Advancing the design of knowledge-building software
1 like651 views
The document discusses advancements in knowledge-building software aimed at enhancing collaborative learning through multiple literacies and public knowledge processes. It focuses on the development of tools for identifying promising ideas in student discourse, detailing design-based research and pilot studies that evaluate these tools' effectiveness. Future directions include improving scaffolding techniques, semantic analysis, and automated detection of misconceptions to foster deeper cognitive engagement.
Advancing the design of knowledge-building software
- 1. Advancing the Design of Knowledge-Building Software Bodong Chen, Monica Resendes, Maria Chuy, Katerine Bielaczyc, Huang-Yao Hong, Marlene Scardamalia, Carl Bereiter
- 2. Multiple Perspectives, Creating Connections Advanced Knowledge Embedded and Multiple Literacies, and Public Knowledge Processes, Rise-Above transformative And Teamwork and Improvable Ideas assessment Graphical view Scaffolds Rise-above Analytic tools Multimedia note Review Build-on Revision Co-authorship Annotation Problem Keyword Reference
- 4. Overview • Tagging Key and Promising Ideas • Scaffolds • Visualization • Concurrent, embedded, and transformative assessment
- 5. Promising Ideas tool Tagging Key and Promising Ideas
- 6. Keywords
- 7. To Tag Promising Ideas? • Promisingness evaluation (Bereiter & Scardamalia, 1993; Bereiter, 2002) • A need to identify promising ideas A case study on Grade 3s’ discourse
- 8. Promising Ideas tool • Design-based research (Brown, 1992; Collins, 1992) Identifying ideas: 1. Click 2. Choose a color 3. Highlight an idea 4. The idea gets highlighted and goes to the box
- 9. Promising Ideas list: Promising Ideas tool 1. Promising ideas listed 2. Choose a specific color list 3. Number of hits 4. Link backwards to notes
- 10. Pilot study 1: Grade 5/6 (Chen et al., 2011) • 83 “promising ideas” from 207 notes 70.00% 60.00% 50.00% 40.00% 30.00% 20.00% 10.00% 0.00% Theories Facts/Evidence Questions
- 11. Export ideas to a new view Export selected ideas to a new view
- 12. Pilot study 2: Grade 3 • Promisingness treatment module • “Progressive use” – two iterations of “writing-tagging-exporting” • Preliminary results – less facts/evidence – improved coverage of domain content 0.3 Cosine similarity measures 0.25 0.2 0.15 soil2011old 0.1 soil2011new 0.05 0 soilbithere soilcurr soilepa soilusda soilwiki Expert corpus
- 14. 1. Frequency of scaffold use Scaffold Tracker 2. Embellishment icons 3. Dynamic filter 4. Live update
- 15. Future directions • “Dynamic promotion” of scaffolds • Epistemic agency?
- 16. Content Classification Semantic Analysis Cognitive Diagnostic Assessment Concurrent, Embedded, and Transformative Assessment
- 17. Automated detection of “ways of contributing”? • Content classification • TagHelper (Rosé et al, 2008)
- 18. Automated detection of misconceptions? • Misconceptions in KF – Healthcare (Lax et al., 2010) Source: http://www.knowledgeforum.com/Kforum/inAction.htm • Semantic Analysis – Focus on predicates (Slotta et al., 1995, 2006)
- 19. Conclusion • A single analysis or technique cannot do it • Technology alone cannot do it • Need for collaboration across teams