Transcript Making Thinking Visible (MTV): Promoting Students’ Model

Students’ Collaborative Model-Building and Peer
Critique On-line.
Janice Gobert
Concord Consortium
[email protected]
Mtv.concord.org
Making Thinking Visible is funded by the the National Science Foundation under grant
No. REC-9980600 awarded to Janice Gobert. The WISE project is funded by the National Science
Foundation by grants awarded to Marcia Linn. Any opinions, findings, and conclusions expressed
are those of the presenters and do not necessarily reflect the views of the National Science
Foundation.
Gobert, NARST, 2003
Summary
I will describe a large-scale design study of 2000 middle and high school students from California
and Massachusetts who collaborated on-line about plate tectonic activity in their respective
location. The students, demographically diverse, participated in this curriculum using WISE,
Web-based Inquiry Science Environment (Linn, 1998), an integrated set of software
resources designed to engage students in rich inquiry activities
The curriculum engaged students in many inquiry-oriented, model-based activities. For example,
students were scaffolded by WISE as they: a) drew initial models of plate tectonic
phenomena in their respective area using WISE; b) wrote explanations of their models and
shared their models and explanations with students on the opposite coast (east vs. west); c)
were scaffolded to critique their peers’ models; d) revised their models based on this
feedback; and e) discussed the differences between E and W coast geology in an on-line
forum.
Data analysis focussed on measuring content gains and characterizing the nature of students’
models and model revisions. Results suggest that this curriculum was successful in fostering
deep content learning. Additionally, the task of evaluating and critiquing their peers’ models
led to both a deeper understanding of the domain as well as fostered students’ epistemologies
of models.
Gobert, NARST, 2003
Forms of Knowledge & Cognitive
Affordances
Knowledge comes in various forms ; Models are one type of knowledge representation.
Different types of knowledge representations offer cognitive affordances, for example, a
continuum here is:
 textual representations, which describe in words various aspects of science phenomena
 diagrams/illustrations of static features of phenomena;
 models and simulations that attempt to show the causal mechanisms as well as dynamic and
temporal features of a phenomenon.
Each of these types of knowledge representations have different information-processing
affordances for learners.
Gobert, NARST, 2003
Information Processing of Various
Knowledge Forms
 One way to think about the information processing of various forms of knowledge
representations is to categorize them in terms of how visually isomorphic they are to the
objects/phenomena that they represent.
 Knowledge representations range from representations that at are not visually
isomorphic to the things that they represent (i.e., a textual description) to models and
simulations that exhibit a greater degree of correspondence to the objects that they
represent both in terms of the structure (i.e., spatial information) and how they function
(i.e., causal and dynamic information).
 In terms of the three knowledge forms outlined previously, this means that textual
representations offer the fewest cognitive affordances for learners and that models and
simulations, on the other hands, SHOULD offer the greatest number of cognitive
affordances for learners.
Gobert, NARST, 2003
Student Difficulty in Learning from Models
 Previously it was thought that diagrams and models would facilitate students’
understanding of difficult science concepts simply by “adding” a diagram or a model to
the textbook’s textual materials.
 However, research has shown that simply adding diagrams and models did not facilitate
learning because it increased cognitive load on learners (Sweller, et al, 1990).
 Also, students lack the necessary domain knowledge in order to guide their search
processes through diagrams/models in order to understand the relevant spatial, causal,
dynamic, and temporal information (Lowe, 1989; Head, 1984; Gobert, 1994; Gobert &
Clement, 1999).
Gobert, NARST, 2003
Scaffolding Learning from Models
Thus, scaffolding is necessary in order to support students’ learning with models, in
particular to support:
 search processes for acquiring rich spatial, dynamic, causal, and temporal information
from models (especially with models in which all information is presented
simultaneously).
 perceptual cues afforded by models in order to promote deep understanding.
 Inference-making with models, again, to promote deep understanding.
(adapted from Larkin & Simon, 1987)
Gobert, NARST, 2003
Scaffolding Framework for Learning with
Models (Gobert & Buckley, in prep.)
Type of Scaffold
Description of Pedagog ical Elements
Representational
Assistance
to guide students' understanding of the
representations or domain-specific
conven tions within model s.
Integration of pieces of
model
can take the form of r eflective qu estions, tasks,
explanations (either provid ed of studentgenerated) int ended to help students' in
integrating component s of the model to come
to a deeper understanding of the aspects of the
model (e.g., spatial, causal, dynamic,
temporal).
Model -based reasoning
supports
refers to tasks, etc., that support students in
reasoning with their model s and revising their
models.
Reconstruct, Reify, &
Reflect
refers to supporting students to refer back to
what they have learned, reinforce it, and then
reflect to mov e to a deeper level of
understanding.
Gobert, NARST, 2003
Making Thinking Visible: A project-based
curriculum for scaffolding rich model-based
learning.
 East and West coast Students’ collaborate on-line about the differences
in plate tectonic phenomena on-line using WISE (Web-based Inquiry
Science Environment; Linn & Hsi, 2000).
 In doing so, students develop…
 Content knowledge of the spatial, causal, dynamic, and temporal features
underlying plate tectonics.
 Inquiry skills for model-building and visualization.
 Epistemological understanding of the nature of scientific models.
See AERA and NARST papers from 2002 for these papers at mtv.concord.org
Gobert, NARST, 2003
Grounded in research in Science Education and
Cognitive Science...
 based on students’ misconceptions of plate tectonics of both the inside
structure of the earth and of the causal mechanisms underlying plate tectonicrelated phenomena (Gobert & Clement, 1999; Gobert, 2000), as well as
students’ knowledge integration difficulties (Gobert & Clement, 1994).
 emphasizes students’ active model-building and scaffolded interpretation of
rich visualizations (Kindfield, 1993; Gobert, 2001; Gobert & Buckley, in
prep.) as strategies to promote deep learning.
 Implemented in WISE (Web-based Inquiry Science Environment) developed
by Marcia Linn & Jim Slotta at UC-Berkeley, which is based on 15 years of
research in science education (Linn & Hsi, 2000).
Gobert, NARST, 2003
Model-based activities and respective
scaffolding for unit: What’s on your plate?
 Draw, in WISE, their own models of plate tectonics phenomena.
 Participate in an on-line “field trip” to explore differences between the East and West
coast in terms of earthquakes, volcanoes, mountains (beginning with the most salient
differences).
 Pose a question about their current understanding (to support knowledge integration
and model-building)
 Learn about location of earth’s plates (to scaffold relationship between plate boundaries
anf plate tectonic phenomena).
 Reify important spatial and dynamic knowledge (integration of pieces of model) about
transform, divergent, collisional, and convergent boundaries.
 Learn about causal mechanisms involved in plate tectonics, i.e., convection &
subduction (scaffolded by reflection activities to integrate spatial, causal, dynamic, and
temporal aspects of the domain).
 Learn to critically evaluate their peers’ models which in turn serves to help them think
critically about their own models.
Gobert, NARST, 2003
Model-based activities and respective
scaffolding for unit (cont’d)

Engage in model revision based on their peers’ critique of their model and what they have learned in
the unit.

Scaffolded reflection task to reify model revision which prompt them to reflect on how their model
was changed and what it now helps explain. Prompts are:
 “I changed my original model of.... because it did not explain or include....”
 “My model now includes or helps explain…”
 “My model is now more useful for someone to learn from because it now includes….”

Reflect and reify what they have learned by reviewing and summarizing responses to the questions
they posed in Activity 3.

Transfer what they have learned in the unit to answer intriguing points:
 Why are there mountains on the East coast when there is no plate boundary there?
 How will the coast of California look in the future?
Gobert, NARST, 2003
Part 1: Content Gain Results
 The students from one class on the West coast were partnered with the students from
two classes on the East coast because of the differences in class sizes. Five such sets or
“virtual classrooms” (referred to as WISE periods) were created in WISE.
 This is analysis of 360 students.
 A significant pre-post gain was found in all five WISE classrooms for content gains.
Gobert, NARST, 2003
WISE Period 1- sig. Content gains
Interaction Bar Plot for contentgain
Effect: Category for contentgain * teacher
8
Cell Mean
7
Fisher's PLSD f or content gain
Eff ect: teacher
Significance Level: 5 %
6
5
A
4
S
T
3
2
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-.322
1.130
.5745
A, T
.643
1.110
.2540
S, T
.964
1.252
.1298
1
0
preCtot
postCtot
Cell
ANOVA Table for contentgain
DF Sum of Squares
teacher
Mean Square
F-Value
P-Value
Lambda
Power
.998
.3745
1.996
.208
2
17.231
8.615
61
526.577
8.632
Category for contentgain
1
130.331
130.331
44.982
<.0001
44.982
1.000
Category for contentgain * teacher
2
22.548
11.274
3.891
.0257
7.782
.680
61
176.740
2.897
Subject(Group)
Category for contentgain * Subject(Group)
Gobert, NARST, 2003
WISE Period 2- sig. Content gains
Interact ion Bar Plot for cont ent gain
Eff ect: Cat egoryf or content gain * t eacher
7
6
Fisher's PLSD f or content gain
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
5
A
4
S
3
T
2
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-1.971
1.307
.0034
S
A, T
-1.603
1.307
.0167
S
1.468
.6209
S, T
.368
1
0
preCtot
postCtot
Cel l
ANOVA Table for cont ent gain
DF Sum of Sq uares
teacher
Subject( Group)
2
102.229
Mean Sq uar e
51.114
F-Val ue
P-Val ue
Lambda
Power
3.946
.0246
7.891
.687
60
777.298
12.955
Categ or y for content g ain
1
115.695
115.695
39.473
<.0001
39.473
1.000
Categ or y for content g ain * teacher
2
38.791
19.396
6.617
.0025
13.235
.911
60
175.860
2.931
Categ or y for content g ain * Subject(Group)
Gobert, NARST, 2003
WISE Period 3- sig. Content gains
Interact ion Bar Plot for cont entgain
Eff ect: Cat egoryf or contentgain * teacher
7
6
Fisher's PLSD f or content gain
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
5
A
4
S
3
T
2
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-1.010
1.300
.1267
A, T
-1.583
1.277
.0155
S, T
-.574
1.448
.4347
1
0
preCtot
postCtot
Cel l
ANOVA Table for contentgain
DF Sum of Squares
teacher
Mean Square
F-Value
P-Value
Lambda
Power
2.525
.0883
5.050
.476
2
60.752
30.376
62
745.837
12.030
Category for contentgain
1
85.178
85.178
26.654
<.0001
26.654
1.000
Category for contentgain * teacher
2
98.937
49.469
15.480
<.0001
30.960
1.000
62
198.133
3.196
Subject(Group)
Category for contentgain * Subject(Group)
Gobert, NARST, 2003
S
WISE Period 4 - sig. Content gains
Interact ion Bar Plot for cont entchange
Eff ect: Cat egoryf or contentchange * teacher
7
6
Cel l Mean
5
A
4
Fisher's PLSD f or content change
Eff ect: teacher
Significance Level: 5 %
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-.784
1.385
.2645
A, T
-2.083
1.360
.0030
S, T
-1.299
1.543
.0982
S
3
T
2
1
0
preCtot
postCtot
Cel l
ANOVA Table for contentchange
DF Sum of Squares
teacher
Mean Square
F-Value
P-Value
Lambda
Power
3.898
.0254
7.796
.682
2
97.656
48.828
62
776.675
12.527
Category for contentchange
1
130.942
130.942
25.019
<.0001
25.019
1.000
Category for contentchange * teacher
2
59.218
29.609
5.657
.0055
11.315
.855
62
324.487
5.234
Subject(Group)
Category for contentchange * Subject(Group)
Gobert, NARST, 2003
S
WISE Period 5 - sig. Content gains
Interact ion Bar Plot for cont ent gain
Eff ect: Cat egoryf or content gain * t eacher
8
7
Fisher's PLSD f or content gain
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
6
5
A
4
S
T
3
2
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-2.394
1.236
.0002
S
A, T
-3.285
1.331
<.0001
S
S, T
-.891
1.446
.2248
1
0
preCtot
postCtot
Cel l
ANOVA Table for content gain
DF Sum of Squares
teacher
Mean Square
F-Value
P-Value
Lambda
Power
13.509
<.0001
27.018
.999
2
256.450
128.225
60
569.514
9.492
Category for content gain
1
82.505
82.505
18.220
<.0001
18.220
.994
Category for content gain * teacher
2
107.916
53.958
11.916
<.0001
23.832
.997
60
271.692
4.528
Subject(Group)
Category for content gain * Subject(Group)
Gobert, NARST, 2003
Part 2: Epistemological Gain Results
 A significant pre-post gain was found in all five WISE classrooms for epistemological
gains.
Gobert, NARST, 2003
WISE Period 1 - sig. Epistemological gains
Interact ion Bar Plot for modelgain
Eff ect: Cat egory for modelgain * t eacher
16
14
Cel l Mean
12
10
A
8
S
Fisher's PLSD f or modelgain
Eff ect: teacher
Significance Level: 5 %
T
6
4
Mean Diff.
Crit. Diff.
P-Val ue
A, S
1.012
1.571
.2047
A, T
.511
1.543
.5139
S, T
-.502
1.739
.5692
2
0
preMtot
postMtot
Cel l
ANOVA Table for modelgain
DF
Sum of Sq uares
2
22.442
11.221
61
988.926
16.212
Categ or y for model gain
1
115.697
115.697
Categ or y for model gain * teacher
2
83.882
41.941
61
439.837
7.210
teacher
Subject( Group)
Categ or y for model gain * Subj ect(Gr oup)
Mean Sq uar e
Gobert, NARST, 2003
F-Val ue
P-Val ue
Lambda
Power
.692
.5044
1.384
.157
16.046
.0002
16.046
.987
5.817
.0049
11.633
.866
WISE Period 2 - sig. Epistemological gains
Interact ion Bar Plot for modelgain
Eff ect: Cat egory for modelgain * t eacher
14
12
Fisher's PLSD f or modelgain
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
10
A
8
S
6
T
Mean Diff.
Crit. Diff.
P-Val ue
A, S
.064
1.632
.9380
4
A, T
-.289
1.632
.7268
2
S, T
-.353
1.827
.7028
0
preMtot
postMtot
Cel l
ANOVA Table for modelgain
DF
teacher
Subject( Group)
2
Sum of Sq uares
2.335
Mean Sq uar e
F-Val ue
P-Val ue
Lambda
Power
1.167
.079
.9244
.158
.061
59
874.504
14.822
Categ or y for model gain
1
311.401
311.401
40.945
<.0001
40.945
1.000
Categ or y for model gain * teacher
2
56.782
28.391
3.733
.0297
7.466
.659
59
448.710
7.605
Categ or y for model gain * Subj ect(Gr oup)
Gobert, NARST, 2003
WISE Period 3 - sig. Epistemological gains
Interact ion Bar Plot for modelchange
Eff ect: Cat egory for modelchange * teacher
14
12
Fisher's PLSD f or modelchange
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
10
A
8
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-.809
1.684
.3437
A, T
.833
1.654
.3207
S, T
1.642
1.876
.0857
S
6
T
4
2
0
preMtot
postMtot
Cel l
ANOVA Table for modelchange
teacher
Subject( Group)
DF
Sum of Sq uares
2
47.195
Mean Sq uar e
23.597
F-Val ue
P-Val ue
Lambda
Power
1.433
.2464
2.866
.285
62
1021.132
16.470
Categ or y for model change
1
366.531
366.531
54.803
<.0001
54.803
1.000
Categ or y for model change * teacher
2
106.362
53.181
7.952
.0008
15.903
.958
62
414.665
6.688
Categ or y for model change * Subject( Gr oup)
Gobert, NARST, 2003
WISE Period 4 - sig. Epistemological gains
Interact ion Bar Plot for modelchange
Eff ect: Cat egory for modelchange * teacher
14
12
Fisher's PLSD f or modelchange
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
10
A
8
S
Mean Diff.
Crit. Diff.
P-Val ue
A, S
-.073
1.392
.9180
4
A, T
-1.589
1.367
.0231
2
S, T
-1.516
1.551
.0552
Lambda
Power
6
T
0
preMtot
postMtot
Cel l
ANOVA Table for modelchange
DF
teacher
Sum of Sq uares
Mean Sq uar e
2
63.678
31.839
62
703.214
11.342
Categ or y for model change
1
190.437
Categ or y for model change * teacher
2
62
Subject( Group)
Categ or y for model change * Subject( Gr oup)
F-Val ue
P-Val ue
2.807
.0681
5.614
.523
190.437
35.768
<.0001
35.768
1.000
65.833
32.917
6.182
.0036
12.365
.889
330.098
5.324
Gobert, NARST, 2003
S
WISE Period 5 - sig. Epistemological gains
Interact ion Bar Plot for modelchange
Eff ect: Cat egory for modelchange * teacher
14
12
Fisher's PLSD f or modelchange
Eff ect: teacher
Significance Level: 5 %
Cel l Mean
10
A
8
S
6
T
Mean Diff.
Crit. Diff.
P-Val ue
A, S
.701
1.631
.3970
4
A, T
-.531
1.758
.5510
2
S, T
-1.232
1.909
.2040
F-Val ue
P-Val ue
Lambda
.840
.4368
1.680
.181
0
preMtot
postMtot
Cel l
ANOVA Table for modelchange
teacher
Subject( Group)
Categ or y for model change
Categ or y for model change * teacher
Categ or y for model change * Subject( Gr oup)
DF
Sum of Sq uares
2
26.202
Mean Sq uar e
13.101
60
936.016
15.600
1
444.676
444.676
75.513
<.0001
75.513
1.000
2
90.227
45.113
7.661
.0011
15.322
.950
60
353.325
5.889
Gobert, NARST, 2003
Power
Portfolio for one pair of students selected
for typical performance….
Gobert, NARST, 2003
Gobert, NARST, 2003
Activity 1 (cont’d): Explain your model.
Gobert, NARST, 2003
Gobert, NARST, 2003
Activity 3: Pose A Question.
Gobert, NARST, 2003
Activity 4: Earth’s Plates.
Gobert, NARST, 2003
Gobert, NARST, 2003
Activity 5: The Mantle.
Gobert, NARST, 2003
Activity 6: Students’ Evaluation and
Critique of the Learning Partners’ Models.
2. Students’ Evaluation and
Critique of the Learning Partners’ Models
 Students read two pieces of text in WISE called “What is a Scientific Model? And
“How to evaluate a model?”
 Students critique learning partners’ models using prompts in WISE. Prompts include:
 1. Are the most important features in terms of what causes this geologic process depicted in this
model?
 2. Would this model be useful to teach someone who had never studied this geologic process
before?
 3. What important features are included in this model? Explain why you gave the model this
rating.
 4. What do you think should be added to this model in order to make it better for someone who
had never studied this geologic process before?
 Prompts were designed to get students to reflect on what causal features should be
included in the model and how useful the model was as a learning/communication tool.
Gobert, NARST, 2003
W. Coast group’s evaluation of E. coast
group’s model
Gobert, NARST, 2003
E. Coast group’s revised model.
Gobert, NARST, 2003
E. Coast group’s revised explanation.
Gobert, NARST, 2003
Notes on model revision.
Gobert, NARST, 2003
Activity 8: What have we learned?
Volcanoes - Mi chel a K., Scott C., Mi ke B.
Why do volcanoe s e ru pt?
Respond - Christ ina V, Peter D., Eri K.
volacanos erupt because plat es collide together and the hot lava bust s from the mantle.
What happens to lava underwater - Nick M., Li l l ai n F., Fi l i pZ.
What happens when the lava spil l s ou t u n de rwate r? How qu ick ly doe s i t cool
Respond - Jessica D., Alexandra M., Colm G.
When the lava goes into the ocean, it cools and hardens causing new plant life because
the ashes are very reach in minerals and nutrient s. A new step of an ocean floor is made.
Is the m an tle an d magma the same thing?- Laura C., Ale x Y.
Respond-Jonathan S, Paul C, Elizabeth V.
Magma is part of the mant le, the part of the mant le that is liquid
Why are som e volcanoes dorm int whi le some are active? Phi l i pW., Jeff G, Christa
P.
Respond - Jonathan S, Paul C, Elizabeth V.
The active ones are over plate rifts. In Hawaii, the islands were formed wh en the plate
they're on drifted of a pocket of magma. It cont inued drifting, and soon many islands
were formed where the magma pocket happend to be relat ive t o the plate, when t he
pocket erupted.
Gobert, NARST, 2003
Gobert, NARST, 2003
Comments on Example 1...
In this example, the studen t s drew a model of volcanic eruption which includes only the
crustal layer of the earth; that is, the inside layers of the earth are not depicted, nor are
there any internal causal mechanisms responsible for volcanic erupt ion included in eit her
the model or explanation. This type a model is called a ÒlocalÓmodel and is consistent
with previous research in t his domain which showed that many studen t s of this age group
have models of plate tectonic phenomena which only include processes on the surface of
the earth, i.e., they do not include the processes and mechanisms inside the earth (Gobert ,
2000). T he correct concept ions that are represented in the model and/or explanation are:
hot magma, movement of magma beyond the volcanic cone, and magma forming new
rock. (For an example of the coding scheme for volcanic erupt ion, see Appendix B.2).
The learning partnersÕcrit ique is very detailed in that it sugges t s that the studentsÕmodel
needs Òlabels, cause, plates, types of volcano, interior, exterior, and what the volcano was
doingÓ. The studen tsÕrevised model includes some the learning partnersÕ suggest ions.
The revised model, includes plates and labels and the studen ts have elaborated on one
type of volcano as reques ted by their learning part ners. More specifically, their
explanat ion it appears the studen t s were trying to depict /describe volcanism due to plat e
convergence 1. The studen ts have also included p late movement and plate frict ion as
causal mechanisms responsible for volcanic erupt ion. Although the revised model only
includes a few addit ional causal mechanisms from t he original, it is a significant advance
over their original model.
Gobert, NARST, 2003
Gobert, NARST, 2003
Comments on Example 2…..
In this example the studen tsÕmodel represent s a misconception, i.e., that a mount ain is
formed and fills up with lava and when it fills up, it erupts. Unfortunately, the learning
partnersÕcrit ique did not include much information upon which a revision could be
based; this is possibly due to them not knowing what to do in the case of an ÒincorrectÓ
model. In the revised model and explanat ion (which we assume is based on the content of
the unit rather than t he learning partnersÕcrit ique), the studen ts have added plate
subdu ction and magma movement as a causal mechanism in how volcanoes are formed
and have also included the concept of pressure as building up within the volcano. It is
import ant to note that although their reasoning here is not ent irely correct, intuit ive
conceptions such as pressure are rich, effect ive pieces of knowledge that can be
effect ively built upon (Clement, Brown, & Zietsman, 1989) and are usable anchors for
developing understanding of convect ion (Gobe rt & Clement ,1994). As such the revised
model represents gain in understanding.
Gobert, NARST, 2003
Gobert, NARST, 2003
Comments on example 3….
In this original model above (left), the studen t s had focussed on the crustal layer of the
earth and had not included what happens inside the earth when mountains are formed;
that is, there is no st ructural informat ion or causal informat ion about the inside of the
earth. Again, this is a ÒlocalÓmodel of plate tect onic phenomena (Gobe rt, 2000) because
it does not include any processes or mechanisms inside the earth. In the crit ique which
was done by their West coast partners, the learning partners requested t hat they label their
model. The revised model includes labels (as sugges ted); it is also a much more detailed
model, sugg est ing that the studen ts learned a great deal from the content in the ÒWhatÕs
on your plate?Ócurriculum. Their new model includes the crustal layer as a Òcut awayÓ
from the cross sect ion view; it also includes convect ion as a causal mechanism in
mountain building (in t he original model there were no causal mechanisms included ).
The inclusion of convect ion as a causal mechanism, the relationship of the convection to
the crustal movement and the locat ion of the convect ion in the correct layers of the earth
(the mant le), in their revised model represent s a significant advance from their earlier
model (Gobe rt, 2000).
Gobert, NARST, 2003
Gobert, NARST, 2003
Comments on Example 4….
In this example, the studen t sÕoriginal model has two views: a cross sect ion view, and a
crustal layer view. Their model and explanat ion include no causal mechanisms in terms
of what happens inside the earth when mountains are formed; thus, it is a local model
(Gobe rt, 2000). In the crit ique from their learning part nersÕ, it was sugg ested that the
students include the direction of movement of the plates. This is a high level comment in
that it reflects that the reviewers knew that t his information was important to the causality
of the system being depicted. The critique also includes comment s related to the model
as a communicat ion tool, i.e., they sugge sted t hat the studen ts include a cross sect ion
view rat her than a birdÕs eye view which is good comment regarding the model as a
communicat ion t ool. T he revised model includes the earth in cross sect ion form with a
cut away t hat includes informat ion about the plates moving toward each other. In addition
the studen ts have added the mant le as a causal mechanism. Although not a significant
advance from the point of view of including more detailed causal informat ion, the revised
model is a bet ter model from a communicat ion standpoint , as was requested by their
learning partners.
Gobert, NARST, 2003
Conclusions
 In most of these programs to date, students are either presented with models to learn
from (Raghavan & Glaser, 1995; White & Frederiksen, 1990) or they are given tasks
which require them to construct their own models (Gobert, & Clement 1994, 1999;
Gobert, 1998; 1999; Penner et al., 1997; Jackson, et al., 1994).
 This research extends a current vein of progressive model-building in science education
(cf., Raghavan & Glaser, 1995; White & Frederiksen, 1990) by having students critique
each others’ models as a way to promote deep understanding.
 Furthermore, all tasks in which students’ are constructing models, are learning with
models, and are critiquing models of their peers are scaffoled using a model-based
scaffolding framework (Gobert & Buckley, in prep.) in order to promote both deep
understanding of the content as well as promote a deep understanding of models in
science and how they are used in theory development.
 It is believed that rich, scaffolded model-based tasks such as these engages students in
authentic scientific inquiry, and as such can significantly scientific literacy.
Gobert, NARST, 2003
To found out more ...
 To view the unit, go to wise.berkeley.edu, click on Member entrance, and for
login enter “AnonyM1” and “try” as your password. Click on “Plate Tectonics:
What’s on Your Plate?”
 To find more information…
 E-mail: [email protected] and get a copy of this paper.
 Other papers are available on this work at mtv.concord.org
 For more on The Concord Consortium contact www.concord.org.
Gobert, NARST, 2003