On pre-test sensitisation and peer assessment to enhance learning gain in science education

*The main part of this thesis focuses on designing, optimising, and studying the embedding of two types of interventions: pre-testing and peer assessment, both supported by or combined with ICT-tools. * Pre-test sensitisation is used intentionally to boost the learning gain of the main intervention, an interactive, multimodal learning environment, designed for the pre-training of science concepts . The results show a high learning gain, especially after applying a pre-test by interaction of the pre-test with the intervention. The learning gain is negligible if no treatment follows the pre-test. The pre-test effect of multiple choice questions is the same as the effect short answer questions. *The gain of writing a scientific report in combination with doing a peer assessment gives an effect size of d=0.876. No learning gain was connected to the writing, whereas the computer-supported peer assessment appeared to be entirely responsible for the measured “average” learning gain with an effect size of d=1.47. *A peer assessment of a complete paper-and-pencil test show an “average” learning gain. When students apply explicit scoring criteria to authentic pre-selected samples of answers of peers in a digital environment a high learning gain was found when students made a pre-test. *In a simulation of chemical reaction kinetics combined with peer support and pre-test sensitisation high learning gain is found. After two months, the effect of pre-testing is still significant. *An alternative is presented to gauge the effectiveness of educational arrangements based on a experimentally found strong power law relationship yi = xi1-B between the pre-test values xi and post-test values, yi. The practical advantage of the proposed method is experimenting with a very small number of participants.