Scaffolding learning activities in real life contexts with collaborative scripts and mobile

computers

Jari Laru,University of Oulu

25.4.2012

Focus of the study

The general focus of this doctoral thesis is to apply theoretical ideas of distributed cognition and scaffolding for mobile computer supported collaborative learning in authentic contexts

Ill-structured problems in this thesis

Ill-structured problem solving was a core task in all experiments in this doctoral thesis. According to Jonassen (2002) problems can be either well-structured, when there is one clear solution and solution path, or ill-structured, when there are unclear problem elements and multiple possible solutions and solution paths.

Theoretical framework

Sirexkat @ Flickr, CC-SA

© Distributed-cognition-maggie-nichols.jpg @ http://cyborganthropology.com/

Distributed cognition

Distributed cognition (Hutchins, 1995; Salomon, 1993) is is a view that cognition does not reside only in person’s head, but distributed among people, artifacts and symbols during thinking, reflection and learning (Salomon, 1993)

Physical distribution

Social distribution

Symbolic distribution

openclipart

The concept of cognitive tools is used to refer to any tool that can support aspects of learner’s cognitive processes (Lajoie, 1993). Jonassen and Reeves (1996) broadens Lajoie’s view of the term, using it to refer to any tools “than enhance the cognitive powers of human beings during thinking, problem solving, and learning” (p.693).

Cognitive tool

Tools for living vs. tools for learning

In order to fit world of distributed cognition where we live and role of mobile devices and applications within it appropriate framework is needed. One fitting approach for this purpose is a distributed view of thinking and learning suggested originally by D. Perkins (1996).

Person-soloPerson+Framework

Person+Surround

F(x)F(x)

F(x)

F(x)

F(x)

Person-solo

Person+Artefact

Higher-order knowledge

Distributed cognitive system

Knowledge

Representations

Retrieval

Construction

Access craharacteristics

F(x)

Exexutive function F(x)

F(x)

Tools for living

Tools for learning

Scaffolds Mindtools

F(x)

F(x)

F(x) F(x)

Executive function F(X)

Person+Surround

F(x)F(x)

F(x)

F(x)

F(x)

Person-solo

Person+Artefact

F(x) A system can further be characterized as dependent on which of its components has the executive function with respect to the task being accomplished.

In the distributed cognition model executive function is distributed by the nature – distributions happen in our surround all the time (Perkins, 1993)

Access characteristicsKnowledge

Representations

Retrieval

Construction

Collaborative learning

Creative Commons

Nature of the learning task is one crucial determinant of successful collaboration (Arvaja, Häkkinen, Eteläpelto, & Rasku-Puttonen, 2000). One of the everlasting challenges for instructional designers is to provide real group tasks and contexts that stimulate questioning, explaining and other forms of knowledge articulation (Järvelä, Häkkinen, Arvaja, & Leinonen, 2003). Such challenge is grounded to an idea that the authenticity of the learning situations and tasks is assumed to be an important factor that can facilitate higher order learning (Brown, Collins, & Duguid, 1989).

Ill-structured problem solving was a core task in all experiments in this thesis.

Collaboratively usable vs. collaborative…suggestion was made to divide tools into the collaboratively usable technology (in which software alone does not scaffold collaboration) and collaborative technology (in which software is designed specifically to support collaborative knowledge construction), based on the instructional and pedagogical aspects of tools (Lipponen & Lallimo, 2004)

Emergent technologiesSupporting learning with

The world is entering into the Age of Mobilism (Norris & Soloway, 2011). New technology tools fit more readily and naturally our lives; increasingly broad, inexpensive, and easy access to Internet wireless devices, and a variety of Web-based personal publishing and social software tools are making computing truly a ubiquitous and “continuous” part of our lives (Roush, 2005, p.49).

Furthermore many researchers have argued that educational use of emergent mobile devices have technological attributes, which provide unique technological, social and pedagogical affordances

Affordances

Evolution of the research on ”mobile learning”

webb-zahn @ Flickr, CC-SA

Scafffolding

XThe fact that students nowadays make use of different electronic devices, which are available ubiquitously and they are called digital natives doesn’t make them good users of the media they have their disposal.

In other words, Perkins (1993) and Salomon (1993) argue that learners do not automatically know how to take appropriate and measured advantage of computer tools when involved in cognitive activities with them.

jari.laru@oulu.fi 23

Scaffolding 1/2 • The concept of scaffolding was first introduced by Wood,

Bruner & Ross (1976) in order to define what kind of instructional processes enables novices to carry out tasks that are beyond their unassisted efforts, thus helping them achieve independent task competence.

• The theoretical foundation of scaffolding comes from ideas concerning the zone of proximal development (ZPD) and sociocultural perspective (Vygotsky, 1986; Wertch, 1998).

• Scaffolding techniques have been used successfully in a number of desktop tools (Quintana, Reiser, Davis, Krajcik, Fretz, Duncan, Kyza, Edelson, Soloway, 2004)

Interlearn (2005)

jari.laru@oulu.fi 24

(distributed) Scaffolding 2/2• Puntembekar & Kolodner (1998) have argud that models of

individual scaffolding are not necessarily applicaple to educational settings in which a group of learners is pursuing a common goal.

• Other up-to-date notions on scaffolding emphasize that it can take a variety of forms - it can be extended to cover physical artifacts and representations, which can serve as cognitive tools that mediate action (Palincsar, 1998; Wertch, 1998), but also to consider peers and social roles as scaffolding agents (Tabak, 2004; Puntembakar & Kolodner, 1998).

=>Puntembekar & Kolodnner (1998) have coined the term distributed scaffolding to refer to such instructional designs that sequence and integrate a variety of social and material supports.

Interlearn (2005)

Collaborative scripts

• With respect to challenges in collaborative learning, Kollar, Fischer & Hesse (2006) have distinguished two classes of scaffolds:

A) scaffolds that provide support on a conceptual level (microscript) B) scaffolds that provide support concerning interactive processes between the collaborators (macroscript)

• Especially in CSCL such scaffolds have been called collaboration scripts (Fischer et. al, 2007; MOSIL, 2004)

FadingChallenging concept of

Concept of fading reveals to be problematic!

• Pea (2004) suggested that scaffolds that do not incorporate fading are actually a part of distributed cognition, or the division of an overall cognitive task into subtasks that can be completed by different people or tools (Hutchins, 1995).

• Fading is not possible at all with it’s original meaning in computer mediated contexts (Belland, 2011)

Scaffolds as a part of distributed cognition

Scripts can be examined through the metaphor of distributed cognition which have been argued to appropriately to apply computer-based scaffolding: ”because the latter do no not simply, but fundamentally change the nature of cognition (Kollar, Fischer & Hesse, 2006; Belland, 2011)

Distributed cognition

Scaffolding

Distributed scaffolding

Scaffolding in the context ofDistributed cognitive system

Microscripts Macroscripts

Integrated learning scripts

Collaborative learning activities

FadingChallenging concept of

Person+Surround

F(x)F(x)

F(x)

F(x)

F(x)

Person-solo

Person+Artefact

Higher-order knowledge

Distributed cognitive system

Knowledge

Representations

Retrieval

Construction

Access craharacteristics

F(x)

Exexutive function F(x)

F(x)

Tools for living

Tools for learning

Scaffolds F(X) Mindtools

F(x)

F(x)

F(x) F(x)ag

Controlling agent(in this case it’sexternal CL script)

Internal CL scriptRemoval

Microscript

Person-soloPerson-solo

Internal script

Person+ Person+ Person+

Task(s)External script

A B

Team+

Microscript

Integrated learning script (macroscript)

Microscript

Internal script

Person-soloC

Internal script

External script

Group members

Deg

ree

of c

ongr

uenc

y

A

B

C

Emergent interaction pattern(s)

• In this thesis script is considered as method to be used during activities in different case studies, not as a pedagogical objective, with goal to be internalized for the future

• However, when script is a method, the internal script is instrumental to play well the external script (each students construct some internal script that will – to some – extent be different from the external script) (Dillenbourg & Jermann, 2006

Script as method

Didactic envelope(macroscripting)

In MOSIL project dillenbourg et. Al (2004) expanded the scope of collaboration scripts presented by Dillebourg (2002) to encompass more than just small group interaction by introducing concept of didactic envelope: “we discriminate the core script from its didactic envelope, i.e. a set of pre- and post-structuring activities” (p.13). Such structuring activities (e.g. introducing the topic, reflecting on what what was discussed, etc.) allow triggering the core mechanisms and enable scripts to be optimally integrated into the lesson plan and are an essential part of macroscripting (Dillenbourg & Jermann, 2006; Dillenbourg & Tchounikine, 2005).

Disturbing collaboration: ”swish”Dillenbourg & Hong (2008) have termed their approach for disturbing collaboration as “Split Where Interaction Should Happen” (SWISH). They summarize it by using three axioms: 1. Learning is result from interactions in which learners have

to engage in order to compensate split introduced by macroscript, i.e constructing shared solution based on materials gained during individual reflections (case study III in this thesis).

2. The nature of “split” thus determines the nature of interactions. Interactions are mechanisms for overcoming task splits

3. The splits can therefore – in reverse engineering- be designed to trigger the very interactions that the designer wants to foster

Person-solo Person-solo

Person-solo Person-solo

Internal script Internal script

Internal scriptInternal script

Macroscript

Microscript n

Microscript n+1

Person+ / team+

Pedagogical designLearning context

Person-solo

Poitinjimmie @ Flickr, CC-SA

Learning design

Aims of the study I/II

The general focus of this doctoral thesis is to apply theoretical ideas of distributed cognition and scaffolding for mobile computer supported collaborative learning in authentic contexts

Aims of the study II/II

• The first aim was to analyze the nature of collaboration in the mobile technology supported settings of collaborative learning (papers I and III-IV) and work (paper II)

• The second aim was to experiment the kind of scaffolding of mobile computer supported collaboration that can enrich collaborative learning (papers I,III-IV) and work (paper II)

• The third aim was to discuss the methodological issues for studying social interactions and collaborative learning in mobile computer supported activities (I-IV)

Research design

The research design combines chracteristics from the design based research (DBR), Case-study approach (Yin, 2003) and situated approach (Greeno, 2006)

FLE3mobile

SmartLibrary

Tiernajack

Flyers

Flyers

Script

Micro: sentence openers

Core activity

Knowledge building

Macro: Free collaboration

Instructional design

Theory

Foundations

Virtual master’s programme

FLE3

ContextMobile tool

Qualitative content analysis

Main methods

Social network analysis 1st analysis

2nd analysis

Mobile toolFieldtrip (K12)

Context

Script

Core activity

Argumentative inquiry learning

Integrated learning script

Instructional design

Theory

Progressive inquiry learningKnowledge building

Theory

Inquiry learningArgumentative learning

1st analysis

2nd analysis

Data

Logs, interviews, questionnaires

Macro: Storyboard + tutors

Micro: Sentence openers

Results & findings

Data

Flyers, audio recordings, mindmaps

Qualitative content analysis

Main methods

Mann Whitney U-test

Core activity

Blended learning EI

Script

Integrated learning script

Results & findings

Higher Education

Context

Instructional design

Mobile tool

Media uploader Google Reader

Social software (students)

Media sharing WeblogWikiRSS Reader

Tools

Social software (class)

Course blogCourse wikiRSS aggregats

1st analysis

Data

Log data, recordings, knowledge test, wiki history, content

Bayesian modeling

Main methods

Qualitative content analysis (mixed)

2nd analysis

Paper IV

Preliminary analysisPaper I

Paper II

Paper III

Theory (analytical lens)

Community of Practise

Case II

Case I

Case III

Laru, J. & Järvelä, S. (2008). Social patterns in mobile technology mediated collaboration among members of the professional distance education community. Educational Media International Journal, 45(1),17-3.

stimulus

Interviews

Log-files Social network analysis (SNA)

Content analysis

Database

Matrices

Results

Overall, the analyses revealed nonparticipative behaviour within the online community.

The social network analysis revealed structural holes and sparse collaboration among participants in the offline community. It was found that due to their separated practices in the offline community, they didn’t have a need for mobile collaboration tools in their practices.

Laru, J. & Järvelä, S. (2008). Social patterns in mobile technology mediated collaboration among members of the professional distance education community. Educational Media International Journal, 45(1),17-3.

STUDY 2: FLYERS

Audio recordings of argumentative discussions

Knowledge claim messages

Content analysis

Top-performers

Results

Mindmap Post-test

Mindmap Pre-test

Mindmap analysis

Low performers

Mann-Whitney U-test

Although the results revealed several shortcomings in the types of argumentation...

….In general, the use of the mobile tool likely promoted important interaction during inquiry learning, but led to superficial epistemological quality in the knowledge claim messages.

Course feed

WikiworkPhototaking

A.Ground C.ConceptualizeB.ReflectD. Reflect & elaborate

E. Review & evaluate F. Co-construct knowledge

G.Monitor Tools used to merge multiple RSS feeds

Merged feedsMultiple feeds

Course blog and wiki Mobile applications

Course level tools

Group level tools

Monitoring tools

Lecture Discussion Blogging Discussion

Collaborative Solo Collaborative

Phase:

Software:

Activity:

Figure 4. Socio-technological design of the course. The idea of making use of each others’knowledge was operationalized in socio-technical design. It consisted of recurrent individual and collective phases in which students used multiple Web 2.0 tools and mobile phones in concert to perform designed tasks. Retrieved from: Jari Laru, Piia Näykki, Sanna Järvelä, Supporting small-group learning using multiple Web 2.0 tools: A case study in the higher education context, The Internet and Higher Education, Available online 28 August 2011, ISSN 1096-7516, 10.1016/j.iheduc.2011.08.004.

Figure 5. Pedagogical design of the course. Groups were required to complete a wiki project by the end of the semester. In order to complete the wiki project, students needed to participate in recurrent solo and collective phases mediated by the use of social software tools and face-to-face discussions in their respective phases. Jari Laru, Piia Näykki, Sanna Järvelä, Supporting small-group learning using multiple Web 2.0 tools: A case study in the higher education context, The Internet and Higher Education, Available online 28 August 2011, ISSN 1096-7516, 10.1016/j.iheduc.2011.08.004.

Wordpress

Results

Wikispaces

Conceptual knowledge post-

test

Flickr Google ReaderShozuDiscussions

Conceptual knowledge post-

test

Knowledge test analysis

Analysis of the engagement

Paired samples t-test

Normalized learning gain

On task analysis

Analysis of Wiki history

Bayesian dependency & classification modeling

Descriptive analysis

In our case, we found that using social software tools together to perform multiple tasks likely increased individual knowledge acquisition during the course.

Bayesian classification analysis revealed that the best predictors of good learning outcomes were wiki-related activities.

Results

Overall, the analyses revealed nonparticipative behaviour within the online community.

The social network analysis revealed structural holes and sparse collaboration among participants in the offline community. It was found that due to their separated practices in the offline community, they didn’t have a need for mobile collaboration tools in their practices.

Single task, free collaboration

Collective task

Existing master’s programs

New master’s program

Knowledge building activity

Microscript

External script

Co-operative learning Microscript: Knowledge building Self-study

Activity: Collaborative knowledge building (progressive inquiry learning)

Internal script Internal script(s)

Activity: Argumentative knowledge construction

prestructuring

Core activity

Post structuring

- Integrated learning script -

CSCL Integrated learning script CSCL

SWISH

Activities: Conceptualizations, reflections, elaborations, knowledge construction

prestructuring

Core activity

Post structuring

Emergent interaction pattern(s)

Emergent interaction pattern(s)

Nonparticipativebehavior

ag

ag

agag

ag

Although the results revealed several shortcomings in the types of argumentation...

….In general, the use of the mobile tool likely promoted important interaction during inquiry learning, but led to superficial epistemological quality in the knowledge claim messages.

Task 1 Task 2 Task 3

Macroscript

Microscript

Storyboard Argumentative discussions

Review and comparisonConclusive discussion

Collective activities Small-group activities

Three independent tasks + didactic envelope

External script

Co-operative learning Microscript: Knowledge building Self-study

Activity: Collaborative knowledge building (progressive inquiry learning)

Internal script Internal script(s)

Activity: Argumentative knowledge construction

prestructuring

Core activity

Post structuring

- Integrated learning script -

CSCL Integrated learning script CSCL

SWISH

Activities: Conceptualizations, reflections, elaborations, knowledge construction

prestructuring

Core activity

Post structuring

Emergent interaction pattern(s)

Emergent interaction pattern(s)

Nonparticipativebehavior

ag

ag

agag

ag

In our case, we found that using social software tools together to perform multiple tasks likely increased individual knowledge acquisition during the course.

Bayesian classification analysis revealed that the best predictors of good learning outcomes were wiki-related activities.

Separated tasks w integrated subtasks+ Swish

Macro

1st topic

Grounding (Lecture) Reflect (Discussion)

Review and evaluate (Discussion)

Collective activities Small-group activities

2nd topic 3rd topic

Individual activities

Conceptualize (Photo [mobile])

Reflect and elaborate (Blog)

Co-construct of knowledge(Wiki)

Core activity

Micro S1

Micro S2

Micro S3

Micro S4

Micro S5

Monitor (RSS [mobile])

Collaboration script

Micro Sn = Microscript nMacro = Macroscript

External script

Co-operative learning Microscript: Knowledge building Self-study

Activity: Collaborative knowledge building (progressive inquiry learning)

Internal script Internal script(s)

Activity: Argumentative knowledge construction

prestructuring

Core activity

Post structuring

- Integrated learning script -

CSCL Integrated learning script CSCL

SWISH

Activities: Conceptualizations, reflections, elaborations, knowledge construction

prestructuring

Core activity

Post structuring

Emergent interaction pattern(s)

Emergent interaction pattern(s)

Nonparticipativebehavior

ag

ag

agag

ag

Person+Surround

F(x)F(x)

F(x)

F(x)

F(x)

Person-solo

Person+Artefact

Higher-order knowledge

Distributed cognitive system

Knowledge

Representations

Retrieval

Construction

Access craharacteristics

F(x)

Exexutive function F(x)

F(x)

Tools for living

Tools for learning

Scaffolds F(X) Mindtools

F(x)

F(x)

F(x) F(x)ag

Controlling agent(in this case it’sexternal CL script)

Internal CL scriptRemoval

Comparison between experiments

Topic

Topic 1

Topic 2

Topic 3

Topic 1

Topic 2

Topic 3

Subtask n Subtask n+1

Subtask n Subtask n+1

Subtask n Subtask n+1

Topic 1

Topic 2

Topic 3

Subtask n Subtask n+1

Subtask n Subtask n+1

Subtask n Subtask n+1

Didactic envelopes (Dillenbourg, 2002) Swish splits (Dillenbourg & Jermann, 2006)

X

X

Free

Conclusion

Simple tools, rich pedagogical practises (Roschelle, 2002)

• Kussakin osatutkimuksessa käytettiin ”parhaita menetelmiä” –

• Design tutkimus vs. Case study.. => • Menetelmät kehittyivät monipuolisemmiksi ja

tarkemmiksi.. Voisikos niitä selittää ilmiön kehittymisellä, menetelmien kehittymisellä..

• Kontekstien muuttuessa kollaboraatio voi muuttua.. Oletko ottanut huomioon..