SlideShare une entreprise Scribd logo
1  sur  37
Hello world!
My baby steps…
PhD: Methods and tools for the evaluation of collaborative learning
activities using time series (2014, University of Patras)
Greece!!!
Growing up…
Here we are!
Postdoc @ HCII since
April, 2016!
“Linking Dialogue with Student
Modeling to Create an Enhanced
Micro-adaptive Tutoring System”
University of Pittsburgh (LRDC)
&
CMU (HCII)
Dialogue as a means for learning
• We aim to develop an adaptive tutorial dialogue
system, guided by a student model that will support
students in learning physics
Research questions
• RQ: What makes tutorial dialogues successful?
Teachers’ adapt the level of discussion to the
student’s “zone of proximal development”
(Vygotsky)
Research questions
• RQ: What makes tutorial dialogues successful?
Teachers’ adapt the level of discussion to the
student’s “zone of proximal development”
(Vygotsky)
• RQ2: How do tutorial dialogues adapt to different
student characteristics and prior knowledge?
Degree of Teacher Control (van de Pol)
Contingent Tutoring (Pino-Pasternak)
Cognitive Complexity (Nystrand, Graesser)
An example would be nice….
RQ: What minimum acceleration must the
climber have in order for the rope not to
break while she is rappelling down the
cliff? (You do not have to come up with a
numerical answer. Just solve for "a"
without any substitution of numbers.)
Chip: a = f / m
T: what's f ?
Chip : f = mg
T: just mg ? how many forces act ont he
climber ?
Chip : mg + T
T: is mg down or up ?
Chip : down and T is up
T: ok so now solve for a again plugging in T
and mg
RQ: What minimum acceleration must the
climber …….
Dale: 500/55 kg=a m/s^2
T: I don't agree - that's the acceleration
that just the pull from the rope would
produce (well once the units are
straightened out it would be). Think a little
more. What is the general rule for finding
acceleration from forces?
Dale : F/m=a
T: and what is the F there?
Dale : tension?
T: No.. the F in F=ma is always the net force
on the object (or group of objects). The
vector sum of all the forces on the object. I
prefer to say "Sum of F= ma" because it's
easier to get it right. So.. if she is sliding
down and the rope is just short of breaking,
what is the *net* force on her?
High performer
Research Objective
To integrate a student model into a
tutorial dialogue system in order to
guide the dialogue more effectively
according to students’ needs
“more effectively”
• Analyze human-to-human tutorial dialogues
• Build a coding scheme to operationalize Level of
Support
The mechanics of tutorial discussions
“we need to identify and
group the features of dialogic
discourse to differentiate the
levels of support”
Method of the study
3 human-to-human dialogues on
Physics / 1 per overall learning
gains level [low/medium/high]
Level of Control
[3-step scale]
Question Category
[18 types]
Level of
Specificity
[3-step scale]
Contingent
Tutoring
[binary]
LOS
Coding
Scheme
Coding scheme - Application
• 4 coders
• 3 dialogues [low/medium/high]
• 19 tutor turns
• Introduction to the coding scheme
• Rating handbook & template
Coding scheme - Results
Dimension Fleiss’ Kappa p-value
Level of control 0.404 4.13e-11
Question category 0.395 0
Level of specificity 0.141 0.0245
Contingency 0.0764 0.415
Lessons learned:
Still unclear how teachers effectively
regulate the level of support
Coding scheme: Lessons learned
• Not easy to interpret the goal of the intervention
• One intervention, multiple goals
• Crucial features:
– New content
– Feedback Information / Information meant to push student
forward
– Degree of detail
Coding scheme
Adaptation and Evaluation
Before After
Level of Control Information related to
student’s answer
(Backward/Forward)
Hints Provision
Question Category Question Category
Level of Specificity Feedback on Correctness
Information related to
feedback
Contingency Contingency
Application and Evaluation
• 10 human-to-human tutorial dialogues
(Physics) – 3 High, 3 Low, 4 Medium
• 2 raters per dialogue
• The raters were given a tutorial on the coding
scheme and detailed instructions
Dimensions Cohen's kappa
D1. Information Provision (B) 0.871
D1. Information Provision (F) 0.843
D2. Hints Provision 0.843
D3. Feedback on correctness 0.826
D4. Information related to feedback 0.764
How to author dialogues
So, you’ve coded it… Now what?
Remember the Research Objective?
“To integrate a student model into a
tutorial dialogue system in order to guide
the dialogue more effectively according
to students’ needs”
Use a student model to decide on what
step and which piece of dialogue to give
next
The line of reasoning
*Jordan, P., Albacete, P., & Katz, S. Exploring Contingent Step Decomposition in a Tutorial Dialogue System.
Can you please tell
me what is the
vertical net force on
the arrow?
What does that mean
with respect to the
arrow’s vertical
acceleration?
RQ: “Suppose the archer is standing at the edge of a high cliff and
shoots his arrow perfectly horizontally with an initial velocity of 50
m/s. Neglecting air resistance, how will the vertical velocity of the
arrow vary during its flight until it eventually hits the ground? “
Move through the line of reasoning
*Jordan, P., Albacete, P., & Katz, S. Exploring Contingent Step Decomposition in a Tutorial Dialogue System.
We need an assessment of the specific skills(KCs) involved in the
specific steps (nodes) (Conati, 2010)
RQ: “Suppose the archer is standing at the edge of a high cliff and
shoots his arrow perfectly horizontally with an initial velocity of 50
m/s. Neglecting air resistance, how will the vertical velocity of the
arrow vary during its flight until it eventually hits the ground? “
High Support: I disagree. Look at
Newton's second law, Fnet = m*a. Is there
a vertical net force acting on the arrow
when it leaves the bow? In our case, the
Fnet is the force of gravity and is applied
on the arrow in the vertical direction. So
Fnet = mg and mg = ma
Low Support: Recall that you just said that
the force of gravity is applied on the
arrow. Would this force produce an
acceleration on the arrow as it leaves the
bow?
A couple of things to consider…
• How to model The Model?
• And how to evaluate it?
How do we model the Model?
• Regression models
• Bayesian networks
• Rule-based models
• Time-series models
• Overlay models (using NLP)
What challenges we face
• Not enough observations per skill
What challenges we face
• Not enough student input
What challenges we face
• Not enough observations per skill
• Not enough student content
• Real-time assessment and prediction requires
efficiency
What challenges we face
• Not enough observations per skill
• Not enough student content
• Real-time assessment and prediction requires
efficiency
How do we model the Model?
• Regression models
• Bayesian model (essentially Bayesian KT)
• Rule-based models
• Time-series models
• Overlay models (using NLP)
Regression models
• Advantages
– Good match for our datasets (multiple skills per step)
– Easy to implement /maintain (for new skills, scenarios)
• Disadvantages
– Multicollinearity issues might affect performance (i.e.
when the predictors highly correlate with each other)
– Overfitting
– Never used for real-time predictions before!
*Cen, H. (2009). Generalized learning factors analysis: improving cognitive models with machine learning.
*MacLellan, C. J., Liu, R., & Koedinger, K. R. (2015). Accounting for Slipping and Other False Negatives in Logistic Models
of Student Learning.International Educational Data Mining Society.
Bayesian models
• Advantages
– Good way to describe complex mechanisms
– Provide better understanding/reasoning for diagnosis
• Disadvantages
– Need of a pre-defined network for every scenario/activity
– Difficult to adapt/maintain (when new skills are added)
BUT
- Bayesian Knowledge Tracing has been used “online” for
inferring mastering of “main” skills per step…
*recent work on individualization/introducing student-
specific parameters into BKT
*Yudelson, M. V., Koedinger, K. R., & Gordon, G. J. (2013, July). Individualized bayesian knowledge tracing models.
In International Conference on Artificial Intelligence in Education (pp. 171-180). Springer Berlin Heidelberg
*Pardos, Z. A., & Heffernan, N. T. (2010, June). Modeling individualization in a bayesian networks implementation of
knowledge tracing. In International Conference on User Modeling, Adaptation, and Personalization (pp. 255-266).
Springer Berlin Heidelberg.
Evaluation
• How to identify the best student model (or
models)?
– Model parameters (when available) – aic, bic
– Error rates (RMSE, MAE) and Learning curves
(Corbett, et.al.,1995)
– Students’ input – observe students’ behavior,
interviews etc.
– Performance indicators[e.g. time, complexity]/cost
Work in progress: Early application of
modeling techniques on existing data
• 3 datasets  1 super set
Skills(Dynamics, Revoice, Summary)  217 skills
• We applied Regression & BKT on all datasets
Work in Progress: Results (per study)
Dynamics (N=5272)
BKT AFM PFM IFM
AIC 6816 4946 5029 4946
BIC 7922 6812 6895 7200
RMSE 0.458 0.396 0.401 0.399
Revoice (N=4613)
BKT AFM PFM IFM
AIC 6621 3954 3908 3913
BIC 7759 5746 5700 6117
RMSE 0.439 0.396 0.391 0.392
Summary (N=6209)
BKT AFM PFM IFM
AIC 7110 5821 5773 5797
BIC 8456 7861 7813 8302
RMSE 0.425 0.392 0.390 0.391
*The Akaike information criterion (AIC)
and the Bayesian information criterion
(BIC) are measures of the relative quality
of statistical models for a given set of
data.
Work in Progress: Results (super set)
Global (N=15644)
BKT AFM PFM IFM
AIC 7110 13103 13123 13056
BIC 8456 17229 17249 17836
RMSE 0.425 0.42 0.421 0.419
Prediction time
(sec) 0.125 0.253 0.223 0.229
Notes
• Overfitting
– For regression models: The more skills, the more
overfitting
– For BKT: practically the same even for large datasets
• Error
– For regression models the error increases along with
the dataset size
– For BKT, practically the same (or even better…)
– For large datasets, error(regression)≈error(BKT)
• Execution Time
– Regression models almost double than BKT
Tutor architecture and Student Model
answer
UpdateProfile
( student)
Compute Next
Steps’
Probabilities
Tutor
Next step
Next step
Profile(student)
Evaluation (answer)
Run Rules
Student Model
Next Steps
alternatives
Future Work
• Implementing the infrastructure for real-time
studies
• Field trials in lab and in classroom
• Further refinements…
But
The end.
If you want to know more, get in touch!
NSH 2602K
ichounta@cs.cmu.edu
(plus cat pictures!)

Contenu connexe

Similaire à Chounta@paws

Analysis of Human to Human Tutorial Dialogues: Insights for Teaching Analytics
Analysis of Human to Human Tutorial Dialogues: Insights for Teaching AnalyticsAnalysis of Human to Human Tutorial Dialogues: Insights for Teaching Analytics
Analysis of Human to Human Tutorial Dialogues: Insights for Teaching AnalyticsIrene-Angelica Chounta
 
formative e-assessment: a scoping study
formative e-assessment: a scoping studyformative e-assessment: a scoping study
formative e-assessment: a scoping studyYishay Mor
 
Adaptive Learning Systems: A review of Adaptation.
Adaptive Learning Systems: A review of Adaptation.Adaptive Learning Systems: A review of Adaptation.
Adaptive Learning Systems: A review of Adaptation.Peter Brusilovsky
 
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdf
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdfShibani Antonette_Augmenting pedagogic writing practice with CLAD.pdf
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdfShibani22
 
Seminar University of Loughborough: Using technology to support mathematics e...
Seminar University of Loughborough: Using technology to support mathematics e...Seminar University of Loughborough: Using technology to support mathematics e...
Seminar University of Loughborough: Using technology to support mathematics e...Christian Bokhove
 
Domain Modeling for Personalized Learning
Domain Modeling for Personalized LearningDomain Modeling for Personalized Learning
Domain Modeling for Personalized LearningPeter Brusilovsky
 
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovskyifi8106tlu
 
Classroom Assessment Techniques
Classroom Assessment TechniquesClassroom Assessment Techniques
Classroom Assessment Techniquesssorden
 
Action Sequence Mining and Behavior Pattern Analysis for User Modeling
Action Sequence Mining and Behavior Pattern Analysis for User ModelingAction Sequence Mining and Behavior Pattern Analysis for User Modeling
Action Sequence Mining and Behavior Pattern Analysis for User ModelingPeter Brusilovsky
 
Flipped education csm_2014_robertson
Flipped education csm_2014_robertsonFlipped education csm_2014_robertson
Flipped education csm_2014_robertsonEric Robertson
 
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...Design & Technology and Computer Science in the CAMAU Project: The Genesis of...
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...David Morrison-Love
 
Integrating an intelligent tutoring system into a virtual world
Integrating an intelligent tutoring system into a virtual worldIntegrating an intelligent tutoring system into a virtual world
Integrating an intelligent tutoring system into a virtual worldParvati Dev
 
Interactive engagement strategies for large classes
Interactive engagement strategies for large classesInteractive engagement strategies for large classes
Interactive engagement strategies for large classesSimon Bates
 
Blueprinting Liz Norman ANZCVS 2021
Blueprinting Liz Norman ANZCVS 2021Blueprinting Liz Norman ANZCVS 2021
Blueprinting Liz Norman ANZCVS 2021Liz Norman
 
Rae t4 d-knowledge-economy-sa-urs-dec2017
Rae t4 d-knowledge-economy-sa-urs-dec2017Rae t4 d-knowledge-economy-sa-urs-dec2017
Rae t4 d-knowledge-economy-sa-urs-dec2017MYRA School of Business
 
Approaches in educational technology
Approaches in educational technologyApproaches in educational technology
Approaches in educational technologyFaKhalid
 
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...Irene-Angelica Chounta
 

Similaire à Chounta@paws (20)

Analysis of Human to Human Tutorial Dialogues: Insights for Teaching Analytics
Analysis of Human to Human Tutorial Dialogues: Insights for Teaching AnalyticsAnalysis of Human to Human Tutorial Dialogues: Insights for Teaching Analytics
Analysis of Human to Human Tutorial Dialogues: Insights for Teaching Analytics
 
formative e-assessment: a scoping study
formative e-assessment: a scoping studyformative e-assessment: a scoping study
formative e-assessment: a scoping study
 
Adaptive Learning Systems: A review of Adaptation.
Adaptive Learning Systems: A review of Adaptation.Adaptive Learning Systems: A review of Adaptation.
Adaptive Learning Systems: A review of Adaptation.
 
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdf
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdfShibani Antonette_Augmenting pedagogic writing practice with CLAD.pdf
Shibani Antonette_Augmenting pedagogic writing practice with CLAD.pdf
 
Seminar University of Loughborough: Using technology to support mathematics e...
Seminar University of Loughborough: Using technology to support mathematics e...Seminar University of Loughborough: Using technology to support mathematics e...
Seminar University of Loughborough: Using technology to support mathematics e...
 
Domain Modeling for Personalized Learning
Domain Modeling for Personalized LearningDomain Modeling for Personalized Learning
Domain Modeling for Personalized Learning
 
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky
2016-05-31 Venia Legendi (CEITER): Sergey Sosnovsky
 
Classroom Assessment Techniques
Classroom Assessment TechniquesClassroom Assessment Techniques
Classroom Assessment Techniques
 
Action Sequence Mining and Behavior Pattern Analysis for User Modeling
Action Sequence Mining and Behavior Pattern Analysis for User ModelingAction Sequence Mining and Behavior Pattern Analysis for User Modeling
Action Sequence Mining and Behavior Pattern Analysis for User Modeling
 
Flipped education csm_2014_robertson
Flipped education csm_2014_robertsonFlipped education csm_2014_robertson
Flipped education csm_2014_robertson
 
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...Design & Technology and Computer Science in the CAMAU Project: The Genesis of...
Design & Technology and Computer Science in the CAMAU Project: The Genesis of...
 
Integrating an intelligent tutoring system into a virtual world
Integrating an intelligent tutoring system into a virtual worldIntegrating an intelligent tutoring system into a virtual world
Integrating an intelligent tutoring system into a virtual world
 
Interactive engagement strategies for large classes
Interactive engagement strategies for large classesInteractive engagement strategies for large classes
Interactive engagement strategies for large classes
 
Blueprinting Liz Norman ANZCVS 2021
Blueprinting Liz Norman ANZCVS 2021Blueprinting Liz Norman ANZCVS 2021
Blueprinting Liz Norman ANZCVS 2021
 
8th sem (1)
8th sem (1)8th sem (1)
8th sem (1)
 
Rae t4 d-knowledge-economy-sa-urs-dec2017
Rae t4 d-knowledge-economy-sa-urs-dec2017Rae t4 d-knowledge-economy-sa-urs-dec2017
Rae t4 d-knowledge-economy-sa-urs-dec2017
 
Approaches in educational technology
Approaches in educational technologyApproaches in educational technology
Approaches in educational technology
 
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...
Chounta Talk @HCII: Linking Dialogue with Student Modeling to Create an Enhan...
 
Concept attainment model
Concept attainment modelConcept attainment model
Concept attainment model
 
Learning Analytics for MOOCs: EMMA case
Learning Analytics for MOOCs: EMMA caseLearning Analytics for MOOCs: EMMA case
Learning Analytics for MOOCs: EMMA case
 

Plus de Irene-Angelica Chounta

Will time tell? Exploring the relationship between step duration and student ...
Will time tell? Exploring the relationship between step duration and student ...Will time tell? Exploring the relationship between step duration and student ...
Will time tell? Exploring the relationship between step duration and student ...Irene-Angelica Chounta
 
Modeling the Zone of Proximal Development with a Computational Approach
Modeling the Zone of Proximal Development with a Computational ApproachModeling the Zone of Proximal Development with a Computational Approach
Modeling the Zone of Proximal Development with a Computational ApproachIrene-Angelica Chounta
 
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...Building Arguments Together or Alone?Using Learning Analytics to Study the Co...
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...Irene-Angelica Chounta
 
An adaptive tutoring system for physics using reflective dialogue
An adaptive tutoring system for physics using reflective dialogueAn adaptive tutoring system for physics using reflective dialogue
An adaptive tutoring system for physics using reflective dialogueIrene-Angelica Chounta
 
LASAD Grammar App: A collaborative application to support students who study ...
LASAD Grammar App: A collaborative application to support students who study ...LASAD Grammar App: A collaborative application to support students who study ...
LASAD Grammar App: A collaborative application to support students who study ...Irene-Angelica Chounta
 
Let's argue over it: Are argumentation skills better learned collaboratively ...
Let's argue over it: Are argumentation skills better learned collaboratively ...Let's argue over it: Are argumentation skills better learned collaboratively ...
Let's argue over it: Are argumentation skills better learned collaboratively ...Irene-Angelica Chounta
 
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project..."From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...Irene-Angelica Chounta
 
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...Multilevel analysis of collaborative activities based on a Mobile Learning Sc...
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...Irene-Angelica Chounta
 
Two make a network: using network graphs to assess the quality of collaborati...
Two make a network: using network graphs to assess the quality of collaborati...Two make a network: using network graphs to assess the quality of collaborati...
Two make a network: using network graphs to assess the quality of collaborati...Irene-Angelica Chounta
 
It's all about time: towards the real time evaluation of collaborative activi...
It's all about time: towards the real time evaluation of collaborative activi...It's all about time: towards the real time evaluation of collaborative activi...
It's all about time: towards the real time evaluation of collaborative activi...Irene-Angelica Chounta
 
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ..."Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...Irene-Angelica Chounta
 
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013Irene-Angelica Chounta
 
Time series analysis of collaborative activities-CRIWG2012
Time series analysis of collaborative activities-CRIWG2012Time series analysis of collaborative activities-CRIWG2012
Time series analysis of collaborative activities-CRIWG2012Irene-Angelica Chounta
 

Plus de Irene-Angelica Chounta (18)

La tartu
La tartuLa tartu
La tartu
 
Will time tell? Exploring the relationship between step duration and student ...
Will time tell? Exploring the relationship between step duration and student ...Will time tell? Exploring the relationship between step duration and student ...
Will time tell? Exploring the relationship between step duration and student ...
 
Modeling the Zone of Proximal Development with a Computational Approach
Modeling the Zone of Proximal Development with a Computational ApproachModeling the Zone of Proximal Development with a Computational Approach
Modeling the Zone of Proximal Development with a Computational Approach
 
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...Building Arguments Together or Alone?Using Learning Analytics to Study the Co...
Building Arguments Together or Alone?Using Learning Analytics to Study the Co...
 
An adaptive tutoring system for physics using reflective dialogue
An adaptive tutoring system for physics using reflective dialogueAn adaptive tutoring system for physics using reflective dialogue
An adaptive tutoring system for physics using reflective dialogue
 
LASAD Grammar App: A collaborative application to support students who study ...
LASAD Grammar App: A collaborative application to support students who study ...LASAD Grammar App: A collaborative application to support students who study ...
LASAD Grammar App: A collaborative application to support students who study ...
 
Let's argue over it: Are argumentation skills better learned collaboratively ...
Let's argue over it: Are argumentation skills better learned collaboratively ...Let's argue over it: Are argumentation skills better learned collaboratively ...
Let's argue over it: Are argumentation skills better learned collaboratively ...
 
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project..."From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...
"From Making to Learning" : Dev Camps as a Blueprint for Re-inventing Project...
 
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...Multilevel analysis of collaborative activities based on a Mobile Learning Sc...
Multilevel analysis of collaborative activities based on a Mobile Learning Sc...
 
Two make a network: using network graphs to assess the quality of collaborati...
Two make a network: using network graphs to assess the quality of collaborati...Two make a network: using network graphs to assess the quality of collaborati...
Two make a network: using network graphs to assess the quality of collaborati...
 
It's all about time: towards the real time evaluation of collaborative activi...
It's all about time: towards the real time evaluation of collaborative activi...It's all about time: towards the real time evaluation of collaborative activi...
It's all about time: towards the real time evaluation of collaborative activi...
 
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ..."Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...
"Μέθοδοι και εργαλεία αξιολόγησης συνεργατικής μάθησης με χρήση χρονοσειρών" ...
 
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013
Chountaetal - team-gaming activity analysis - @ectel meets ecscw 2013
 
Time series analysis of collaborative activities-CRIWG2012
Time series analysis of collaborative activities-CRIWG2012Time series analysis of collaborative activities-CRIWG2012
Time series analysis of collaborative activities-CRIWG2012
 
Chounta avouris limassol2011
Chounta avouris limassol2011Chounta avouris limassol2011
Chounta avouris limassol2011
 
Chounta avouris arv2011
Chounta avouris arv2011Chounta avouris arv2011
Chounta avouris arv2011
 
Incos2010 irene
Incos2010 ireneIncos2010 irene
Incos2010 irene
 
Katsini etal 2010_sfhmmy_postersmall
Katsini etal 2010_sfhmmy_postersmallKatsini etal 2010_sfhmmy_postersmall
Katsini etal 2010_sfhmmy_postersmall
 

Dernier

Mixin Classes in Odoo 17 How to Extend Models Using Mixin Classes
Mixin Classes in Odoo 17  How to Extend Models Using Mixin ClassesMixin Classes in Odoo 17  How to Extend Models Using Mixin Classes
Mixin Classes in Odoo 17 How to Extend Models Using Mixin ClassesCeline George
 
Sociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning ExhibitSociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning Exhibitjbellavia9
 
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...pradhanghanshyam7136
 
How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17Celine George
 
Google Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptxGoogle Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptxDr. Sarita Anand
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdfQucHHunhnh
 
Food safety_Challenges food safety laboratories_.pdf
Food safety_Challenges food safety laboratories_.pdfFood safety_Challenges food safety laboratories_.pdf
Food safety_Challenges food safety laboratories_.pdfSherif Taha
 
How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17Celine George
 
Python Notes for mca i year students osmania university.docx
Python Notes for mca i year students osmania university.docxPython Notes for mca i year students osmania university.docx
Python Notes for mca i year students osmania university.docxRamakrishna Reddy Bijjam
 
FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024Elizabeth Walsh
 
Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...Association for Project Management
 
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxSKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxAmanpreet Kaur
 
Dyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxDyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxcallscotland1987
 
ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701bronxfugly43
 
ICT role in 21st century education and it's challenges.
ICT role in 21st century education and it's challenges.ICT role in 21st century education and it's challenges.
ICT role in 21st century education and it's challenges.MaryamAhmad92
 
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptxMaritesTamaniVerdade
 
Fostering Friendships - Enhancing Social Bonds in the Classroom
Fostering Friendships - Enhancing Social Bonds  in the ClassroomFostering Friendships - Enhancing Social Bonds  in the Classroom
Fostering Friendships - Enhancing Social Bonds in the ClassroomPooky Knightsmith
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxAreebaZafar22
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and ModificationsMJDuyan
 

Dernier (20)

Mixin Classes in Odoo 17 How to Extend Models Using Mixin Classes
Mixin Classes in Odoo 17  How to Extend Models Using Mixin ClassesMixin Classes in Odoo 17  How to Extend Models Using Mixin Classes
Mixin Classes in Odoo 17 How to Extend Models Using Mixin Classes
 
Sociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning ExhibitSociology 101 Demonstration of Learning Exhibit
Sociology 101 Demonstration of Learning Exhibit
 
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...Kodo Millet  PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
Kodo Millet PPT made by Ghanshyam bairwa college of Agriculture kumher bhara...
 
How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17How to Create and Manage Wizard in Odoo 17
How to Create and Manage Wizard in Odoo 17
 
Google Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptxGoogle Gemini An AI Revolution in Education.pptx
Google Gemini An AI Revolution in Education.pptx
 
1029-Danh muc Sach Giao Khoa khoi 6.pdf
1029-Danh muc Sach Giao Khoa khoi  6.pdf1029-Danh muc Sach Giao Khoa khoi  6.pdf
1029-Danh muc Sach Giao Khoa khoi 6.pdf
 
Food safety_Challenges food safety laboratories_.pdf
Food safety_Challenges food safety laboratories_.pdfFood safety_Challenges food safety laboratories_.pdf
Food safety_Challenges food safety laboratories_.pdf
 
How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17How to Give a Domain for a Field in Odoo 17
How to Give a Domain for a Field in Odoo 17
 
Python Notes for mca i year students osmania university.docx
Python Notes for mca i year students osmania university.docxPython Notes for mca i year students osmania university.docx
Python Notes for mca i year students osmania university.docx
 
FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024FSB Advising Checklist - Orientation 2024
FSB Advising Checklist - Orientation 2024
 
Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...Making communications land - Are they received and understood as intended? we...
Making communications land - Are they received and understood as intended? we...
 
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptxSKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
SKILL OF INTRODUCING THE LESSON MICRO SKILLS.pptx
 
Dyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptxDyslexia AI Workshop for Slideshare.pptx
Dyslexia AI Workshop for Slideshare.pptx
 
ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701ComPTIA Overview | Comptia Security+ Book SY0-701
ComPTIA Overview | Comptia Security+ Book SY0-701
 
ICT role in 21st century education and it's challenges.
ICT role in 21st century education and it's challenges.ICT role in 21st century education and it's challenges.
ICT role in 21st century education and it's challenges.
 
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
2024-NATIONAL-LEARNING-CAMP-AND-OTHER.pptx
 
Fostering Friendships - Enhancing Social Bonds in the Classroom
Fostering Friendships - Enhancing Social Bonds  in the ClassroomFostering Friendships - Enhancing Social Bonds  in the Classroom
Fostering Friendships - Enhancing Social Bonds in the Classroom
 
ICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptxICT Role in 21st Century Education & its Challenges.pptx
ICT Role in 21st Century Education & its Challenges.pptx
 
Mehran University Newsletter Vol-X, Issue-I, 2024
Mehran University Newsletter Vol-X, Issue-I, 2024Mehran University Newsletter Vol-X, Issue-I, 2024
Mehran University Newsletter Vol-X, Issue-I, 2024
 
Understanding Accommodations and Modifications
Understanding  Accommodations and ModificationsUnderstanding  Accommodations and Modifications
Understanding Accommodations and Modifications
 

Chounta@paws

  • 2. My baby steps… PhD: Methods and tools for the evaluation of collaborative learning activities using time series (2014, University of Patras) Greece!!!
  • 4. Here we are! Postdoc @ HCII since April, 2016!
  • 5. “Linking Dialogue with Student Modeling to Create an Enhanced Micro-adaptive Tutoring System” University of Pittsburgh (LRDC) & CMU (HCII)
  • 6. Dialogue as a means for learning • We aim to develop an adaptive tutorial dialogue system, guided by a student model that will support students in learning physics
  • 7. Research questions • RQ: What makes tutorial dialogues successful? Teachers’ adapt the level of discussion to the student’s “zone of proximal development” (Vygotsky)
  • 8. Research questions • RQ: What makes tutorial dialogues successful? Teachers’ adapt the level of discussion to the student’s “zone of proximal development” (Vygotsky) • RQ2: How do tutorial dialogues adapt to different student characteristics and prior knowledge? Degree of Teacher Control (van de Pol) Contingent Tutoring (Pino-Pasternak) Cognitive Complexity (Nystrand, Graesser)
  • 9. An example would be nice…. RQ: What minimum acceleration must the climber have in order for the rope not to break while she is rappelling down the cliff? (You do not have to come up with a numerical answer. Just solve for "a" without any substitution of numbers.) Chip: a = f / m T: what's f ? Chip : f = mg T: just mg ? how many forces act ont he climber ? Chip : mg + T T: is mg down or up ? Chip : down and T is up T: ok so now solve for a again plugging in T and mg RQ: What minimum acceleration must the climber ……. Dale: 500/55 kg=a m/s^2 T: I don't agree - that's the acceleration that just the pull from the rope would produce (well once the units are straightened out it would be). Think a little more. What is the general rule for finding acceleration from forces? Dale : F/m=a T: and what is the F there? Dale : tension? T: No.. the F in F=ma is always the net force on the object (or group of objects). The vector sum of all the forces on the object. I prefer to say "Sum of F= ma" because it's easier to get it right. So.. if she is sliding down and the rope is just short of breaking, what is the *net* force on her? High performer
  • 10. Research Objective To integrate a student model into a tutorial dialogue system in order to guide the dialogue more effectively according to students’ needs “more effectively”
  • 11. • Analyze human-to-human tutorial dialogues • Build a coding scheme to operationalize Level of Support The mechanics of tutorial discussions “we need to identify and group the features of dialogic discourse to differentiate the levels of support”
  • 12. Method of the study 3 human-to-human dialogues on Physics / 1 per overall learning gains level [low/medium/high] Level of Control [3-step scale] Question Category [18 types] Level of Specificity [3-step scale] Contingent Tutoring [binary] LOS Coding Scheme
  • 13. Coding scheme - Application • 4 coders • 3 dialogues [low/medium/high] • 19 tutor turns • Introduction to the coding scheme • Rating handbook & template
  • 14. Coding scheme - Results Dimension Fleiss’ Kappa p-value Level of control 0.404 4.13e-11 Question category 0.395 0 Level of specificity 0.141 0.0245 Contingency 0.0764 0.415 Lessons learned: Still unclear how teachers effectively regulate the level of support
  • 15. Coding scheme: Lessons learned • Not easy to interpret the goal of the intervention • One intervention, multiple goals • Crucial features: – New content – Feedback Information / Information meant to push student forward – Degree of detail
  • 16. Coding scheme Adaptation and Evaluation Before After Level of Control Information related to student’s answer (Backward/Forward) Hints Provision Question Category Question Category Level of Specificity Feedback on Correctness Information related to feedback Contingency Contingency
  • 17. Application and Evaluation • 10 human-to-human tutorial dialogues (Physics) – 3 High, 3 Low, 4 Medium • 2 raters per dialogue • The raters were given a tutorial on the coding scheme and detailed instructions Dimensions Cohen's kappa D1. Information Provision (B) 0.871 D1. Information Provision (F) 0.843 D2. Hints Provision 0.843 D3. Feedback on correctness 0.826 D4. Information related to feedback 0.764 How to author dialogues
  • 18. So, you’ve coded it… Now what? Remember the Research Objective? “To integrate a student model into a tutorial dialogue system in order to guide the dialogue more effectively according to students’ needs” Use a student model to decide on what step and which piece of dialogue to give next
  • 19. The line of reasoning *Jordan, P., Albacete, P., & Katz, S. Exploring Contingent Step Decomposition in a Tutorial Dialogue System. Can you please tell me what is the vertical net force on the arrow? What does that mean with respect to the arrow’s vertical acceleration? RQ: “Suppose the archer is standing at the edge of a high cliff and shoots his arrow perfectly horizontally with an initial velocity of 50 m/s. Neglecting air resistance, how will the vertical velocity of the arrow vary during its flight until it eventually hits the ground? “
  • 20. Move through the line of reasoning *Jordan, P., Albacete, P., & Katz, S. Exploring Contingent Step Decomposition in a Tutorial Dialogue System. We need an assessment of the specific skills(KCs) involved in the specific steps (nodes) (Conati, 2010) RQ: “Suppose the archer is standing at the edge of a high cliff and shoots his arrow perfectly horizontally with an initial velocity of 50 m/s. Neglecting air resistance, how will the vertical velocity of the arrow vary during its flight until it eventually hits the ground? “ High Support: I disagree. Look at Newton's second law, Fnet = m*a. Is there a vertical net force acting on the arrow when it leaves the bow? In our case, the Fnet is the force of gravity and is applied on the arrow in the vertical direction. So Fnet = mg and mg = ma Low Support: Recall that you just said that the force of gravity is applied on the arrow. Would this force produce an acceleration on the arrow as it leaves the bow?
  • 21. A couple of things to consider… • How to model The Model? • And how to evaluate it?
  • 22. How do we model the Model? • Regression models • Bayesian networks • Rule-based models • Time-series models • Overlay models (using NLP)
  • 23. What challenges we face • Not enough observations per skill
  • 24. What challenges we face • Not enough student input
  • 25. What challenges we face • Not enough observations per skill • Not enough student content • Real-time assessment and prediction requires efficiency
  • 26. What challenges we face • Not enough observations per skill • Not enough student content • Real-time assessment and prediction requires efficiency
  • 27. How do we model the Model? • Regression models • Bayesian model (essentially Bayesian KT) • Rule-based models • Time-series models • Overlay models (using NLP)
  • 28. Regression models • Advantages – Good match for our datasets (multiple skills per step) – Easy to implement /maintain (for new skills, scenarios) • Disadvantages – Multicollinearity issues might affect performance (i.e. when the predictors highly correlate with each other) – Overfitting – Never used for real-time predictions before! *Cen, H. (2009). Generalized learning factors analysis: improving cognitive models with machine learning. *MacLellan, C. J., Liu, R., & Koedinger, K. R. (2015). Accounting for Slipping and Other False Negatives in Logistic Models of Student Learning.International Educational Data Mining Society.
  • 29. Bayesian models • Advantages – Good way to describe complex mechanisms – Provide better understanding/reasoning for diagnosis • Disadvantages – Need of a pre-defined network for every scenario/activity – Difficult to adapt/maintain (when new skills are added) BUT - Bayesian Knowledge Tracing has been used “online” for inferring mastering of “main” skills per step… *recent work on individualization/introducing student- specific parameters into BKT *Yudelson, M. V., Koedinger, K. R., & Gordon, G. J. (2013, July). Individualized bayesian knowledge tracing models. In International Conference on Artificial Intelligence in Education (pp. 171-180). Springer Berlin Heidelberg *Pardos, Z. A., & Heffernan, N. T. (2010, June). Modeling individualization in a bayesian networks implementation of knowledge tracing. In International Conference on User Modeling, Adaptation, and Personalization (pp. 255-266). Springer Berlin Heidelberg.
  • 30. Evaluation • How to identify the best student model (or models)? – Model parameters (when available) – aic, bic – Error rates (RMSE, MAE) and Learning curves (Corbett, et.al.,1995) – Students’ input – observe students’ behavior, interviews etc. – Performance indicators[e.g. time, complexity]/cost
  • 31. Work in progress: Early application of modeling techniques on existing data • 3 datasets  1 super set Skills(Dynamics, Revoice, Summary)  217 skills • We applied Regression & BKT on all datasets
  • 32. Work in Progress: Results (per study) Dynamics (N=5272) BKT AFM PFM IFM AIC 6816 4946 5029 4946 BIC 7922 6812 6895 7200 RMSE 0.458 0.396 0.401 0.399 Revoice (N=4613) BKT AFM PFM IFM AIC 6621 3954 3908 3913 BIC 7759 5746 5700 6117 RMSE 0.439 0.396 0.391 0.392 Summary (N=6209) BKT AFM PFM IFM AIC 7110 5821 5773 5797 BIC 8456 7861 7813 8302 RMSE 0.425 0.392 0.390 0.391 *The Akaike information criterion (AIC) and the Bayesian information criterion (BIC) are measures of the relative quality of statistical models for a given set of data.
  • 33. Work in Progress: Results (super set) Global (N=15644) BKT AFM PFM IFM AIC 7110 13103 13123 13056 BIC 8456 17229 17249 17836 RMSE 0.425 0.42 0.421 0.419 Prediction time (sec) 0.125 0.253 0.223 0.229
  • 34. Notes • Overfitting – For regression models: The more skills, the more overfitting – For BKT: practically the same even for large datasets • Error – For regression models the error increases along with the dataset size – For BKT, practically the same (or even better…) – For large datasets, error(regression)≈error(BKT) • Execution Time – Regression models almost double than BKT
  • 35. Tutor architecture and Student Model answer UpdateProfile ( student) Compute Next Steps’ Probabilities Tutor Next step Next step Profile(student) Evaluation (answer) Run Rules Student Model Next Steps alternatives
  • 36. Future Work • Implementing the infrastructure for real-time studies • Field trials in lab and in classroom • Further refinements… But
  • 37. The end. If you want to know more, get in touch! NSH 2602K ichounta@cs.cmu.edu (plus cat pictures!)

Notes de l'éditeur

  1. I was born and raised in Greece where I studied electrical and computer engineering and did my phD in the university of patras. My phd was on CSCL and in particular I studied the use of time series to model and assess collaboration quality for collaborative learning activities
  2. After my phD, I moved to germany where I worked as a postdoc at the university of duisburg-essen with prof. Ulrich Hoppe and the Collide group. During my time there, we focused on the use of learning analytics for mirroring and guiding purposes. In particular we studied the use of network analytics to support communication in MOOCs and the use of interaction analysis to support coordination in maker events & hackathons.
  3. On April 2016 I moved in Pittsburgh and ever since I am a proud postdoctoral researcher at the HCII institute at the Carnegie Mellon University Here I work on a collaborative project between the university of pittsburgh and cmu
  4. The project’s goal is to link dialogue with student modeling in order to create an enhanced micro-adaptive tutoring system. I am very lucky to work with excellent researchers such as pamela jordan, sandra katz and particia abacete from the learning research and design center of the university of pittsburgh and bruce mc laren from cmu
  5. In particular we want to use an intelligent tutor to support students who learn physics through dialogue. In contrast to other ITS and cognitive tutors where the student has the initiative and the tutor responds to student input by providing recommendations and feedback, in our system the tutor is responsible for guiding the student through the learning scenarios. So what we practically aim to do is to build a physics story teller that will guide the student through adaptive lines of reasoning. And by “adaptive” we mean that we want to adapt the dialogue both in terms of structure but also in terms of content and depending on the student’s characteristics. The project is building on top of the successful practice of RIMAC that is a physics tutorial dialogue system developed by the university of pittsburgh and has been widely used to support school students. What is particularly interesting and innovative in this project is that -to our knowledge- no other adaptive dialogue tutor has used a student model to guide the student appropriately.
  6. There are two fundamental research questions that guide our project. The first one is: what makes tutorial dialogues successful. One can argue that learning is a social process and therefore the interaction between the teacher and the student can benefit and scaffold learning. What is also very important to remember here is that the teachers have the ability to assess their students’ performance and knowledge and to adapt the conversation to the right level in order not to set the bar too high but at the same time challenge the student, that is to address the student within her zone of proximal development where actually learning takes place.
  7. Our second research question focuses on what changes during the adaptation of the tutorial dialogues to different characteristics? What are the specific factors and dialogical features that adapt to match the zone of proximal development A lot of research has been done in this field and some studies pointed out important dimensions that affect tutorial dialogues such as the level of teacher control, the contingent tutoring structure of dialogues and the cognitive complexity
  8. These are two dialogues from two different students on the same reflection question and with the same tutor. The differences between those dialogues are obvious even by visual inspection. The one on the left is limited, using few words and symbols while the one on the right is extended and elaborate, the teacher uses terminology and provides explicit feedback, challenges the student to try harder etc. For the record, the dialogue on the right belongs to a high performing student
  9. To come back to our research, our objective is to integrate a student model in order to guide tutorial dialogue more effectively. And this brings us to our second question: what means “more effectively”?
  10. In order to explore deeper the mechanics of tutorial discussions we analyzed human to human tutorial dialogues in order to identify the features of dialogic discourse that change to support students of different levels. Additionally, we attempted to build a coding scheme to operationalize –using these features- the level of support, i.e. the way we need to address a student during a tutorial dialogue in order to successfully scaffold the learning process.
  11. As a first attempt, we analyzed 3 h2h dialogues that belong to different learning gains. For this analysis we applied different coding schemes that came up from our literature review. In particular we code our dialogues on four dimensions, the level of control which describes how much new content the teacher provides and how much control the teacher has over the students’ answer (I,e, whether he gives our the answer or tries to elicitate it etc. The level of specificity - this dimension refers to whether the tutor provided specific and focused information to the student. The contigent tutoring dimension - contingent tutoring takes place when the tutor challenges the student with questions or comments that are at or above her potential and non-contingent tutoring happens when the tutor poses questions and tasks that are lower than the student’s potential. The question category – 18 question categories (verification, concept completion etc) to code teacher’s intervention
  12. This is an example of what our coding scheme looked like
  13. However the results were discouraging. It was still hard for us to define how teachers effectively adapt the dialogue to the appropriate level for tutoring and in addition, we also discovered several factors that we didn’t take into account in this first approach
  14. These are some of the lessons learned: The Analysis revealed that experts usually had different opinions about what the goal of the intervention was. In some cases, they even stated that most likely the teacher didn’t have a specific goal but was instead trying to assess the student’s knowledge state. Frequently, the experts stated that a specific intervention served multiple goals. Teachers often provided feedback and guidance in one dialogue turn and this caused mismatches to the codings From the analysis we also identified important features that should be include into the coding schema such as the amount of new content, the provision of feedback and the degree of detail the teacher gives and requires from the answer.
  15. Provide appropriate, adaptive dialogic support: What is appropriate for whom? guidelines for feedback provision
  16. So now, we have a rough idea what “effective support” means in terms of tutorial dialogues and what are the features of tutorial discourse that are adapted to match students levels of knowledge and understanding. And we know how *roughly* to write dialogues for students of different knowledge levels. However our research objective is to use a model in order to guide students through the dialogue and what this means is that we need of a student model that will decide the optimal line of reasoning based on student’s performance and skills
  17. What we see here is an extract of the plan network of a tutorial dialogue for responding to the RQ, taken from Pam’s, Sandy’s and Patty’s recent paper where they argued when it is helpful to expand the dialogue and when not. I use it here to describe the application of the student model (SM). The dialogue always begins with a reflection question that the tutor sets to the student. Depending on the student’s answer we choose how the tutor should respond and where should the tutor move the dialogue choosing from a pool of alternatives
  18. By initiating a special dialogue depending on the current assessment of the student (Low/Medium/High) – and for that we need an overall assessment of the student profile (i.e. not only when it comes to particular skills but the general picture) –macro-assessment and/or By choosing appropriate nodes, i.e. node 11 instead of node 4 (again depending on the level of the student) – and here we need an assessment for the skills involved in this step – micro-assessment An additional question here is whether these question “levels” also represent concepts of different levels (this is where it links to Patricia’s and Dennis’ hierarchies). This is something that should be answered by the content experts and the teachers. To my understanding, when you decompose a concept you move down to more basic concepts. I do not know whether this is the case here.
  19. These are the modeling approaches that we have taken into consideration, so far. First, I will go through the ones that I do not consider as appropriate or good enough and my reasonings.
  20. These are the modeling approaches that we have taken into consideration, so far. First, I will go through the ones that I do not consider as appropriate or good enough and my reasonings.
  21. .
  22. The idea here is to propose a general architecture in which any of the models can be inserted and tested. The model here is implemented as a “plug-in” for the tutor (on the right), using as input the student profile, the evaluation of student answer in the current turn, the possible next steps (entry points). Depending on the type of the model, additional information on the entry points might be required (i.e. related KCs etc). As an output, the model will provide an updated profile of the student and the proposed next step.