Systematisch Ontwerpen voor complex leren

1
Systematisch Ontwerpenvoor complex leren

• Jeroen van Merriënboer
• Open Universiteit Nederland
• Studiemiddag CRWO
• 24 november 2000, VSNU, Utrecht

2
inhoud

• Inleiding 13.30-14.15
• complex leren
• het 4C/ID-model
• (Actieve) bezinning 14.15-15.15
• opdracht maken globale blauwdruk
• Discussie 15.30-16.30

3
Complex learning

• diagnosing X-ray pictures
• engineering software
• maintaining aircraft
• analyzing policies based on simulation models
• diagnosing faults in chemical plants
• designing electronic circuits
• controlling air traffic
• operating complex pieces of medical equipment
• analyzing statistical data
• examining patent applications

4
some characteristics

• constituent skills
• goal-directedness
• simultaneous learning processes
• lengthy and effortful process

5
constituent skills
Examining patent applications
Performing substantive examination
Issuing the communication or vote (including B09)
Re-examining the application
Preparing the search report
Examination of amendments
Discussions with applicant
Writing further communication or refusal
Analyze applications
Perform the search
Write B09 (pre-examination result)
Determine
Use
Evaluate
Determine
Novelty/
Other EPC
Determine
Classify
search
search
search
inventive
requirements?
claimed
main features
application
strategy
tools
results
subject-matter
step?
of invention

• co-ordination
• integration

Determine
Determine
Lack of
Compare
Select relevant
invention
invention
unity?
documents with
documents
described/
claimed
invention
inventive
concept
6
goal-directedness

• knowledge-based (controlled)
• problem solving in its rudimentary form
• working backward
• schema-based (controlled)
• knowledge of subgoals controls behavior
• finding means for reaching subgoals
• skill-based (automatic)
• routine
• working forward

7
simultaneous learning processes

• schema construction
• leads to the development of schema-based
  behaviors
• non-recurrent constituent skills
• rule-/schema automation
• leads to the development of skill-based behaviors
• recurrent constituent skills

8
lengthy effortful process

• at least 100 hours to reach basic level of
  proficiency
• may take thousands of hours of practice to become
  a real expert
• learners have to invest large amounts of effort
  in acquiring the skill
• cognitive capacity constraints

9
Outline of 4C/ID-model

• working on authentic learning tasks in real or
  simulated task environments

learning tasks
complex learning outcomes competencies

• simultaneous, integrated
• process of
• schema construction
• rule/schema automation
• attitude formation

• cases
• projects
• problems
• ......

• transfer

10
4 components

• learning tasks
• backbone of training program
• supportive information
• JIT-information
• part-task practice

11
component 1learning tasks (whole-task practice)

• schema construction
• cognitive strategies (how-to-approach)
• mental models (how-it-works)
• rule/schema automation (to some degree......)

12
Task classes support

• higher complexity for each new task class
• diminishing support within a task class

13
component 2 supportive information

• systematic approaches to problem solving
• modeling examples
• help to develop cognitive strategies
• conceptual/structural/causal models
• case studies
• help to develop mental models

14
component 3JIT-information

• for recurrent aspects of learning tasks
• present precisely when necessary

15
component 4part-task practice

• repetition
• JIT-info
• cognitive context

16
decompose the complex skill
sequence skill clusters and/or task classes
analyze non-recurrent aspects
analyze recurrent aspects
identify procedures or rules
identify cognitive strategies
identify mental models
identify prerequisite knowledge
refine sequence of task classes
(5)
(3)
(1)
(2)
(4)
17
(No Transcript)
18
(Actieve) bezinning
...the ability to produce a coherent and
appropriate sequence of case studies and problems
is a key feature in the design of constructivist
learning environments (Collins, Brown,
Newman, 1987)
19
Three-level sequencing in the 4C/ID-model

• skill clusters in a skills hierarchy (macro)
• task classes (basic sequencing level/meso)
• learning tasks within each task class (micro)

20
Meso-level sequencing

• basic level for sequencing
• simple-to-complex sequence of task classes
• whole-task sequencing
• each task class reflects all constituent skills
• allow the learners to quickly develop a holistic
  vision upon the whole task
• focus on co-ordination and integration of
  constituent skills

21
Example (patent examination task classes)
?

• write a vote (granted) for a clearly written
  application yielding only X and Y-documents
  (marginally relevant)
• write a communication for a clearly written
  application yielding many A-documents (competing)
• write a vote (granted) for a long, verbose
  application yielding X and Y-documents
• etcetera

22
Macro-level sequencing

• only if you cannot find a task class that is easy
  enough to start your training with
• define and sequence skill clusters
• meaningful/interrelated sets of constituent
  skills
• reflect realistic or authentic tasks
• sets preferably overlap
• first cluster allows learners to quickly start
  practice on non-trivial tasks

23
E
A
B
C
D

• forward chaining
• one-by-one A - B - C - D
• snowballing A- AB - ABC - ABCDE
• backward chaining
• one-by-one D based on results of ABC - C based on
  results of AB - B based on results of A - A
• snowballing D based on results of ABC - CD based
  on results of AB - BCD based on results of A -
  ABCDE

Confront with useful models from the beginning on
24
Patent examination
1
2
3
Performing substantive examination
Issuing the communication or vote (including B09)
Re-examining the application
Preparing the search report
Examination of amendments
Discussions with applicant
Writing further communication or refusal
Analyze applications
Perform the search
Write B09 (pre-examination result)
Determine
Use
Evaluate
Determine
Novelty/
Other EPC
Determine
Classify
search
search
search
inventive
requirements?
claimed
main features
application
strategy
tools
results
subject-matter
step?
of invention
Determine
Determine
Lack of
Compare
Select relevant
invention
invention
unity?
documents with
documents
described/
claimed
invention
inventive
concept
25
example (patent examination skill clusters)

• perform a substantive examination based on an
  application and given search report
• write a B-09 (pre-examination results) and
  perform a substantive examination, based on an
  application and given search results
• perform the complete patent examination

and then define task classes for each skill
cluster...
26
ExampleTask classes for cluster 1

• Application and search report given. Writing
  votes for clear applications, search report
  correctly citing A-documents only.
• Application and search report given. Writing
  communications for clear applications. Search
  report citing X and Y-documents.
• Application and search report given. Writing
  communication. Application is long and verbose
  search report citing A-documents only.

27
Micro-level sequencing

• develop/select the concrete learning tasks for
  each task class
• the set of learning tasks within a task class
  shows a high variability
• support that is given to learners diminishes if
  they acquire more expertise for a given task class

28
Build a preliminary outline for a training
program, which consists of a sequence of learning
tasks. It concerns a course for students who must
learn to search for literature. The goal of the
training is that students must be able to make a
selection from scientific literature that is
relevant for a particular research question.
Sequence both on the macro and meso level. Below
is a description of an expertss
approach Searching for Literature An experts
approach An expert literature searcher will
first select one or more suitable databases in
which to search for relevant literature. This may
be a simple database of articles the researcher
gathered before, a database available at the
institute where the researcher is working, or a
large database with all articles that appeared in
a particular field of research. In order to
select a suitable database, the researcher must
determine at least two things. First, the field
of research that is pertinent to the research
question must be identified. And second, the time
frame that is relevant to the research question
must be identified Only the literature of the
last few years or also articles that have been
published further in the past? After selecting a
database, the researcher formulates his or her
search query. Therefore, it is necessary to
translate the research question into relevant
search terms. In order to do so, it is possible
to consult a so-called thesaurus, that is, a list
with relevant search terms for a particular
database or even a particular field of research.
Typically, a number of search terms is combined
in the search query. Literature databases
typically use Boolean operators for combining
search terms. These are words like AND, OR and
NOT. The more search terms are used, the more
complex the search query is. Then the search
query is executed. The researcher operates the
software that is used to search the database.
This software includes a fixed number of commands
for performing the search. The researcher also
has to determine which fields of records in the
database are used for the search. In the simplest
case, the search is only performed on the title
of the articles. But it is also possible to
search in the abstracts of the articles or even
the whole text (full text search). Finally, a
selection is made from the results yielded by the
search query. On the basis of the abstract of
each article, it is estimated if this particular
article is really relevant for the research
question. The more articles have been found, the
more difficult it is to make a good selection.
29
(No Transcript)
30
Discussie

• Enige mogelijke topics
• Soorten leertaken (support)
• Media
• Toepassing model in HO
• Dynamische taakselectie
• Computer-ondersteund ontwerpen
• Relevant onderzoek
• ....................