations, namely, the maintenance of self-esteem and self-concept and the need to see study time as efficient in process and productive in outcome. What Duchastel calls the "effin factor" appears to be at work here: the effort-to-interest trade-off that operates when a student has to make judgements about what efforts are justified for which information (Duchastel, 1990, p.138): "The effin factor characterizes the intrinsic motivation of the student, otherwise known as epistemic curiosity". All these issues have to be met in an environment that is not only interpersonally complex but lacking in the aural and visual cues that guide assessments of peer behaviour and intentions. It is appropriate therefore to discuss this environment in more detail, with regard to the two characteristics of CMC that separate it from audio-conferenced or face-to-face classes. Those characteristics are the reliance on written text for communication and the event of waiting.

(C. Day Lewis, cited in Bereiter & Scardamalia, 1987, p. 22)

Emig's research led her to argue (1977) that writing is important to thinking when it is integrative, connective, active, self-rhythmed, and capable of revision and review. Biggs summarizes prima facie reasons for how writing helps to clarify thinking: it "externalizes thought...[giving opportunity for] reconsidering...or revising", it demands clarity because of its distance from the cue-rich context of face-to-face communications, it demands considerable cognitive activity, and its expression calls for attention to the audience, not the author (Biggs, 1988, p.188). In a later and indirect form of support for that argument, Bereiter & Scardamalia (1987) explain two models of writing to learn - the knowledge-telling model and the knowledge-transforming model. The latter model is appropriate for graduate-level learning because it involves higher order thinking skills and the re-working, shaping, integrating and transforming of the bits and pieces of information that arrive in the conscious mind in a rather scattered way but which appear to have some relevance - "driblets" of information in Aldous Huxley's term (Bereiter & Scardamalia, 1987, p. 10), or "nuggets of knowledge" (Whalley, 1990, p. 62). The transforming process may involve changes in the expression of that information as well as in the information itself. Bereiter & Scardamalia argue that two problem-solving spaces are therefore set up in the student's mind: one is for problems about the information (the knowledge space), the other is for problems about the expression of ideas (the rhetorical space). The student/writer oscillates or moves back and forth deliberately between the spaces as she/he thinks about the driblets and copes with various problems of expression and relevance. This model makes great sense to me personally because it named the way I work through driblets to get an insight or a feeling of coherence. Indeed, almost all my useful incoming information is formed as driblets!

CMC messages are full of driblets - and dross too. The student therefore has to select and separate out the dross, at least at a gross level, before beginning the processes of knowledge transformation and reflective thinking. References to reflective thinking in the interviews drew the fastest comments, so I suspect that the knowledge-transforming model may prove fruitful for research into current conceptualizations of cooperative learning and into CMC learning strategies. Certainly the model appears useful in terms of investigating the students' qualitative conceptions of how they learn. Researchers would have to take into account one factor in understanding the model. Bereiter & Scardamalia stress that the key issue in using the model is the number of simultaneous demands on a person's attention. What happens in learning strategy terms when students' short term memories have to deal with more than the recognized five or seven chunks of information? Knowledge-transformation in the CMC context has to be accomplished at the same time as coping with the ongoing and fast-moving flow of driblets and dross.

Biggs' research on writing appears to be a fruitful base for further research in CMC learning contexts. Keeping the Bereiter & Scardamalia content and rhetorical spaces and the two-way reflective movement between those spaces, he developed a model of essay writing that includes three aspects or stages in writing: intentions - feelings and orientations (deep or surface), the creative processes of parawriting - planning and composing, and the analytical processes of the actual writing - writing a draft, reviewing and revising, with their demands for reflection, critique, and correctness in spelling, syntax, lexicon and style (Biggs, 1988). In a CMC context, with its often fast pace of incoming messages and the need to give thoughtful and public responses in a relatively short time, the student may experience quite a load of pressure, especially if she or he is a reflective type of student.

The concept of cognitive load has recently been discussed in the cognitive science literature (Chandler & Sweller, 1991) and it appears to be relevant in understanding the qualitative nature of student load in a CMC context. The Chandler & Sweller definition of cognitive load refers not to quantity of information but to tasks which are ultimately dysfunctional but have to be carried out initially in order to clear the ground for the important tasks of cognition; e.g., "if students unnecessarily are required to mentally integrate disparate sources of mutually referring information such as separate text and diagrams" (Chandler & Sweller, 1991, p. 293). The results of their study around the load involved in the composition of messages for permanent and public display have similarities with a smaller study conducted with some Open University (UK) students who took the same CMC course used for Mason's research (Grint, 1992). Grint reported that writing messages involved four aspects - composing the text, analyzing its content, rendering that content "absolutely unambiguous" and making "explicit as much as possible" (p. 165). Chandler & Sweller's definition of cognitive load may therefore have to be modified for CMC to account for the quantity and fragmentation of incoming information to process within time limits and the feelings of pressure to produce grammatically correct and substantive written communications. A changed definition of cognitive load might also refer to attempts to develop conditions for the creative "volatility of conversation" (Grint, 1992, p. 163) without producing prattle or what one Open University student called "chewing gum for the eyes" (Grint, 1992, p. 160).

One way of mapping and explaining the impact of the stressors may lie with the Power-Load-Margin formula (McClusky, 1970). The formula shows a margin that gives an adult surplus power if the power available to cope with load is greater than the load to be carried. Power to cope with load comes from internal resources, e.g., the knowledge and skills from life experience, and a repertoire of appropriate learning strategies, and external resources, e.g., health, social networks, etc. The load to be carried comes from the demands of intrinsic and extrinsic goals and drives to action, the external demands of adult life role responsibilities, the attributes of the learning tasks, and the qualities of the conditions in the learning context. Main (1979) stressed how load and power are related in learning contexts and how they determine the role of the resulting margin. If the adult is to remain psychologically healthy and have the energy, time and space to learn, the ratio of load to power has to be such that a positive margin exists for investment in and maintenance of the work of learning. With my interviewees, the continuing presence of such a margin cannot be assumed. There are bound to be times when there is a negative margin because of increased load. That load may come from a variety of sources and act in different ways. Novices, for example, have to learn both software and course content at the same time at the beginning of the course. The seven stressors listed above constitute a load to be carried, so also may be the results of a failure to use strategies when they were needed (Garner, 1990). Other sources of load may be linked with some of the existing research on reading. The somewhat fragmented nature of the messaging experienced by many interviewees reduces predictability in the discourse (Kirby, 1988, p. 254). Even good readers may suffer reduced processing ability if the capacity in their working memory is used up in coping with the fragmentation of ideas in the messages. And Biggs' research would appear to reinforce this suggestion that under these conditions, a student's ability for information processing at the `deep' level is reduced. Biggs (1988, p. 237) points out that the generation of thematic meaning from text (deep processing) happens at the highest levels of analysis and can be carried out only after successful completion of the lower levels such as generation of meaning from early ideas. If those early ideas are presented in rather fragmented forms, then an extra load is expected. There will be times during a CMC mode course when a negative margin creates a need for a temporary withdrawal or the use of exceptional measures to control information.

Figure 5-1 maps the power and load forces from the interviewees' perceptions of the strengths, weaknesses and features of CMC. The figure would appear initially to contain enough power resources to ensure a safe margin. But my second thoughts and intuition indicate three phenomena that may influence how the power and load forces operate. One phenomenon is the asymmetrical nature of the effects of the medium. Examples include: one person's time flexibility to respond at will is another's time delay when a fast response is wanted; one person's freedom to write as much as desired is another's oppression from the resulting information overload; and one person's convenience of computer access at home is another's need to get away from home responsibilities and relax with peers to rekindle energies. Another study would be needed to display the power of each effect and the operation of the dynamics of each antinomy.

The second phenomenon that influences power and load forces is time - the time taken to become proficient in using the CMC software to the point where power can be used to control information. The interviewees who were novices to CMC said that for the first month of the three month course they experienced some stress from having to learn CMC skills, as well as learning the course content. Other interviewees may not have had adequate skills, and certainly the software has its limits.

The third phenomenon concerning power and load forces is the process of change and adaption from learning in a face-to-face class to a text-only environment. The CMC context may be experienced as fast-paced because of the message frequency. It may also be experienced as difficult to regulate because of the loss of aural and visual cues. Inadequate data exist in this study to allow for any comparison with face-to-face contexts. I did play with some gross word counts as a way of comparing the extent of messaging in CMC and face-to-face learning contexts. Say, for example, that the average CMC message length is 10 lines with 13 words per line: that is 130 words per message. In course #1, having 1,629 notes (refer to Table 4.1), the estimate here is 211,770 words. In course #2, having 1,797 notes (refer to Table 4-2), there could be 195,000 words to process. Would a face-to-face context show as many words? The calculation here has to be different. The normal rate of speech is taken by audio experts to be 125 words (approximately) per minute (Barr, 1992). Subtracting 20 minutes for the refreshment break in a three hour class, one could estimate that in a 150 minute (total) time of actual talking (excluding silences, for which I have allowed 10 minutes over three hours), the number of words spoken could be approximately 18,000. Over 12 weeks of a course, that number would become 216,000. These numbers match reasonable well with the CMC contexts of this study, but they represent a very simplistic approach. Any adequate analysis would have to distinguish the functions of the words, e.g., social maintenance or sophisticated cognition. So the students in the CMC context have to use all their sources of power to create a helpful margin for contributive energy. How are these sources developed? I suspect that the margin for contributive energy in CMC is created more by load reduction than by power increase. Load reduction may be achieved in several ways. Interviewees choose to selectively neglect some information in order to keep it to a manageable level, to use "electronic skins" or other means to reduce losses in self esteem, and to generally work hard to keep up with the flow and contribute. Another load reduction mechanism deals with the load of frustration: it is to reduce expectations of helpful and timely peer behaviour - expect less, get less and therefore work less. Here the "effin factor" is at work again.

II. Support Strategies
Develop and maintain "a good internal state".

C. Metalearning strategies
To monitor cognitive processes and to respond to various learning tasks with appropriate strategies

C.1 Plan
- select goals
- activate memory schemata
assess depth of processing required
- estimate processing demands for the task
C.2 Attend
- focus on instructional materials
- search memory for relevant information
- contrast new information with memorized information
- validate new information by confirming with old
C.3 Encode
- clarify and elaborate new information)
- Refer to Information Processing Strategies in section A
C.4 Review
- confirm correctness of learning
- repeat new learning via recall and practice
- revise as needed
C.5 Evaluate
- assess clarity and coherence of instructional materials
- determine the relative importance of new information

D. Preparation and Execution strategies
To create optimal psychological conditions
D.1 Preparation strategies
- establish mood
- relax (for concentration)
- reduce anxiety (about new information)

D.2 Execution strategies
- manage concentration
- manage time
- set up goals to assess progress of learning

1.1 Goals:
Use observation and analysis to focus awareness; use questions to hypothesize and demonstrate; use internal debates to understand and revise learning.

1.2 Key elements:
Perceiving - see, listen, make sense of context
Analyzing - recognize elements, principles and relationships in situations
Proposing - think own thoughts, assess evidence, develop propositions
Imaging - create new ideas, images, solutions, procedures, realities
Reflecting - assess results of above activity, decide about possible revisions to thinking.

2. Emotion

"...responding with feeling, developing commitment and acting with confidence" p. 83.

2.1 Goals: Deal with the inhibiting and enabling inner states; "build the intensity of caring, pleasure, and power ... involved in passionate pursuit"(p. 84-85).

2.1 Key elements:
Feeling - experience authentic emotional response to situation and know causes of that response.
Clarity - express feeling as interest in own environment, discover personal goals, develop commitment to learning action.
Confidence - feel worthy and capable, become confident that own plans will materialize, develop a capacity and desire for success.
Determination - develop drive and toughness to take risks, solve problems and overcome obstacles.
Intuition - trust the unspoken feelings that own ideas are correct/appropriate

3. Action

"Successful performance requires learning to act almost without thinking... learning to apply what we know" (p. 85).

3.1 Goals:

"making decisions, designing action plans, taking the risk to act, anticipating the action skills...securing feedback...recognizing... success" (p. 87).

3.1 Key elements:
Decision making - Assess choices available for situation, make appropriate decision.
Initiative - getting the energy and drive to be pro-active
Practice - rehearsal skills required to improve actual skills
Problem solving - use context feedback to see problems and devise solutions
Influence - use political and leadership skills to influence events.

Dear

This letter introduces a personal research project, and seeks your consent to be interviewed as part of the data gathering for the project.

I am an Ed.D. student in the Department of Adult Education of O.I.S.E. which is the Graduate Department of Education of the University of Toronto. To complete the degree I must carry out a research project and submit the analysis of the results to an examining committee which is chaired by Dr. Lynn Davie.

The topic of my research is individual reactions to and the use of learning strategies within the computer-mediated communication environment of an OISE course. In particular, I am interested in what strategies course participants used in their own individual information processing, in interacting with the course resources, in establishing a comfortable and productive state of mind, and in any self-monitoring of learning that may go on. Additional areas of interest are the possible impacts on participants of the unique characteristics of a computer-mediated learning environment, and the extent to which the passage of time may influence how participants behaved.

Those who agree to discuss their reactions and experience will be interviewed by me in three interviews approximately 4 weeks apart between January and April 1989. Each interview is expected to last somewhere between 60 to 90 minutes. A longer time period would be available, of course, should the interviewee and interviewer think it necessary. The timing and method of the interview will be at the convenience of the interviewee, and I expect to use the telephone for interviews of participants who live out of range of convenient face-to-face interviews. The interviews will be recorded only to ensure that no information is lost.

You are assured that your responses will be confidential to me. The interview tapes will not be used by anyone else, and any information that may link the discussions to you will be disguised - for example, pseudonyms will be used in any quotations to illustrate points being made, and the findings will be reported in categories of data, e.g., strategies, issues, or themes; that process will prevent the identification of individual interviews. In this way there will be no chance that your comments will be heard or seen by Dr. Davie or anyone else at OISE who is concerned with student grading. In short, you will have protection from any inadvertent evaluation. You will have an opportunity during the data analysis to check my interpretations of what you said; again, at a time of your convenience. You will not be responsible for any telephone or other communications costs -- these are my personal responsibility.