Using Sequential Pattern Mining to Understand How Students Use Guidance While Doing Scientific Calculations

Abstract

AbstractIn natural science education, experiments often lead to the collection of raw data that need to be processed into results by doing calculations. Teaching students how to approach such calculations can be done using digital learning materials that provide guidance. The goal of this study was to investigate students’ behaviour regarding the use of guidance while doing scientific multi-step calculations, and to relate this behaviour to learning. Sequential pattern mining was used to i) identify students’ behaviour patterns while doing calculations in an online learning environment, ii) study the relation between use of guidance and success on first attempt at submitting a calculated value, iii) study the relation between students’ use of guidance and learning gain, and iv) study the relation between students’ use of guidance and prior knowledge. Data showed that all students frequently used the guidance provided in the learning task. Moreover, students who used the option to check their intermediate calculations and students who studied worked examples were more likely to successfully complete the calculation on their first attempt than students who did not use this guidance. Guidance in the form of hints was used frequently. However, using the hints did not result in more success at the first attempt. We did not find a relation between learning gain and use of guidance, but we did find a trend that students with a low prior knowledge used more guidance compared to students with a high prior knowledge. The results of this study imply that providing hints and intermediate calculations is of utmost importance for students to independently complete scientific multi-step calculations.