Problem-solving strategies

Automatically Diagnosing Learners' Strategies for Solving Algorithmic Problems in Lower Secondary Education

Computer Science tutors are frequently confronted with questions like "What way did learners use to reach their solution approaches?" or "What did programming novices think while taking their last implementation steps?". Answering such questions will help to support learners on their further way to a solution or to make a more differentiated evaluation of their solution approaches than just stating "completely correct" respectively "wrong". Searching answers to these questions previous studies in this field apply methods which only enable a subsequent analysis of the learner’s solution process at a significant cost of time. In this work a procedure is developed by which an automated online identification of learners’ problem solving strategies is realizable when they complete algorithmic tasks using the programming environment "Kara, the programmable ladybug". The basis for this is provided by methods and algorithms out of the field of pattern recognition, which have been applied successfully in automatic speech recognition for a long time. These are adapted to the problem set here and embedded in the programming environment as a separate module. A study validating the results of the developed identification tool shows very high agreement with the findings of human observers. Based on the results of this analyzing tool individualized system feedback is designed adapted to the learners’ problem solving strategy, which supports learners on their further way to a solution. With regard to aspects of the attribution theory the feedback messages are optimized for maintaining learners’ motivation during their whole problem solving process. While developing the methods and tools a high transferability to other programming environments was ensured as far as possible. What is also easily conceivable is applying these methods with programming environments out of the area of computer science. In addition the developed procedure is employable in further studies to analyze the dependency of learners’ preferred problem solving strategies on factors like their age, their gender or their performance in other subjects.

Team members

Dr.-Ing. Ulrich Kiesmüller

Dr.-Ing.

Position:
former Academic Staff
Homepage:
Telephone:
+49 9831-88319-0
Fax:
+49 9831-88319-30
Prof. Dr. Torsten Brinda

Prof. Dr.

Email
Room:
SA-222
Telephone:
+49 201 18-37248
Fax:
+49 201 18-36897

Publications

  • Kiesmüller, Ulrich: Prozessbegleitende automatisierte Identifizierung von Problemlösungsstrategien Lernender beim Lösen algorithmischer Probleme mit Programmierumgebungen in der Sekundarstufe I (3). Shaker Verlag, Aachen 2013. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Brinda, T.: Automated Online Identification of Learner Problem Solving Strategies - A Validation Study. In: Bezáková, D.; Kalaš, I. (Ed.): Informatics in Schools: Situation, Evolution and Perspectives – Proceedings of the 5th International Conference on Informatics in Schools (ISSEP 2011). Comenius University, Bratislava 2011. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Sossalla, S.; Brinda, T.; Riedhammer, K.: Online Identification of Learner Problem Solving Strategies Using Pattern Recognition Methods. In: ACM (Ed.): Proceedings of the 2010 ACM SIGCSE Annual Conference on Innovation and Technology in Computer Science Education (ITiCSE 2010). ACM Press, New York 2010, p. 274-278. LABEL-FOR-eidrisexport
  • Kiesmüller, U.: Automatisierte, prozessbegleitende Identifizierung der Problemlösestrategien Lernender unter Verwendung von Mustererkennungsmethoden. In: Diethelm, I.; Dörge, C.; Hildebrand, C.; Schulte, C. (Ed.): Didaktik der Informatik - Möglichkeiten empirischer Forschungsmethoden und Perspektiven der Fachdidaktik (Didaktik der Informatik - Aktuelle Forschungsergebnisse. 6. Workshop der GI-Fachgruppe "Didaktik der Informatik"). Köllen, Bonn 2010, p. 93-104. LABEL-FOR-eidrisexport
  • Kiesmüller, U.: Diagnosing Learners' Problem Solving Strategies Using Learning Environments with Algorithmic Problems in Secondary Education. In: ACM Transactions of Computing Education, Vol 9 (2009) No 3, p. 1-26. LABEL-FOR-eidrisexport
  • Kiesmüller, U.: Prozessbegleitende, automatisierte Identifizierung der Problemlösestrategien von Lernenden beim Lösen algorithmischer Probleme. In: Koerber, B. (Ed.): Zukunft braucht Herkunft - 25 Jahre "INFOS - Informatik und Schule". Köllen, Bonn 2009, p. 169-180. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Brinda, T.: Automatically identifying learners' problem solving strategies in-process solving algorithmic problems. In: ACM (Ed.): Proceedings of the 14th annual ACM SIGCSE conference on Innovation and technology in computer science education 2009, Paris, France July 06 - 09, 2009 (14th Annual Conference on Innovation and Technology in Computer Science Education (ITiCSE 2009), Paris). ACM Press, New York 2009, p. 354-354. LABEL-FOR-eidrisexport
  • Kiesmüller, U.: Diagnosing Learners' Problem Solving Strategies Using Learning Environments with Algorithmuc Problems in Secondary Education. In: Pears, A.; Malmi, L. (Ed.): Proceedings of the 8th International Conference on Computing Education Research - Koli Calling 2008. Uppsala Universitet, Uppsala, Schweden 2009, p. 16-24. LABEL-FOR-eidrisexport
  • Kiesmüller, U.: Automatisierte Identifizierung der Problemlösestrategien von Programmieranfängern in der Seundarstufe I. In: Brinda, T.; Fothe, M.; Hubwieser, P.; Schlüter, K. (Ed.): Didaktik der Informatik - Aktuelle Forschungsergebnisse - 5. Workshop der GI-Fachgruppe "Didaktik der Informatik". Köllen, Bonn 2008, p. 33-42. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Brinda, T.: Diagnosing problem solving strategies of programming novices in secondary education automatically?. In: IFIP (Ed.): Proceedings of the Joint Open and Working IFIP Conference on ICT and Learning for the Net Generation - LYICT 2008. 2008. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Brinda, T.: How Do 7th Graders Solve Algorithmic Problems? A Tool-Based Analysis. In: ACM (Ed.): roceedings of the 13th Annual Conference on Innovation and Technology in Computer Science Education (ITICSE 2008, Madrid). ACM Press, New York 2008, p. 353-353. LABEL-FOR-eidrisexport
  • Kiesmüller, U.; Brinda, T.: Werkzeuggestützte Untersuchung der Vorgehensweisen von Lernenden beim Lösen algorithmischer Probleme. In: Eibl, C.; Magenheim, J.; Schubert, S.; Wessner, M. (Ed.): Die 5. e-Learning Fachtagung Informatik - DeLFI 2007. Köllen, Bonn 2007, p. 295-296. LABEL-FOR-eidrisexport

Talks

  • Kiesmüller, U.: Prozessbegleitende automatisierte Identifizierung der Vorgehensweisen von Lernenden beim Lösen algorithmischer Probleme in der Sekundarstufe I - Eingeladener Vortrag; Herbsttagung des MNU-Landesverbandes Franken, 17.09.2009, Universität Bayreuth, Bayreuth.