Paper Summary
Share...

Direct link:

Using Prompts to Scaffold Metacognition in Applied Problem Solving

Sun, April 19, 12:25 to 1:55pm, Marriott, Floor: Sixth Level, Northwestern/Ohio State

Abstract

Objectives: Metacognitive abilities play an important role in applied problem-solving, especially among novice learners or when faced with unfamiliar problem situations (Davidson, Deuser & Sternberg, 1994). Essential metacognitive abilities include here, e.g., problem identification, planning and monitoring progress and evaluating all stages of the problem solving process. Here, computer-based learning environments could contribute to support metacognitive processes during problem solving within certain domains (e.g., mathematics, cf. Lazakidou & Retalis, 2010). Within our approach we analyzed the influence of metacognitive prompting during a computer-based learning environment designed to acquire applied problem solving skills within the domain of educational psychology. While prompting has to be shown an effective instructional device (cf. Bannert, 2009), we examined here also the use of metacognitive strategies as a kind of habitus. The main objectives of this research was to find out how the use of metacognitive learning strategies influences applied problem solving skills and how it interacts with the use of metacognitive prompting as a direct instructional intervention. We expect here that the use of metacognitive learning strategies will lead to increased performance within the learning environment. Metacognitive prompts are only expected to be helpful when the use of metacognitive learning strategies is rather low.

Methods: Within a one-factorial experiment we compared the influence of prompting on the problem-solving quality and knowledge acquisition within a computer-based learning environment about the topic “Attribution processes and learning”. In the experimental group (n=20) each screen providing background information required for case solution was followed by a prompt asking a learner to reflect the information and their progress (cf. Azevedo & Hadwin, 2005; e.g., with regard to goal specification). The control group (n=20) did not have this intervention. Prior to the intervention a pre-test has been administered assessing participants’ prior knowledge and use of metacognitive learning strategies (using the questionnaire suggested by Wild & Schiefele, 1994).

Results and Significance: Analysis of variance with repeated measurement revealed that all participants increased significantly from pre- to post-test (η2ps = .81), nevertheless there was so significant interaction indicating that the intervention did lead to increased performance (η2ps = .005). Using linear regression analyses separate for both groups revealed a positive significant model for the experimental group with pre-test performance and use of metacognitive strategies as significant predictors for post-test performance (R2= 0.57). With regard to the control group, only pre-test performance was a significant predictor in the model (R2= 0.57) and the use of metacognitive strategies failed a significant level (p= 0.36). This result is rather surprising, because it reveals an Aptitude-Treatment-Interaction-Effect opposite to what we expected. It seems that the prompting did not work as compensation for learners with rather low use of metacognitive learning strategies but rather facilitated and activated those learners that already used metacognitive strategies. The findings implicate that it might be helpful to provide a solid repertoire in metacognitive abilities first (e.g., by means of training) and then, as a second step, provide scaffolding approaches like prompting.

Authors