Computer Supported Intentional Learning

Scardamalia, M. and Bereiter, C.

Notes from the CSILE papers assigned in class

noel@gaia.gatech.edu (Noel Rappin)

For Educational Technology, Winter 1994

Knowledge Building Communities

We seem to all agree that schools are failing in producing the learning
we want.  So far, we have tried to tailor the learning environment to the
sophistication of the student, but Scardamalia suggests that it may not
be appropriate for the level of the student to determine the student's
environment. It may be that the implicit expectations of the environment
is a causative agent is determining learning.

Environments which produce consistant excellence include sports and
business. These environments require continuous adaptation which causes
growth. These environments are classified as second order environments.
By contrast, schools in most cases require limited adaptation, and hence
limited growth.

Present day research communities seem to exemplify these second order
environments.  Alfred Whitehead noted the 'disciplined progress' and
synergistic results of the precursor of today's research communities
in the universities of Germany near the turn of the century.  Scardamalia
classifies these highly successful groups as knowledge building communities.

The belief is - if the environment of our students is modeled after
these knowledge building communities - the sophistication of the
learning exhibited by our students will be similar to the sophistication
of learning exhibited by scientists and researchers.

The goal of making individual students perform like scientists is not novel.
Previous attempts have included increasing the sophistication of the tools
the students use in the form of microworlds, tutoring systems, etc.  Arguably,
these techniques have failed. Therefore, it may seem that changing the
classroom to be more like a research center is ill-timed and inappropriate.

As long we believe that the sophistication of the students determines the
appropriate environment for the students, dramatically this kind of
transformation of the classroom is untennable.  However, if the environment
is key in determining the sophistication of the student, such a change could
be revolutionary.

If we are to restructure the learning environment of students after
a research community, we must determine what about the research
communty makes it a knowledge building community.
The defining characteristics of knowledge building communities 
are the production and refining of objective knowledge in a social setting.
For research communities, this means the journal process.  The journal
process, through peer review, and the publication of articles produces and
refines objective knowledge.

Salient Features of the Journal Process

The salient features of the journal process are its focus on problems and
depth of understanding, the decentralized, open knowledge environment, and
the productive interaction within broadly conceived communities.  Focusing on
problems and depth of understanding move the focus from topics or categories.
It has been shown that both of these are weak learning strategies.

The decentralized and open learning environment is a shift away from the
structured environment in most schools.  This environment is useful for
the professional community in that it can engage busy people, distribute
work among members, sustains increasingly advanced inquiry and ensures that
the group is working on the forefront of their collective understanding.
In the typical didactic classroom, the teacher is outside the the learning
process, and either advanced or slow students are often not engaged.
In knowledge building communities, no one stands outside the learning process,
but everyone has a role where they are active participants in learning.

Knowledge building communities are not specific but broadly concieved because
this ensures a second-order environment. Within a niche, participants
may beome comfortable and stop adapting.  The standard in a broadly concieved
community constantly raised due to the accomplishments of others, and the
social dimension of the environment captures the interest of participants.

The promise is that if students are immersed in an environment where goal
are dynamic yet clear, and there is motivation beyond purely epistemic
motivation, students will exhibit sophisticated learning at the level of
today's top researchers.

It is obvious that modern technology is not necessary to the existence of
knowledge building communities.  These communities have existed for over
100 years.  However, modern technology is important because it makes the
existance of these communities feasable in today's classroom.  One example
of this technology is CSILE.  This will be evaluated next.


***  Computer Supported Intentional Learning Environments (CSILE)  ***

(Summary of paper by Scardamalia et. al. presented by Nitin 'Nick' Sawhney)

Definition:  CSILE is an "educational knowledge media system", with a 
common database to store, retrieve and link information from several media.  
This expression is used to refer generically to environments that foster the 
ability of students to exert intentional control over their own
learning. CSILE 
supports intentional learning by providing a means for a group of students to 
build a collective database of their thoughts using written notes and
pictures      .  
Pictures can be "zoomed" in or out and written text can be labeled in a number 
of ways to facilitate reflection and allow the notes to appear in multiple 
contexts.  The notes can also be placed on a timeline or attached to a
spot on a picture.  

     CSILE was initially developed for college students, yet the first-year 
implementation focused on two grade fix-six classes.  CSILE was made 
available on 16 networked computers.  It is designed for eventual use at all 
grade levels and for all school curricula, as well as other conventional school 
activities.  An objective was to develop specifications for a wide range of 
educational software to support reflection, problem solving, and learning.  It 
promotes educational objectives like planning, monitoring, goal-setting, and is 
aimed at developing the intelligence of the user instead of making software as 
intelligent as possible.

Four persistent characteristics of passive or immature learners is listed to 
understand better strategies for active learning:

     1)  Organization of mental activities around Topics vs. Goals
     2)  Focus on surface features (vs. depth?)
     3)  Use of Straight Ahead vs. Recursive features
     4)  Additive vs. Transformational approach to learning

     A computer system can provide maximum support for developing higher-
order abilities, by active vs. passive learning strategies.  The authors argue that 
an intelligent tutoring system is the wrong approach to address this issue, for it 
is not the computer but the student that must do the diagnosing, goal-setting, 
and planning.  The computer provides the facilitating structure and tools to 
enable students to make maximum use of their own intelligence and knowledge 
(procedural facilitation).

Eleven principles (based on recent cognitive research) are suggested for 
designing computer environments that support intentional learning.

1.	Make Knowledge Construction Activities Overt
2.	Maintain Attention to Cognitive Goals
3.	Treat Knowledge Lacks in a Positive Way
4.	Provide Process-Relevant Feedback
5.	Encourage Learning Strategies other than Rehearsal
6.	Encourage Multiple Passes through Information
7.	Support Varied Ways for Students to Organize their Knowledge
8.	Encourage Maximum Use and Examination of Existing Knowledge
9.	Provide Opportunity for Reflectivity and Individual Learning Styles
10.	Facilitate Transfer of Knowledge across Contexts
11.	Give Students More Responsibility for Contributing to Each Other's 
Learning

The application of these principles to CSILE as well as observations from trials 
in the first year is described below:

1.	Make Knowledge Construction Activities Overt

WHY?:  So that students become aware of them and are better able to carry 
them out deliberately.
HOW?:  Students select the mental activities they wish to engage in and they 
have a consequence within the functioning of the software.
CSILE?:  Students identify notes that they or others created by icons (student-
designed) that represent different thinking types.

2.	Maintain Attention to Cognitive Goals

WHY?:  To help students anticipate what they will learn and what they will do 
en route to attaining that goal.  These should be cognitive goals (learning, 
finding out, etc.) rather than task goals (scoring, finding treasure, etc.).
HOW?:  Students should be called on to state their goals and possibly to 
specify a time at which they think they will have reached their goal.
CSILE?:  Planning is one of the thinking types (selected with student-designed 
icon) supported by prompts to indicate cognitive goals, plans to pursue them, 
and target dates.  Students schedule deadlines for investigating questions using 
the timeline.  In the future a calendar function will be implemented to provide 
an overview of learning goals and check progress of goals on current log-on 
date.

3.	Treat Knowledge Lacks in a Positive Way

WHY?:  Knowing what one does not know is a kind of metaknowledge, 
without which knowledge is severely limited.
HOW?:  Educational software should allow students to identify what they don't 
know or are curious about. Identification of knowledge lacks should have 
positive consequences within the functioning of the program to help motivate 
and enhance possibilities for achieving goals.
CSILE?:  The notes enable students to raise questions reflecting interests or 
curiosity.  Class activities, procedural facilitation's, and feedback provisions 
within CSILE are used to encourage serious question formulation.

4.	Provide Process-Relevant Feedback

WHY?:  Focusing on process rather than products is a more useful educational 
strategy.
HOW?:  Design partner or team activities in which one member has the job of 
monitoring processes and is provided with computer support for doing so.
CSILE?:  (indirectly provided)  Teachers can obtain reports of the number and 
kind of notes (and thinking types) produced by each student, based on which 
useful feedback can be provided to the students.  Future versions will provide 
reports to students with summarized data and self-ratings.

5.	Encourage Learning Strategies other than Rehearsal

WHY?:  Students are more successful when rehearsal and memorization 
strategies are supplemented with strategies for comprehensive understanding.
HOW?:  Asking comprehension questions, arranging tasks, and cloze 
procedures using paraphrases instead of previously read material.
CSILE?:  students review notes and compile them in categories of "questions I 
must answer" and also pose questions for peers to answer.  This activity is 
supported by interactive environments and diagrammatic facilities allowing 
students to pose different hypothesis(notes and illustrations) about the 
questions.

6.	Encourage Multiple Passes through Information

WHY?:  Naive students have a strong tendency not to go back over 
information.  Their reluctance to revise compositions or check math work 
displays single-pass strategies.  HOW?:  Educational software should help 
students develop multiple pass strategies.  Ways should be sought to make it 
worth-while for students to call back information or reconsider its use in a 
different context.
CSILE?:  Database searches enable the same note to appear in different 
contexts.  Scratchpad, note-editing, and copying functions encourage students 
to reprocess previously recorded information.

7.	Support Varied Ways for Students to Organize their Knowledge

WHY?:  Other ways of organizing information besides hierarchical list 
structures must be devised to help students better grasp arguments, lines of 
thought, visualize relationships or themes.
HOW?:  Alternative ways of structuring knowledge such as timelines, graphs, 
maps, narrative sequences, story grammar structures, concept nets, and casual 
chains should be considered.
CSILE?:  There are provision for different representations of knowledge 
accessible in the same database.  Currently implemented alternatives include 
maps, diagrams, timelines, freely constructed pictorial, as well as written notes. 
Diagrams are hierarchically embedded so that students can zoom in or out at 
different levels of detail and attach notes at any level.

8.	Encourage Maximum Use and Examination of Existing Knowledge

WHY?:  Real-world tasks require that students not be confined by a limited 
body of knowledge.
HOW?:  Educational software may draw from larger databases or support 
more open-ended tasks in which the student may draw on knowledge from a 
variety of sources.  The feedback should be provided by the teacher and not 
the computer.
CSILE?:  A large knowledge-base is constructed by the students themselves.  
Students are not only responsible for putting knowledge into the system but 
also for evaluating, interrelating, labeling, sorting, and reorganizing it.  Future 
versions will incorporate prepackaged databases and video technology.

9.	Provide Opportunity for Reflectivity and Individual Learning Styles

WHY?:  Personal learning styles of different students are not suited to the pace 
and "publicness" of classroom learning.
HOW?:  Educational software must permit private, self-paced learning as well 
as provide students time, opportunity, and peace to think and reflect on the 
material.  Thus it should not be busy motivating the student or controlling 
his/her thoughts.
CSILE?:  It provides (if student chooses) anonymity in posing and answering 
questions.  It provides a scratchpad for notes, a facility for placing notes on 
hold while other ideas are being pursued, and editing database notes.

10.	Facilitate Transfer of Knowledge across Contexts

WHY?:  Compartmentalization of learning is blamed for students failure to 
transfer knowledge across new contexts.
HOW?:  Educational software can cut across curricular lines to apply 
knowledge from several disciplines (in microworlds and games).
CSILE?:  It allows cross subject searches and clustering of notes into common 
themes (over time) leading to more powerful cross-domain understandings.  
Future versions will incorporate intelligent database management to support 
interconnections among various notes.

11.	Give Students More Responsibility for Contributing to Each Other's 
Learning

WHY?:  Encourage cooperative learning.
HOW?:  Educational software should help students recognize what they are 
learning and provide aggregate data to monitor the learning progress of the 
class as a whole.
CSILE?:  1) Students respond to other students requests for information, 
confusion, self-rating, etc.  2) All students are involved in preparatory thinking 
of goals for the class.  3) By getting students to participate in what has been 
traditionally been regarded as "teaching" rather than "learning" activities, 
CSILE can promote higher levels of knowledge (associated with being able to 
teach what one has learned).

Conclusion:  The authors claim that despite the use of CSILE by the students 
for over a period of almost 8 months, they never got tired of it and continued 
to use it for the rest of the school term.  Students used the system to elaborate 
models and hypothesis, delve into difficult and texts, seek deeper explanation, 
and elaborate confusion.  The authors are confident that core design of CSILE 
is appropriate and flexible enough to foster intentional learning across the 
school curricula and for children at all grade levels.