Abstract
Argumentation is an important aspect of the scientific education. The purpose of the present studies was to explore the influences of two variables termed 'the argumentation strategy' and 'the kind of science concept' on student's argumentation in the web-based learning environment. There were two types of argumentation strategies (individual or collaborative) and two kinds of science concepts (descriptive or theoretical) identified for our experimental investigations. The quasi-experimental method was applied in which 138 7th grade students were invited and then assigned into either individual group (N=68) or collaborative group (N=70) randomly. They then participated into a web-based learning pro-gram we developed for six science topics, including three descriptive and three theoretical topics. Several scaffoldings were embedded into the program explicitly and systematically, such as multi-media, a discussion block, and scripts in order to promote students' learning engagement. The students' arguments pro-posed during the web-based learning, including claims, warrants, backings and rebuttals were clarified, coded, and then analyzed. The results showed the argumentation strategy has a significant effect on students' argumentation only in the theoretical topics. The post-hoc analysis showed the students in the collaborative group outperformed their peers significantly in the individual group. In theoretical science topic, the students needed more scaffoldings to acquire an in-depth understanding which supports them to construct multiple types of arguments. Moreover, the students in both groups made significant improvements no matter for the descriptive or the theoretical topics, implying the web-based learning pro-gram was effective for argumentation learning.