Building: Pinnacle
Room: 2F-Harbourside Ballroom
Date: 2016-07-03 03:30 PM – 05:00 PM
Last modified: 2016-05-22
Abstract
Introduction
The competency to solve problems collaboratively within a group, defined as collaborative problem solving (CPS), is a key 21st century skill, and contributor to university and career readiness (Rosen & Tager, 2013). International large-scale education assessments recently incorporated CPS competency testing to internationally evaluate the extent of CPS proficiency in students. The Programme for International Student Assessment integrated CPS assessment firstly in 2015, and measured CPS by twelve conceptually different CPS skills assessed by a computer-simulated human-agent (H-A) approach, in which students communicate via predefined messages with agents while solving tasks in collaboration.
Objectives
As CPS assessment is a new research field, and the H-A approach conducted in a large-scale setting in PISA 2015, we aimed the first objective of theoretically revisiting the PISA 2015 approach and underlying 12-cells matrix, and further evaluate strengths and weaknesses in comparison to fellow large-scale approaches. In our second objective, we aimed to empirically embed an instrument for measuring collaborative behaviour (COLBAS) (Krkovic, Wüstenberg & Greiff in press), which covers partial aspects of the PISA 2015 approach, into the PISA 2015 H-A approach.
Design/Methodology
For the first objective, COLBAS was theoretically mapped onto the PISA 2015 CPS approach and specific partial aspects identified. For the second objective, COLBAS was empirically embedded by applying a latent class analysis (LCA) to an existing dataset of COLBAS results from 484 seventh graders (Krkovic, Wüstenberg & Greiff in press).
Results
Results have shown, that the PISA 2015 CPS approach seems reliable and entails key strengths in comparison to fellow approaches. Further, the LCA classified students to (1) compensating, (2) efficient and (3) low-performing (non-) collaborators of CPS behaviour on the basis of cognitive performance and motivational differences, which might be types of collaborators in PISA 2015.
Conclusion
In conclusion, as the LCA provided specific information on students’ collaboration and problem solving aspects, we suggest an LCA for analysing PISA 2015 results when they are published in 2017.
References
Rosen, Y., & Tager, M. (2013). Computer-based assessment of collaborative problem-solving skills: Human-to-agent versus human-to-human approach. Research & Innovation Network, Pearson Education.